University of Southern California Dissertations and Theses

Integrated geochemical and hydrodynamic modeling of San Diego Bay, California

(USC Thesis Other)

Integrated geochemical and hydrodynamic modeling of San Diego Bay, California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content INTEGRATED GEOCHEMICAL AND HYDRODYNAMIC MODELING OF SAN DIEGO BAY, CALIFORNIA by Jian Peng A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment o f the Requirements for the Degree DOCTOR OF PHILOSOPHY (GEOLOGICAL SCIENCES) May 2004 Copyright 2004 Jian Peng Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 3140535 INFORMATION TO USERS The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. UMI UMI Microform 3140535 Copyright 2004 by ProQuest Information and Learning Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code. ProQuest Information and Learning Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements I would like to thank Dr. Teh-Lung Ku for his patience and guidance over the years. I am really lucky to have him as my advisor, because I have not met a better person in terms o f kindness, grace, and knowledge. Dr. Douglas Hammond has provided equal amount o f advice and guidance as Dr. Ku. His skeptical eyes had pushed me to do better, and his critics greatly helped me improve my thesis. Drs. Teh-Fu Yen and Lowell Stott, as the other two members in my Ph.D dissertation committee, have been helpful throughout the years. The classes I took from Dr. Yen’s Department o f Civil Engineering were almost enough for another M aster’s degree. Dr. Stott sparkled my interest in paleoceanography so much that I almost chose paleoceanography for my Ph.D dissertation. Special acknowledgement is given to Dr. Eddy Zeng o f Southern California Coastal Water Research Project (SCCWRP). He was once a member in my Ph.D dissertation committee but was unfortunately kicked out due to scheduling conflicts. However he has helped me both academically and professionally, including helping me secure a job in SCCWRP. I also want to thank Dr. Ralph Cheng o f USGS for sharing his TRIM model with computer codes and other files. His advice for me to get a driver’s license (i.e. to take a course on fluid mechanics) before he lent his car (TRIM model) to me proved to be essential in the long run. I also thank Dr. Pei-Fang Wang o f SPAWAR o f US Navy for transferring the TRIM files to me. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Michelle Chambers o f Western Regional Climate Center provided climate data o f Lindburg Field for the past few years, Michael Garrod o f Sweetwater Authority gave me the flow data on the Sweetwater River sinee 1980, and Bonnie Baker o f Cabrillo National M onument (CNM) gave me permission to plant my atmospherie fallout eolleetors in the pristine area o f Point Loma. They are gratefully acknowledged for their support. My field samplings were managed by Dario Diehl o f SCCWRP. Crew members of Sea Watch o f USC, and o f Metro and M onitor II o f the City o f San Diego were thanked for their assistanee during many sampling cruises. David Tsukada o f SCCWRP and Hongyu Huang helped with sample proeessing and analyses. Miguel Rincon performed ICP-MS analyses for trace metals and Pb-recovery measurements. Drs. Hongchun Li and Shangde Luo are my brothers as well as semi-advisors. Hongchun helped me through some very difficult times. Shangde is one o f the most underrated researchers in the academic world, and I was lucky to have learned a lot from him. Cindy Waite, Vardui Ter-Simonian, John Yu, and Barbara Grubb in the main office o f the Department o f Earth Sciences are thanked for their assistance on bureaucratie. 1 1 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. computing, and teaching affairs. Dr. Don Gorsline was thanked for a couple o f free football tickets that turned me into a fervent fan o f Trojan football. Many other faculties, staff, and grad students are also thanked here for their kindly support over the years that made my time at USC one o f the most memorable period in my life. This thesis is dedicated to my family, especially to my wife Yaling (Jacline) Teng. She was very much like a single mom during the final period o f my dissertation, taking care o f the whole family in San Diego when I was in Los Angeles in weekdays and sometimes during weekends. As a competent hiostatistician, she even helped me with the statistical analyses o f my data. IV Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents Acknowledgements ii List o f Tables x List o f Figures xi Abstract xii Chapter 1 Introduction 1 1.1 General Statement 1 1.2 The use o f radionuclides as chemical tracers 3 1.3. Previous studies on particle dynamics and particle- radionuclide interactions 10 1.4 Previous studies on hydrodynamics o f coastal seas 16 1.5 Recent progresses hydrodynamic modeling o f chemical species in coastal seas 19 1.6 Outline o f the dissertation 20 Chapter 2 Study Site 22 2.1 Introduction 22 2.2 Geological setting 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.3 Climate and hydrography 25 2.4 Historical pollution o f SDB 28 Chapter 3 Sampling and Analytical Methods 31 3.1 Sample collection 31 3.2 Sample processing and analyses 46 Chapter 4 Results 52 4.1 Results o f June-July 1999 sampling cruises 52 4.1.1 CTD Profiles 52 4.1.2 Activities o f total ^'°Po and ^*^Pb 53 4.1.3 Water contents o f sediment cores 55 4.1.4 TOC-TON profiles o f sediment cores 56 • • 9 1 n 4.1.5 Profiles o f activities o f excess Pb, ^^"^Th and ’^’Cs in sediment cores 61 4.1.6 In-situ pump sampling o f suspended particles and PCBs 65 4.2 Results o f winter 2000 sampling cruises 69 4.2.1 Activities o f dissolved and particulate ^ 'V o a n d ^ 'V b 69 4.2.2 Result o f three sediment traps deployed during Feb. 17-Feb.29 71 4.3 Results o f samplings around SDB in June- December 2002 74 V I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.4 Results o f September 6, 2002 sample cruise 82 4.5 Results o f tidal cycle samplings 93 4.5.1 Result from July 6-7 sampling trip 93 4.5.2 Variation o f water column TSS, V o, and with tidal phase. 95 • 91 0 4.6 Atmospheric fallout o f Pb measured by collectors and salt marsh sediment cores 100 4.7 Salinity and trace metal distributions and variations with tides 102 4.8 Summary of geochemical studies 108 Chapter 5 Numerical Modeling o f Tidal Hydrodynamics o f San Diego Bay 110 5.1 Introduction 110 5.2 Description o f TRIM model 113 5.2.1 Goveming equations 114 5.2.2 The numerical model 119 5.2.3 ELM for treating the convective terms 126 5.3 Modifications of the model parameters 130 5.4 Schemes o f geochemical modeling o f non conservative radionuclides in San Diego Bay 134 5.4.1 Goveming equations for geochemical modeling 135 5.4.2 Quantification o f non-conservative terms for radionuclides 13 8 Vll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.5 Particle dynamics modeling 139 5.5.1 Stochastic processes as sources o f suspended sediments 141 5.5.2 Sediment resuspension due to tidal currents 143 5.5.3 Settling o f suspended particulate materials 145 5.6 Scheme o f simulation o f particle scavenging of radionuclides 148 5.7 Coupling o f sediment and water columns 153 5.8 Results o f geochemical modeling o f SPM and Particle- reactive radionuclides ^’°Pb, ^^Vo, and 155 5.8.1 SPM simulation result 156 5.8.2 Bottom shear stress simulation 159 5.8.3 Simulation results o f radionuclides 161 5.8.4 Simulation results o f precipitation effects 172 5.8.5 Simulation result o f scavenging and deposition 174 5.9 Concluding remarks on the results o f geochemical- hydrodynamic modeling 176 Chapter 6 Discussions 179 6.1 General statements 179 6.2 Mass balances o f water, ^'®Pb, ^'^Po and ^^‘ ^Th in San Diego Bay 180 6.2.1 Water balance in San Diego Bay 182 vm Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.2.2 Mass balance o f 185 6.2.3 Mass balance o f 190 6.2.4 Mass balance o f suspended particulate matter in SDB 193 6.3 Residence times o f ^'^Pb and in San Diego Bay 195 6.4 Recycling o f ^'V o in water eolumns 198 6.5 Stripping and trapping o f ^'®Pb inside SDB 200 6.6 Environmental significance o f particle scavenging and tidal exchange 201 6.7 Future studies 203 6.8 Summary and conclusions 205 References 208 Appendix I List o f stations around San Diego Bay sampled during June-December 2002 225 Appendix II RS-TRIM program flow chart 230 IX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Chapter 3 Table 3-1 Summary o f all sampling cruises/trips Table 3-2 Water sample collection log during summer 1999 Table 3-3 June 25, 1999 box core collection log and sample description Table 3-4 Description o f filter-retained particles in water samples o f September 6, 2002 cruise 31 35 37 44 Chapter 4 9 1 n Table 4-1 Sedimentation rates and enrichment factors o f Pb for summer 1999 sediment cores Table 4-2 Summary o f sediment trap data 64 72 Chapter 5 Table 5-1 Harmonic constants for the 16 major harmonic constituents in San Diego Bay Table 5-2 List o f model parameters used by other researches 131 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Chapter 1 Figure 1-1 The role o f suspended particles in estuarine biogeochemical processes 11 Chapter 2 Figure 2-1 Sketch o f San Diego Bay area 23 Figure 2-2 M ajor faults near San Diego Bay area 24 Chapter 3 Figure 3-1 Locations o f stations for sampling cruises during summer 1999 and winter 2000 34 Figure 3-2 Locations o f stations around SDB sampled during June-December 2002 40 Figure 3-3 Station locations for September 6, 2002 sampling 43 cruise. Chapter 4 Figure 4-1 CTD data for station 5 (a) and station 3 (b) 52 X I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ABSTRACT San Diego Bay (SDB) is a typical semi-closed, low inflow estuarine. Our previous geochemical studies have shown that there are systematic differences between waters inside SDB and in the open ocean in terms o f activities o f particulate and dissolved 910 910 994 particle-reactive radionuclides ( Pb, Po and Th) and suspended particulate matter (SPM). The difference in SPM concentrations is found to have caused the distinction between the two waters. Vigorous tidal exchange has caused temporal variations o f the above chemical species and SPM concentrations, especially in north SDB. Sediment ^'°Pb inventory in north SDB was more than 12 times higher than the atmospheric input. A process we defined as “stripping” may be responsible for the extraordinarily high inventory. It is a process in which seawater with higher dissolved 91A Pb activity from the open ocean mixes with particle-abundant SDB water through tidal exchange and is scavenged efficiently to the sediment. To establish a connection between the behavior o f particle-reactive radionuclides and tidal hydrodynamics, an existing 2-D hydrodynamic model is modified and integrated with geochemical and particle dynamic sub-models. It is the first study on the hydrodynamic modeling o f naturally occurring, particle-reactive radioisotopes, which have been used as tracers as oceanic processes for decades. The geochemical and particle dynamic submodels include sediment resuspension, settling, and transport; xii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. particle scavenging o f radionuclides, river and atmospheric input o f freshwater, radionuclides and SPM, radioactive decay, and in-situ production. The scavenging o f radionuclides by particles is simulated as an equilibrium process in which the distributions o f radionuclides are determined by distribution coefficient Kd. Equipped with the boundary conditions provided by the previous geochemical studies, the model was able to reproduce the observed spatial and temporal variations o f SPM and radionuclides in our previous geochemical studies. Moreover, model reproduced higher inventories o f ^‘* ’Pb in north SDB, indicating that the model is functioning correctly over a long time-scale. This study is a pioneering effort to the geochemical-hydrodynamic modeling o f a dynamic coastal environment. The novel modeling approach improved our understanding o f the behavior o f non-conservative, particle-reactive radiotracers and expanded the potential application o f these radiotracers in the studies o f coastal environments. Xlll Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1 Introduction 1.1 General Statements Coastal seas and estuaries play an important role in global biogeochemical cycles by transporting terrestrial materials (fresh water, suspended solids, nutrients and contaminants) to the ocean. Coastal seas are the boundaries between land and sea and the underlying continental shelves consists o f mostly terrestrial sediments. Due to their shallow depth, strong interaction with bottom sediment, and physical and chemical impacts from terrestrial sources, coastal seas differs significantly from open seas in physical, chemical and biological aspects. The physics o f coastal ocean is characterized by vigorous wave action, tide, strong bottom friction and sediment resuspension. Chemically, coastal ocean is strongly influenced by dissolved and particulate chemicals with terrestrial sources, and its salinity, oxygen level, and many trace elements or compounds may bear significantly different patterns from those of the open ocean. Due to high nutrient load, coastal ocean has much higher productivity, which affects its physical and chemical properties (Broecker and Peng, 1982). Radioisotopes such as ^'®Pb, ^^'^Po, and ^^"^Th have been used in the studies o f coastal oceans (Broecker and Peng, 1982; Bhat et al, 1969; Moore, 1992; Cochran, 1992) due to their short half-lives and strong particle affinity. Short half-lives o f these Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radionuclides allows them to trace the transient oceanic processes in the coastal seas. Their particle affinity helps their role as a chemical tracer for particle dynamies. Parallel to traditional physical and geochemical studies, numerical modeling o f the coastal oceans has seen fast progresses in recent years (Martin and McCutcheon, 1999; James, 2002, and references therein). Numerical modeling o f oceanic processes is achieved by solving hydrodynamic equations using computer programs in 2-D or 3- D domain. The underlying hydrodynamics in oceanic processes is largely well understood. There are still a number o f issues specifically related to the numerical solutions to the hydrodynamic equations, such as the selection o f 2-D or 3-D representation, spatial and temporal resolution, approximation and simplification assumptions, and the stability and convergence o f numerical solutions (Martin and McCutcheon, 1999; Gerald and Wheatley, 1994). Numerical modeling provides a framework o f physical processes, based on which a number o f other oceanic processes can be studied, including chemical, biological, biogeochemical processes, and sediment dynamics. However, there are only limited studies on coupled geochemical and hydrodynamic modeling o f reactive chemical species in shallow seas (James, 2002 and references therein; Tang and Adams, 1998; Dortch, 1998; Sheng, 1996), and no such studies on particle-reactive, non-conservative 234 210 210 radionuclides such as Th, Pb, and Po. This is in dire contrast with the extensive studies and applications o f these radionuclides as chemical tracers for a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. wide range o f oeeanic processes (Bacon et al, 1988; Baskaran et al, 1997; Benninger, 1976; Broecker et al, 1973; Cherry et al, 1975; Cherry and Heyraud, 1979; Cochran, 1992; Kadko, 1993; Kaufman et al, 1981; Martin et al, 1978a,b; Masque et al, 2002; Santschi et al, 1980; Sarin et al, 1994; Tanaka et al, 1983; Turekian et al, 1974). 1.2 The Use o f Radionuclides as Chemical Tracers The studies o f coastal oceans have benefited significantly from the use o f chemical tracers, especially uranium-series radionuclides. The radioisotopes under consideration evolve through the decay chain as follows (half lives are shown in brackets); (4.5 billion yr)-> ^^"^Th (24.1 d)-> ..(interm ediates)..^^^R n (3.8d)->.. (interm ediates)..^ ^'°Pb (22yr) ( i n t e r m e d i a t e ) . .( 1 3 8 d ) - ^ ^^^Pb (stable). U, as the ultimate parent o f all radionuclides in the decay chain shown above, is released from the continental weathering and transported through rivers into the ocean. In the solution, uranium is often eomplexed with carbonate, sulfate or dissolved organics (Langmuir, 1978) . In estuarine environments, uranium is largely mixed conservatively when fresh water and sea water meet (Martin, 1978a and 1978b). Ku et al (1977) found that there was a good linear relationship between salinity and U activity, and this relationship enables researchers to infer the activities o f U through salinity, which is easily measured. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. is strongly particle-reactive and tends to be scavenged to particles in a very short time, so its removal rates increase with decreased distance from shore (Bhat et al, 1969; Cochran, 1992). Previous studies have showed that the particle reactivity o f ^^"^Th is slightly higher than that o f ^'°Po and significantly higher than ^'®Pb (Sarin et al, 1992). Most of the applications o f ^^"^Th in the coastal oceans are to use it as an indicator of mixing o f recently deposited sediments and as a proxy for other particle-reactive species. For example, the relationship between the conservative U and particle- reactive ^^"^Th allows calculation o f the residence time o f ^^"^Th (Aller and Cochran, 1976), and the result can be used to infer the removal rate o f other particle-reactive species with similar particle affinity. In the sediment column, due to the short half-life o f it should exist only in the top few millimeters o f sediments if no mixing occurs. As a result, the existence o f ^^"'Xh in the subsurface sediment may indicate • ♦ » 9 X 4 • sediment mixing, and the mixing depth is indicated by the Th profile (Santschi et al, 1979). ^^^Ra tends to adsorb to particles in freshwater and exists in seawater primarily in 226 dissolved form. In estuarine environments, Ra activities were found to be higher 9 9 ^ « than both in river water and seawater due to desorption of Ra from river-borne particles (Moore, 1992) and from coastal sediments as well, and this is in fact a Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • • • 226 210 • significant source o f oceanic Ra. However, as a source of Pb in estuarine and coastal seas, ^^^Ra is relatively unimportant and often negligible compared to other sources (Benninger, 1978; DeM aster et al., 1986; Carpenter et al., 1981). ^'°Pb may come from several sources in either dissolved (defined hereafter as < 0.47pm) or particulate forms (>0.47pm). One important souree is from atmospheric 9 9 7 9 1 A fallout after deeaying from Rn. The atmospherie fallout o f Pb ean be measured through soil inventory or direct collection o f precipitation, and the contribution from this source can be dominant in coastal oceans (>80%, Benninger, 1978). Rivers 9 1 0 supply Pb in both particulate and dissolved forms to estuaries and coastal seas. 210 Usually river water has higher concentrations o f Pb than coastal sea due to its atmospherie origin (Benninger, 1978; Carpenter et al, 1981; Benoit, 1988). Carpenter 9 1 0 et al. (1982) measured Pb o f 26 samples in Columbia River (Washington) with an average activity o f 11.2dpm/100L, and Benninger’s (1978) measurements of Connecticut River gave ^'®Pb activities o f 10-80dpm/100L. The activities o f ^'*’Pb seemed to correlate with the length o f time it stays in the water column, with upstream part o f river having higher activities than that o f downstream (Rama et al, 1961). This is because ^'°Pb is constantly removed by particles in river water along its 9 1 0 way. For the same reason, lakes generally have lower Pb activities than rivers draining to them (Benoit, 1988, and references therein). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1 n Besides atmospheric fallout and river input, Pb in coastal waters can also be produced in-situ from ^^^Ra in the water column or supplied from offshore waters through mixing or coastal upwelling. However they could vary in different oceanic 91 0 settings. For example, Amazon River estuary sees 30-68% o f its Pb inventory coming from the boundary scavenging from the offshore waters (DeMaster et al, 1986). Similar findings were made for W ashington coast by Carpenter et al (1981), 9 1 n where an overwhelming portion o f Pb in the sediments was found to be from offshore waters via coastal upwelling and scavenging by iron and manganese oxides near the sea floor. The source most difficult to quantify is from offshore waters because o f the stochastic nature o f the coastal meso-scale circulation and the enormous difficulty in calculating the exact amount o f offshore water that participate in the exchange and mixing. 9 1 0 Rama (1961) and Lewis (1977) have shown that Pb is rapidly removed from 9 1 0 solution onto detrital particles, and the residence time for dissolved Pb in surface oceans can be as short as 2 days (Santschi et al, 1979). Benninger (1978) found that ^'®Pb activities in Long Island Sound correlated well with the amount o f suspended 91 0 solid in the water column, and there was virtually no dissolved Pb. Similarly, 50- 70% o f total ^'°Pb was in particulate form in Santa Barbara Basin (Krishnaswami et 910 • al, 1975). For the same reason, the dissolved phase Pb activities deerease with the distance to the shore, as shown clearly by a systematic study o f the global ocean Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Nozaki et al, 1976). Similar observation was made for the G ulf o f California and eastern North Pacific waters (Bruland et al, 1974), where total ^'°Pb concentrations decreased from 0.09dpm/L in the offshore open ocean to O.Oldpm/L in the water far inside the G ulf o f California. The profile o f excess Pb in coastal and estuarine sediment column depends on the input from water column as well as the post-depositional processes such as mixing and erosion. Due to the large portion o f atmospheric input of ^^*^Pb in the overall ^'Vb inventory in the coastal sediments, it is important to quantify the amount of atmospheric input rate o f ^'®Pb in the local environmental settings. The common procedures to measure the atmospheric input include direct collection o f atmospheric fallout or measure the sediment column inventory o f ^*‘ ’ Pb of salt marsh or forest soil, which are considered to be good ‘natural collectors’ o f atmospheric ^'°Pb (Fuller, 1982). In the calculation o f sedimentation rates in coastal environments, an accurate measurement o f atmospheric fallout rate is not necessary as long as the input o f ^'°Pb to the sediment column is largely constant (Robbins, 1978). However, if the mass 9 1 0 balance for Pb is desired, it is necessary to do the measurement. ^’°Po is produced from its precursor ^'*^Pb, but it is generally deficient relative to ^'*’Pb in the surface o f the open ocean due to its preferential removal, mainly by biogenic particles, leading to a typical ^'Vo/^*Vb ratio o f about 0.5 (Bacon et al, 1976, 1988; Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Nozaki et al, 1976; 1990, 1997, 1998; Cochran, 1992; Kim, 2001; Sarin et al, 1994). It was found that ^’°Po can be bioaccumulated by an order o f 1 0 "^ in pbytoplankton and zooplankton as opposed to the factor o f 10^ for (Heyraud et al, 1976; Cherry and Heyraud, 1979; Kadko, 1993; Radakovitcb et al, 1999). As a result, the 710 710 Po/ Pb activity ratio in the particulate fraction in surface waters is usually >1 (3 for pbytoplankton and 12 for zooplankton), while the reverse situation is observed for the dissolved fraction (Shannon et al, 1970; Turekian et al, 1974; Kbarkar et al, 1976; Masque et al, 2002). ^*^Po can also come from the atmosphere fallout following the same path o f ^'^’Ph, but there is significant disequilibrium between and ^’®Po 7 1 0 because o f the short residence time (7-40 days) o f Pb in the atmosphere (Moore et al, 1974; Turekian et al, 1977; Appleby and Oldfields, 1992;). The observed 710 7 1 0 Po/“ Pb ratios in rainfall and in ground-level air were about 0.1-0.2 (Appleby and Oldfields, 1992, Fuller, 1982). This could be one o f the reasons why there was deficiency o f ^'**Po in the surface waters (Shannon et al, 1970, Turekian et al, 1974). 71 0 7 1 0 The deficiency o f Po in the surface waters is often accompanied by excess o f Po 71 0 • relative to Pb in subsurface, where falling biological particles are recycled and ^*®Po released back into water column (Radakovitcb et al, 1999; Kadko, 1993; Masque et al, 2002). Benoit (1988) found that in an oligotrophic lake, ^’°Po was deficient relative to ^’*^P b in the epilimnion, but ^'^Po/^'^'Pb ratio ean reach as high as 7 1 0 3.8 in metalimnion in early fall. This phenomenon manifested that Po could be in Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. « 9 1 D . disequilibrium with Pb, and the disequilibrium seems to be largely controlled by biological processes. However, in the coastal and estuarine environment, especially for the shallow and dynamic coastal seas, there are relatively few published data (Masque et al, 2002; Radakovitcb et al, 1999), but a general pattern is that excess of "910 910 Po over Pb is more often observed, and lateral transport o f water becomes important in the mass balance o f these radionuclides. In this study, the relationship 9 10 9 10 between Po and Pb will be emphasized. For a given radionuclide, the time rate o f change (TROC) in a 2-dimensional domain can be expressed mathematically by the following differential equation: - ^C-q, + Pr+S (10) (I-l) Where D 5 5 d ^ ^ ^ substantial derivation in the x-y plane Ds is the diffusivity coefficient in the horizontal plane (implying good mixing and isotropic condition) X C defines radioactive decay o f the radionuclide; qs is the amount o f the radionuclide scavenged by particles Pr is the amount contributed from the decay of its parent Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S 10 is the combined terms o f sources and sinks o f the radionuclides other than specified above. They can be river or atmospheric input, bioaccumulation, chemical source/sink, etc. The mass balance considerations for each o f the above radionuclides in the coastal environments should include the above terms. Based upon a good understanding of the mass balance o f these radionuclides, oceanic processes such as mixing and circulation, sediment transport, and particle scavenging can be studied with better confidence using these radionuclides as tracers. 1.3 Previous Studies on Partiele Dynamics and Particle-radionuelide Interactions Suspended particulate matters (abbreviated as SPM hereafter) are ubiquitous in coastal waters and they play an important role in coastal biogeochemical processes (Turner and Millward, 2002, and references therein). Processes that affect the nature and amount o f SPM include particle in-situ generation/dissolution, external input, resuspension, settling, and transport. In the present study, in-situ generation and dissolution is assumed unimportant; particles are assumed to be entrained in the water during transport, so only resuspension and settling, and processes that affect them will be considered in the mass balance o f particulate matter in the water column. 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Suspended materials are the locale o f adsorption/desorption o f particle-reactive chemical species, and resuspension and settling o f suspended materials are important source and sink o f the radionuclides. As a result, particle dynamics control the source, fate and transport o f the particle-reactive radionuclides in water column and across the sediment-water interface. Biological processes such as primary productivity, nutrient cycling and biogenic detritus transport usually involve particulate matter in one way or another. However, due to the large variations in particle sizes, composition and complex environmental conditions that directly affect the source/sink and transport o f particles, quantitative studies o f particle dynamics are often difficult and almost always based on empirical approaches (Partheniades, 1965; Sheng et al, 1992; Lumborg and Windelin, 2003). diagenetic remobilization water adsorption precipitation heterotrophy precipi -tation decay, exudation desorption dissolution degradation adsorption partitioning settling sediments suspended particles capture sorting ingestion resuspension Egestion Excretion egestion, excretion suspension feeders dissolution-absorption Figure 1-1 Schematic representation o f the role o f suspended particles in estuarine biogeochemical processes (from Turner and Millward, 2002). 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. A taste o f the eomplexity o f processes involved with particulate matter is offered by the following simplified scheme (Turner and Millward, 2002) as shown in figure 1-1. There are so many processes involved that it is impossible to take into consideration all the processes as shown above in the present study. A realistic way is to single out the most important processes or to combine several processes to make them easier to handle. Empirical approximations are often used instead of theoretical calculations due to the heterogeneity o f particulate matter and complexities o f environmental parameters. Additionally, the involvement o f biological processes (through suspension feeders) is often important but it is difficult to quantify. As a result, the biological aspects in sediment dynamics are often omitted for simplicity. However, due to the largely empirical treatment o f many particle-related processes, the biological processes are in fact implied in the overall sediment dynamics. Similar approach will be adopted in the particle dynamic investigations in the present study. Sediment resuspension could result from bottom currents, bioturbation, sediment failure and wave action, or human-related activities such as dredging and boat propelling. Other particle-producing processes include primary production, chemical precipitation and coagulation o f colloids. Obviously, sediment resuspension increases the particle concentration and the concentration o f particle phase radionuclides, and may strengthen scavenging and eventually drive down the overall concentration o f particle-reactive radionuclides after settling. The difficulties in the studies o f sediment 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resuspension lie in the complex physics involved in the process. The physical parameters involved in sediment resuspension include direct impact by current, gravity, van der Waals force among particles, electro-magnetic (gluing) forces among particles, and supporting forces from underlying particles. If all o f these parameters are considered, the model becomes too complicated to be practical (Taki, 2001). As a result, sediment resuspension will be modeled empirically in the present study. The most common simplification o f the resuspension process comes from the observation that there seems to be a threshold value for the bottom shear stress beyond which resuspension will occur, and the amount o f resuspension is proportional to the ratio between bottom shear stress and a critical shear stress, with an empirical parameter that has to be determined experimentally (McDonald and Cheng, 1994; Wang, 2002). It was found through experiment (Torfs et al, 2001) that particles need a smaller shear stress to be eroded after breaking loose. When the shear stress increases further, erosion rate (in unit o f weight o f sediment/unit area/unit time) will increase linearly with shear stress (Torfs et al, 2001). For a more detailed discussion about sediment resuspension, please refer to chapter 5. With biological processes not considered explicitly in this study, the production and sink of SPM are schematically shown in equation 1-2; 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resuspension Sediment 4 ^ ----------- p. SPM (1-2) settling Resuspension and settling, as the two opposing proeesses, determine the coneentrations o f SPM o f different sizes and eompositions. Flocculation and disintegration are also important processes that change the sizes o f SPM and eventually play significant roles in the settling o f SPM and adsorption-desorption o f dissolved species. However, the latest studies o f flocculation and disintegration (Partheniades, 1993; Lick, 1994) had to rely heavily on empirical parameters. In the present study, this process is not treated explicitly. Rather, the empirical treatment of particle dynamics has taken into consideration o f the partiele-particle interaction implicitly. Particle settling is another important processes determining the SPM and radionuclide concentrations. The settling process is determined by gravity difference between sinking particles and seawater, and the velocity will be affected by particle shape, flow field, and water viscosity. The settling o f particles can largely be calculated by Stokes Law (see equation 5-26 in chapter 5) with some simplification assumptions (i.e. the shape o f particles are assumed to be spherical and water flow is laminar). However, when there is a large range o f particle sizes and composition and when the flow field in water column is not well defined, it is often difficult to calculate the 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. amount o f sediment settling even though the physical laws underlying particle settling are well-defined. As a result, empirical or semi-empirical schemes are routinely used to define sediment processes (McDonald and Cheng, 1994; Lumborg and Windelin, 2003; Le Normant et al, 1998). Due largely to the uncertainties in particle dynamics, the process o f particle scavenging o f radionuclides and other heavy metals in oceanic environments, is a subject o f ongoing studies (Broecker and Peng, 1982; Cochran, 1992, and references therein; Yen, 1999; Quigley etal, 2001). To treat particle scavenging as a chemical reaction, interactions between water and suspended sediment as shown in figure 1-1 can be written in the form o f a chemical reaction; ki M"+ + SPMi M------------------ ► (M:SPMi)"’ ' (1-3) k-i There are several processes involved in the above reactions; 1. Adsorption o f metal ions with SPM with equilibrium constant ki,2,...n for suspended particle species 1, 2, ...n; 2. Desorption o f metal from SPM with equilibrium constant k-i, k.2,...k.n for metakSPM complex M;SPMi, M ;SPM 2, ... M;SPMn; 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3. Implied in the above equation, the rate o f the adsorption-desorption reactions is determined by the rate constants for reactions between metal ions and SPM i, SPM2, ...SPMn. In practice, the scavenging process cannot simply be treated as a chemical process because o f the heterogeneity o f SPM. It is possible to study the process under ideal conditions (Hummel, 1997; Zhu and Anderson, 2002). In natural environments, it is often necessary to employ empirical approaches in order to model the processes closely and at the same time, keep the model simple enough in order to handle. This approach will be adopted and will be described in subsequent chapters. 1.4 Previous Studies on Hydrodynamics o f Coastal Seas There have been extensive studies o f coastal seas using hydrodynamic methods on processes such as circulation, tides and waves (Largier, 1995; Casulli and Cheng, 1992; Prandle, 1991; van de Kreeke, 1988; Cheng et al, 1984; Cheng et al, 1993, Feng et al, 1986a & b). In these studies, hydrodynamic calculation or numerical models were used to study current velocity, water level variation, the effects o f bottom friction on the flow field, and other related processes. Unlike the studies on the open ocean, the studies on hydrodynamics o f shallow coastal seas, must treat the sea floor as an integral part o f the model. In fact, the quality o f models is largely differentiated by their treatments o f the interactions between water flow and sea floor. The reason 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. for this is that the hydrodynamic models o f coastal oceans have to be based on the equations o f momentum conservation (Navier-Stokes equations) besides continuity equation. In the momentum equations, bottom friction is usually the most important components o f the net forces exerted on the parcel o f water. As the result, the flow processes can vary from weakly nonlinear to highly nonlinear (Feng et al, 1986a). Recent developments o f physical oceanographic studies o f coastal oceans have benefited from the numerical hydrodynamic models, which in turn benefited from recent development in computing capabilities. With hydrodynamics and fluid mechanics o f seawater relatively well established, the limitations o f hydrodynamic modeling are largely from computing power, cost o f computing and efficiency o f simulation. The performance o f the model also is affected by its complexity, e.g. grid sizes, time interval, the use o f Lagrangian scheme to track the movement o f water parcels, and the numerical schemes to solve sparse matrices o f enormous sizes for a prolonged time. Obviously, many estuarine hydrodynamic issues can be better addressed by three-dimensional. Future research in estuarine modeling will soon make robust and efficient three-dimensional models available to replace the two- dimensional depth-averaged models in estuarine hydrodynamic studies. However, all current three-dimensional hydrodynamic models suffer a huge drawback o f high computing cost and low simulation efficiency, i.e. runtime/real-time ratio, which renders them impractical for simulation o f coastal processes for longer time scale. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In lieu o f a full three-dimensional estuarine hydrodynamic model, the needs of estuarine hydrodynamic research can largely be met by a depth-averaged two- dimensional model. However, the model must be robust and efficient, so that fine spatial resolution can be introduced to properly represent the usually complex basin bathymetry. Shallow coastal estuaries are usually well-mixed vertically; however, the range o f salinity can vary between the values o f fresh water and that o f the oceanic water along the axis o f the estuary. For the low-inflow estuaries, hypersaline condition might result in the inner side o f the estuary in summer season due to excessive evaporation (Largier et al, 1997). The shallow shoal are often exposed at low tides and submerged at high tides. Thus the treatments o f baroelinic forcing and wetting and drying o f shoals in a tidal cycle are two o f the most important issues that need special treatment in 2-D depth-averaged hydrodynamic models. With these issues resolved, 2-D models can be computationally more efficient to enable simulations spanning much longer time scales. O f course, like all numerical models, there should be rigorous model calibration and verification processes to ensure the model predict the observational data reasonably well (Cheng et al, 1993; Wang et al, 1998). In the numerical methodology aspect, depth-averaged tidal flows in coastal estuaries and tidal embayments are governed by the shallow-water equations that are usually solved by time-stepping methods. The most commonly used method o f solution for 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the shallow-water equations has been a form o f alternating-direction-implicit (ADI) method, which is a compromise between the explicit and implicit method (Casulli, 1990). In recent years, more robust solution techniques for shallow water flows, which are economically competitive with ADI methods, have been developed and used in practical applications. These methods include a semi-implicit formulation (Casulli, 1990) as well as fully implicit, time-splitting methods (Officier et al, 1987). 1.5 Recent Progresses in the Modeling o f Chemicals in Coastal Seas The modeling o f conservative chemical species is relatively well-established based on hydrodynamic models. Recently, there has been increasing interest in the applications o f hydrodynamic modeling on the studies o f non-conservative chemical species, especially those related to environmental issues (Liu et al, 1998). Broecker et al (1973) pioneered the idea o f using particle-reactive radionuclides to mimic the activities of pollutants that have similar particle affinities. Due to their different half- lives, these radionuclides provide good handle for oceanic processes o f different time scales. So they prove to be a valuable tool o f studying coastal biogeochemical processes as well as coastal environmental issues. In our previous work (Zeng et al, 2002, Peng et al, 2003) we employed this idea on the environmental issues in San Diego Bay by using particle-reactive radioisotopes ^'V b, ^'‘ *Fo and ^^'^Th as proxies o f polychlorinated biphenyls (PCBs). Many contaminants, due to their similarities in particle affinity with particle-reactive radionuclides, can be reasonably mimicked by 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. these radionuclides in terms o f particle scavenging, settling, and transport with suspended sediments. In fact, the use o f naturally-occurring, particle-reactive radionuclides has the advantage that their source and sink terms are more easily defined than the man-made contaminants, so the factors that influences their dispersion and transport can be defined with more confidence. However, there has been no studies up to now on the hydrodynamic modeling o f particle-reactive, naturally-occurring radionuclides in coastal waters despite their established value as oceanic tracers, perhaps due to a requirement o f interdisciplinary knowledge on chemical and physical oceanography, fluid mechanics, numerical modeling, and particle dynamics. So, the present study represents the first attempt to use a hydrodynamic model to simulate the behavior o f particle-reactive radionuclides. The simulation o f particle dynamics will be an integral part o f the model. It is hoped that this study will shed light on further studies on other chemical tracers or contaminants in coastal environments. 1.6 Outline o f the Dissertation In this thesis, 1 established a coupled geochemical-hydrodynamic model that combines geochemical behaviors o f naturally occurring, particle reactive radionuclides with a benchmark numerical hydrodynamic model. The geochemical model was based on my study o f the profiles o f ^'°Pb, ’^’Cs, and ^^"^Th in the sediment, as well as the distribution and transport o f particulate and dissolved ^’* ’Pb, 20 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ^’^Po and in water columns o f San Diego Bay, a semi-closed embayment in southern California. The organization o f the chapters is as follows. First, an introduction to the study site, San Diego Bay, is given in chapter 2, then the methods and results o f the geochemical studies are presented in chapter 3 and chapter 4, respectively. In chapter 5, the hydrodynamic model used in this study is introduced with an outline o f its governing equations and numerical schemes, then the geochemical and particle dynamics modules are described in detail, followed by the modeling results. Chapter 6 presents discussions o f important issues, including mass ‘ 7 1 0 7Td. balance calculations for Pb and Th for San Diego Bay, followed by a summary and conclusions o f the overall study. 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2. Study Site 2.1 Introduction San Diego Bay (abbreviated as SDB hereafter) is a semi-enclosed embayment on the southwestern corner o f California near the US-Mexican border. The bay is roughly 25 km along its axis and varies in width from 0.5 to 4 km. The area o f the bay is approximately 4.3x10^ m^ at mean lower low water with an average depth o f approximately 6.5 m (Wang et al, 1998). The portion of the bay south o f the Coronado Bridge is commonly referred to as south bay, and the region north o f the bridge to the bay’s coimection to the Pacific Ocean at Zuniga Point is called north bay. Sometimes the term o f “central bay” is used to indicate the portion o f SDB near the Coronado Bridge, ranging from the Seaport Village Restriction to the north and the Naval Stations to the south. The south bay is broader and shallower than the remainder o f the bay. Water depths range from 1 to 4 m outside o f the main shipping channel, which is dredged to a depth o f about 12 m (figure 2-1). The bay, bordered by four cities including San Diego with an estimated total population o f 1.2 million, supports a large number o f recreational, commercial, and naval facilities. 2.2 Geological setting The geology o f the San Diego area is to a large extent dictated by plate tectonics due to its location on the passive plate boundary between Pacific and North American 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Point Loma S an D iego Airport k - , > H artiO f (slan^ North SDB A ircraft C m m r Cruiseiir«:r Dovwjtown San Diego V North Island Naval Base Z o d ig a P o in t Paiteta I Creek u > \ M % \ \ V CtioKa® Creek Sweetwater > City of Chula Vista Pacific Ocean I Otay River n juana Rivar Figure 2-1. Sketch o f San Diego Bay area, indicating major geographic features plates (Abbot, 1999). The available geological records trace only as far as Jurassic, when most o f North America was under shallow water, and volcanic activities were frequent, persisting until the Paleocene and formed plutonic-rock-dominated Peninsular Ranges and their volcanic-rock foothills to the west. Since the Paleocene, 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. due to the decreasing angle o f subduction, volcanic activities largely moved further east. Since then, the San Diego area was dominated by erosion, which, along with frequent sea level changes until the late Quaternary and continuous uplift o f the land, produced the characteristic marine terraces and sea cliffs along Southern California coastline, including the San Diego area (Abbot, 1999). The topography for most o f the San Diego region is characterized by uplifted sea floors sloping genfly to the west and dissected by west-flowing rivers that have carved significant canyons. But the regional pattern is disrupted by the Rose Canyon fault zone, which is part o f the San Andreas Fault zone. The Rose Canyon fault zone seems to die out in San Diego Bay (Abbot, see figure 2-2). As a result, the tectonic Figure 2-2. M ajor faults near San Diego Bay area 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. pressures are passed through a major right step with transfer o f movement to the offshore Descanso fault (Figure 2-2). San Diego Bay should be the result o f this stretching, as are other topographical features in the vicinity (Abbot, 1999). 2.3 Climate and Hydrography The climate o f the San Diego region is characterized as semi-arid or Mediterranean- type. Annual precipitation is about 25 cm and occurs almost exclusively between November and April. Evaporation exceeds precipitation in all seasons except for winter and totals approximately 160 cm per year (Chadwick et al, 1999). Air temperature ranges from 14.4 to 22.2 °C throughout the year. Creeks and rivers feed fresh water and particulate materials into SDB primarily in the winter. Major ones include Sweetwater River (drainage area: 540 km^), Otay River (360 km^), Chollas Creek (70 km^), and Paleta Creek (10 km^) (Chadwick et al, 1999). However, according to data from the Sweetwater Authority (Garrod, personal correspondence), Sweetwater Dam traps over 75% o f the drainage area o f Sweetwater River, and the flow over the dam to SDB has been sporadic and essentially nonexistent in recent years. Similarly, there are two large reservoirs (upper and lower Otay River Reservoirs) on Otay River. As a result, the actual storm water draining into SDB is significantly reduced. Ground water is not a significant source o f fresh water to the bay (Chadwick et al, 1999). 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based on the above information, SDB can be classified as a typical low-inflow estuary. Even though it is semi-closed, the salinity difference between bay water and open ocean is usually small. Common salinity values are 33.3 % o at the mouth, increasing to about 35.5 %o in the south bay. The communication between bay water and open ocean is mostly provided by currents generated by semidiurnal tides (Largier, 1995). The tidal range from mean lower-low water to mean higher-high water is about 1.7 m with extreme tidal amplitudes of about 3 m. Bay water temperatures range from about 14 to 25°C throughout the year (Largier, 1995). SDB waters are vertically well-mixed throughout the year. Residence times o f water increase from about 5 to 20 days in mid-bay to over 40 days in south bay (Largier, 1995; Wang etal, 1998). The waters outside SDB belong to the Southern California Bight, defined as “an open embayment o f the Pacific Ocean” with a western boundary defined by the California Current, on the north by Point Conception, on the south to Cape Colnett, Baja California (Carlucci et al, 1986). The ridges and troughs within the Bight generally align northwest-southeast. Its irregular bottom topography, characterized by a steep slope and narrow (3-4km offshore) continental shelf, has a profound influence on its circulation. The oceanic currents within Southern Califomia Bight belong to the California Current system but bear very different patterns from the California Current itself (Jackson, 1986). The main currents within the Southem Califomia Bight are the 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. northward-flowing Southern Califomia Countercurrent on the surface and northward- flowing subsurface water, with the former showing high seasonal variabilities (Jackson 1986). The biological productivity o f San Diego Bay is at least a factor o f 10 higher in the bay relative to the surrounding ocean waters, presumably as a result o f nutrient loading from near-shore sources (Chadwiek et al, 1999). Marine habitats in San Diego Bay provide shelter and other life support for 66 kinds o f finfrsh, over 200 species o f invertebrates, 72 species o f birds (Needham, 1983). The shallow subtidal zone constitutes about 58 percent o f the bay water surface area while the deep subtidal zone makes up about 42 percent. The 4.8x10 m o f intertidal zone, about 10 percent o f the bay water surface area, is composed o f 54% mudflat, 29% tidal marsh and salt flats, and 17% sandy beach (Needham, 1983). Bottom types include mud, sand, rocky, and combinations o f these three. The bottom types o f north bay are predominantly coarse grain sands and rocky. Those o f south bay are generally mud bottom while those o f the central bay are mud and sandy. Eelgrass is the single most important habitat for San Diego Bay shellfish. While many finfish and shellfish hateh and develop in the shallow areas o f south bay and remain permanent there, others move out into the deeper waters o f the bay or beyond into the Pacific Ocean. In this way, the bay serves to replenish some offshore species by either direct migration or by larval disbursement (Needham, 1983). 27 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The sediments o f San Diego Bay eohsist primarily o f mud, silt, and sand. In general, the grain size distribution o f sediments in north bay is coarser, while those o f the south bay are finer. However, the distribution can vary from this general pattern with some areas almost exclusively sand such as that off North Island and areas that are predominantly silt such as those within Shelter Island. While distribution o f sedimentary materials in the bay is reasonably consistent with sources and depositional environment, two redistribution processes operate. The first process is dredging that artificially alters areas o f high deposition through the wholesale removal o f material. The second process is both a man-made and natural process in which sediments are resuspended into the water column and redistributed by currents. Natural resuspension occurs from tidal- or wind-generated currents moving over the bottom in shallow regions such as south bay. Man-made resuspension results from ship propeller wash that occurs primarily during ship movements in and out o f pier areas or even in the deeper mid-channel region during the transit o f large ships such as aircraft carriers (Hyman et ah, 1995). Sedimentary organic carbon concentrations in the bay have been measured in the range o f 1 to 5%. In general, the higher organic carbon values tend to be associated with finer grained material (Krom et al., 1989). 2.4 Historical Pollution o f SDB Past and present pollution sources o f San Diego Bay include sewage, industrial (commercial and military) wastes, ship (commercial, pleasure craft, and military) 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. discharges, urban runoff, and accidental spills. Hydrography (flow and behavior o f a water body) plays an important role in their transport and ultimate settlement. During the post-war period (1950s-1960s), more than 50 million gallons per day o f sewage and industrial wastes were disposed o f in the bay. By 1963, a massive collection, treatment, and ocean sewage disposal system was constructed, and sewage discharge to the bay ended. This change resulted in a return o f native fish and shellfish species, as well as the re-colonization by valuable plants needed by marine life. The removal o f the sewage from the San Diego Bay allowed the Bay to become a valuable recreational resource. The dredge and fill history o f the bay has reclaimed about 90% of available marshlands and 50 percent o f intertidal lands (Peeling, 1974). Particularly intensive bay dredging activity occurred during 1941 to 1945 to accommodate Navy operations. Much o f the bay was deepened during that period. Extensive dredging was also implemented in shoreline areas in 1942, 1951, 1952, 1956, 1964, and 1975. Smaller dredging projects were completed in 1955, the late 1960s, the early 1970s, 1985, and 1987. Current sources o f pollution to San Diego Bay include underground dewatering, industries, marinas, U.S. naval installations and urban runoff. In varying degrees these sources have led, and continue to lead, to contamination o f some San Diego Bay 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediments. Although sewage input to the bay has been stopped, both permitted and non-permitted sources (e.g., runoff, spills, etc.) continue to impact the waters and sediments o f the bay (Chadwick et al, 1999). Currently known chemicals in the bay include: arsenic, copper, chromium, lead, cadmium, selenium, mercury, tin, manganese, silver, zinc, tributyltin, polynuclear aromatic hydrocarbons (PAHs), petroleum hydrocarbons, polychlorinated biphenyls (PCB), chlordane, dieldrin, and the polychlorinated pesticide DDT (SDIWQP, 1990). There is a generally decreasing trend for many contaminants (Fairey et al, 1998). However, SDB remains labeled as one of the most contaminated embayments in the United States (Fairey et al, 1998), so the environmental quality o f the SDB area has been a constant issue for the local government, public agencies, and residents-. 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3. S am p lin g and A n a ly tica l M eth od s 3.1 Sample Collection Field sampling was conducted during June-July 1999, January-March 2000, June 2002 to July 2003. There were 26 sampling cruises or trips in total. Due to the large number o f sampling cruises/trips and large number and variety o f samples collected, they are described in chronological sequence for clarity. Table 3-1 lists sampling date, stations sampled and types o f samples collected for all the sampling cruises/trips during June 1999-July 2003. Table 3-1 Summary o f all sampling cruises/trips Date Station Sample type Note June 15, 1999 9,8,a7 water June 18, 1999 7,6,a6,a5 water June 22, 1999 1,2,3,4,5,6,8 Sediment core June 25, 1999 5,4,a4,2,a2,al,l water June 29,1999 l,3,a3 water February 17, 2000 3,4,5 water February 22, 2000 2, a l, 1 water February 22, 2000 4 Sediment trap February 25, 2000 6 water February 25, 2000 3 sediment trap February 29, 2000 8,9 water 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-1. Summary of all sampling cruises/trips (continued) Date Station Sample type Note February 29, 2000 2 Sediment trap June 2, 2002 u s e Atmospherie Fallout Gollected since May 31,2002 June 12, 2002 SM Bl water June 13, 2002 u s e Atmospheric Fallout (.91dpm) Golleeted since May 31, 2002 June 23, 2002 SMB2 water July 6, 2002 OBP, IP, SDR, OTR, SWR water July 11,2002 eW , SI, HIW, HIE water July 26-27, 2002 SI water Sampled 6 times July 27, 2002 OTR (low tide) water September 6, 2002 70,80,90,100,60 50,40,30,20,10 water September 6, 2002 70,80,90,100,60 50,40, 20 Surface sediment Collected by Van Veen grab September 6, 2002 90,30,10 Surface sediment Collected by gravity corer November 8, 2002 MM Rain water November 9,2002 SI, GP, SDR water December 26, 2002 TJR, SWR, e e , OTR, SI, HIW, HIE, GP, WSB, ESB water 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-1. Summary of all sampling cruises/trips (continued) Date Station Sample type Note December 26, 2002 TJRl Sediment core Sampling site is a Salt plain February 16-17, 2003 SI, GP, HIW, HIE Water SI and GP were sampled 10 times March 2,2003 CNM Atmospheric fallout Collected since July 6, 2002 February 17, 2003 TJR2 Sediment Core Sampling site is a bushy lowland July 28,2003 u s e Atmospheric fallout Collected since May 2002 On June 15,18, 25 and 29, 1999, water samples were taken from stations 1-9 and a l- a7 by 30-L Niskin bottles. The sampling stations are shown in figure 3-1, and the sampling collection log is shown in table 3-2. All the samples were collected during rising low high tides with ranges o f about 1 meter. 1-3 samples were taken from each station, depending on the depth o f the station. 5 gallon (18.9 L) plastic bottles were used to store water samples from Niskin bottle casts. Each plastic bottle was cleaned multiple times with alternate rinsing with hydrochloric acid and double deionized water (DDIW). Each sample, upon collection, ^^**Th, ^*^^Po, stable lead and FeCE were added to each sample. ^^°Th and ^'^^Po were spikes for subsequent measurement of 209 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 'a4 a5 # 6 • s Figure 3-1. Locations o f stations for sampling cruises during summer 1999 and winter 2000. Sediment cores were collected at stations 1-6 and station 8 in summer 1999 cruises. o " X A 1 n Th and Po, respectively; stable lead was used to calculate the recovery o f lead in the analytieal proeess in order to ealeulate ^'^Pb; FeCh was used as a eo-preeipitant. After spiking, each sample was vigorously shaken to mix well with spike solution, transported to USC Geochemistry Lab immediately after each cruise and processed promptly (sample processing procedures are described in the next section). In the ensuing sampling cruises/trips, the processing and analyses of water samples followed the same procedure as described above. 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-2 W ater sample collection log during summer 1999 Date time (hhmm) station number station depth(m) sampling depth (m) salinity (psu) water temp(*’C) June 15 0850 9 4.0 3.2 29.6 21.14 June 15 0920 8 7.2 7.0 37.4 21.41 June 15 0930 8 7.2 4.0 34.46 21.34 June 15 0945 a7 12.0 11.3 37.4 20.43 June 15 1000 a7 12.0 5.3 37.73 20.74 June 15 1015 a7 12.0 1.0 37.78 20.73 June 18 0915 9 4.0 NA 39.0 23.09 June 18 0920 9 2.7 NA 38.7 23.04 June 18 1000 7 10 8.7 36.5 20.76 June 18 1015 7 10 0.3 37.0 20.65 June 18 1030 6 9.3 8.3 36.5 20.31 June 18 1040 6 9.3 0.3 37.0 20.64 June 18 1053 a6 3.7 2.0 37.0 21.00 June 18 1110 a5 12.0 10.0 37.0 20.60 June 25 0723 5 10.0 8.0 34.0 20.0 June 18 1120 a5 12.0 0.3 37.0 20.74 June 25 0745 5 10.0 0.3 34.0 20.0 June 25 0800 4 11.7 0.3 NA NA June 25 0815 4 11.7 8.0 NA NA June 25 0830 a4 13.7 13.0 NA NA June 25 0835 a4 13.7 6.3 NA NA June 25 0835 a4 13.7 0.3 NA NA June 25 0850 2 8.3 0.3 NA NA 35 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-2 Water sample collection log, 1999 (continued) Date Time (hhmm) Station Numbe r Station depth(m) Sampling Depth (m) Salinity (psu) Water T em p(V ) June 25 0903 2 8.3 7.7 NA NA June 25 0903 a2 9.3 0.3 NA NA June 25 0913 a2 9.3 7.3 NA NA June 25 0930 al 13.3 0.3 NA NA June 25 0940 al 13.3 12.7 NA NA June 25 0943 1 11.7 0.3 NA NA June 25 0943 1 11.7 11.0 NA NA June 25 1020 2 11.0 11.0 NA NA June 25 1020 2 11.0 0.3 NA NA June 29 0830 1 12.0 0.3 NA NA June 29 0840 1 12.0 10.3 NA NA June 29 0905 3 12.0 0.3 NA NA June 29 0915 3 12.0 10.7 NA NA June 29 0925 a3 10.7 0.3 NA NA June 29 0936 a3 10.7 NA NA NA June 29 0947 a3 10.6 10.0 NA NA On June 22, 1999, sediment samples were collected from stations 1-6 and station 8 (refer to figure 3-1 for locations) aboard the RV Sea Watch using a box corer with a size of 50cm(L)x40cm(W)x50cm(H). Station 7 was abandoned for sediment sampling after 6-7 attempts due to garbage that covered the sampling site. Each box core, upon retrieval, was visually checked to ensure an undisturbed sediment-water interface. 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Overflowed sediment cores were discarded, so were those with insufficient length (<15cm). Two subcores were taken from each box core using 9.5cm i.d. acrylic plastic tubes. It was observed that the shortening o f the section due to subcoring was minimal (l-2cm for a 30cm core). The subcores were kept in iee boxes until being transported to the USC Geochemistry lab, where they were kept in refrigerators or in a cold room at 4 °C until analysis. The set o f subcores for organic analysis were transported to the chemistry lab o f the Southern California Coastal W ater Researeh Project (SCCWRP) and stored at -2 0 * ^ C until analysis. The sampling log including a description o f each sediment sample is shown in table 3-3 below. Table 3-3. June25, 1999 box core collection log and sample description Station Sampling Time Notes/Sediment Core Description 8 11:20am Top 0-15cm soft silty mud; 15-25cm: mixed with <10% black coarse sand; below 25cm: 100% silty clay. Little bioturbation in the entire eore. 7 10:45am (sampling failed for 6 times due to garbage spread over the whole area; gave up sampling at this site) 6 12:17pm Top 0-20cm: sandy with little bioturbation, homogenous. 0-6cm is light-colored; increasingly muddy downward; 21-25cm: bivalve debris layer; below 27cm: light-colored clay. 5 12:44pm surface covered by algae mats; 0-5cm: light-colored silt; bioturbated at surfaee; no bioturbation at depth. 37 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-3. June25, 1999 box core collection log and sample description (continued) Station Sampling Time Notes/Sediment Core Description 4 1:25pm Top 0-5cm: abundant shrimp/invertcbratcs; sandy and hard, light-colored, muddy below 5cm; a layer of hard chunky clay at 9cm; broken shell layers under 10cm; 2 live razor clams at 13 cm; another one at 18cm; 21-24cm: shell debris layer. 3 1:57pm Top 0-5cm: light-colored sandy clay; no sign o f bioturbation; 5-12cm: ycllowish/black alternate layers; little bioturbation at depth but sand-filled tunnels were found at 18cm; a shell debris layer at 16cm; 2 2:45pm Top 0-5cm: light-colored silty sand; turns blackish- brown downward; black muddy sand below 12cm. 1 3:15pm Top 0-5cm: hard, coarse and sandy, light colored; brownish black downward; abundant vertical holes by bivalves and nematodes. Thread-like worms and hard-bodied invertebrates; shell debris layer at 11cm. In sampling cruises during February 17 to February 29, 2000, only water samples were collected from stations 1-6, 8, 9, and al using the same scheme as described above. However, only one set o f duplicate samples was collected from each station because it was found from summer 1999 samples that there was little difference between water samples collected at different depths. One set o f the duplicate samples 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. were spiked in the same way as described above, the other set o f duplicates were kept intact until being transported to USC Geoebemistry Lab. They were vaeuum filtered to remove particulate matter (>0.47mm) within 24 hours, then the filtered samples were spiked in the same way as described before. There was a major storm during February 20-21 (2.1 inches), and a smaller one on February 23 (0.7 inches), and the water was turbid during February 22-25 but became rather clean on February 29. Three sediment traps were deployed and retrieved during February 17-22, February 22-25, and February 25-29, at stations 4, 3 and 2, respectively. The traps were made o f acrylic plastic tubes o f 9.5cm i.d. with a length o f 40cm. They were sealed at one end and covered with 0.5x0.5cm metal mesh at the other end to prevent large “swimmers” from entering the trap. About 50g potassium chloride was added in the trap before deployment to prevent scavengers from consuming the trapped materials. At the same time, the added salt could keep a higher density at the bottom o f the trap so that the turbulence was minimum. The traps were deployed vertically, 1.5m from the sea floor. During both June 1999 and February 2000 sampling seasons, Infiltrex 100 water pumping systems (Axys Environmental Systems Ltd., Sidney, BC, Canada) were deployed at depths o f 1.5m and 5m above the sea floor from stations 1-9. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. During June to December 2002, a number o f sampling trips were made to stations around San Diego Bay. Locations o f these stations are shown in figure 3-2. Not shown are the atmospheric fallout collection site at Los Angeles, two water sample collection sites at Santa Monica Bay, and one site at M ira Mesa (a city about 30 miles north of San Diego), where rain water sample was collected. HIW •H IE LC W CNM \ H \ S WSB ^ »<ES8 / V Pond ft OBP Figure 3-2. Locations o f stations around SDB sampled during June-December 2002 except for the September 6, 2002 cruise (see figure 3-3). 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • ')\0 Atmospheric fallout o f Pb was collected at both Los Angeles and San Diego. The collectors were made out o f 5-gallon (18.9L) Sparklett® water jars with the bottom removed and mouth sealed. The shape o f the jar (40cm deep with 25cm i.d.) minimizes the turbulent disturbance, and the funnel-shaped bottom (i.e. the mouth and the neck o f the water jar) was able to trap the fallen particles, preventing loss of collected fallout. Metal mesh with pointed edge was bound outside the collector to discourage birds from resting on the collector. The jars were put inside a wooden frame for stability. Los Angeles site (labeled ‘U SC’) was on the roof o f the Electron Microscope Lab. The collection time periods at this site were from May 31 to June 2, 2002; May 31 to June 13, 2002, and June 13, 2003 to July 28, 2003, respectively. The San Diego site was collected in station CNM located at Cabrillo National M useum at Point Loma (Figure 3-2) from July 6, 2002 to March 2, 2003. On June 12 and 21, 2002, two water samples were collected at SM Bl (location: 34°01.284’N, 118°31.473’W) and SMB2 (33°39.730’N, 118°01.479’W), with the help o f Steve Colbert during two separate sampling cruises to Santa M onica Bay. The samples were collected for comparison with the coastal sea offshore San Diego. They were not spiked on site before being transported to USC Geochemistry Lab. On July 6, 2002, water samples were collected at stations OBP (Ocean Beach Pier), IP (Imperial Beach Pier), SDR (San Diego River Channel, OTR (Otay River), and 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SWR (Sweetwater River). OBP and IP were fishing piers that extend into the Paeific Ocean for about 550m and 480m, respectively, and the water depth were about 40 feet at both locations. At all stations, water samples were collected by an acid-precleaned plastic bucket, and stored in 5-gallon plastic jars. A similar sampling technique was used on July 11 sample collections, which covered stations CW (Convention Way Fishing Pier), SI (Shelter Island Fishing Pier), HIW (Harbor Island west end), HIE (Harbor Island east end). To determine the effect o f tidal pattern on the radioisotope activities in the water column, multiple samples were taken at station SI from one station during an entire tidal cycle on July 26-27. Three o f the six samples were spiked in order to measure radionuclide activities in dissolved and particulate phases. On September 6, 2002, a major sampling cruise was carried out. Stations sampled during this cruises are stations 10-100 for both water samples and surface sediments. Water samples were collected with 30-L Niskin bottles. For deep stations (i.e. stations 50-100 with depths >40 ft), 2 samples were taken at about 0.6m from surface and about 2m from the bottom sediment. Only one sample was taken at approximately 3m below the surface at stations 10-40. Surfaee sediment samples were eolleeted by van Veen Grabs, which typically collected 5-10cm surface sediments with some disturbance on the sediment surfaee. A gravity corer was deployed at all stations, but 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediment cores were eolleeted successfully only at stations 90, 30, and 10. W ater samples and sediment cores were processed as described before, and the properties of the suspended particles in these water samples were described in table 3-4. Sediment samples collected by van Veen Grab were analyzed as single sediment samples. It should be noted that the van Veen grab collects only the top 5-10 cm o f sediments. 4 0 • 6 0 100 Figure 3-3. Station locations for the September 6, 2002 sampling cruise. Sediment cores were taken from stations 10,30 and 90. Van Veen grabs were deployed at other stations. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-4 Description of Filter-Retained Particles for September 6, 2002 Samples Station SPM Description 10 Sandy particles observable at the bottom o f the bottle. Large amount of zooplankton swimming. Fair amount o f phytoplankton. 20 A few live copepods, many live smaller zooplankton. Large amount o f microalgae. 1 mm long fish larvae. 30 No record 40 Large amount o f copepods and algae, not much sandy particles. 50-2m Moderately large amount o f plankton 50-16m No Record 60-2m sandy particles abundant. Large amount o f zooplankton. 60-14m Fair amount o f phyto-zooplankton, abundant sandy particles 70-2m few green-yellowish colored phytoplankton, a few small zooplankton 70-16m very sandy, high amount o f particles, mostly fine sand, probably Niskin bottle hit bottom 80-2m fairly abundant Green-brownish colored phytoplankton, a few small zooplankton, few sandy particles 80-32m some green-brownish phytoplankton, a few large zooplankton, few sandy particles 90-2 very clean. Few phyto-zooplankton, very light greenish-yellow color 90-53m Very clean but a little more TSS than corresponding surface sample consisting fine clay tinted by light green color. 100-2m few green-yellowish phytoplankton, a few small zooplankton, no sandy particle 100-16m fairly abundant green-brownish phytoplankton, a few small zooplankton, few sandy particles 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. With the realization that tidal pattern and precipitation may significantly affect the radioisotopic activities in water columns, the samplings after the September 6, 2002 cruise were focused on tidal and precipitation patterns. More samples were also taken from around SDB in order to get a more complete pattern o f the entire area. On November 8, 2002, rainwater was collected during precipitation event (0.33 cm) with a clean bucket at M ira Mesa, about 15 miles north o f the City o f San Diego. Then water samples were collected from stations SI, GP and SDR during the storm. During a sampling trip o f the entire San Diego Bay area on December 26, 2002, water samples were collected from stations TJR, SWR, CC, OTR, SI, HIW, HIE, GP, WSB, ESB. Refer to figure 3-2 for sampling locations. W ater samples were collected by preeleaned plastic bucket and stored with 2.5 gallon plastic bottles. All samples were not spiked on site in order to measure radionuclide activities in particulate and dissolved phases. During the sampling trip, a sediment core was collected at station TJR I located in a dry fluvial plain on the west side o f the Monument Road inside the Border Field State Park. The core was taken by hammering a 9.5 cm i.d. acrylic plastic tube into the sediment. About 25 cm o f sediment core was taken. There was a minor disturbance o f the top sediment during sampling. During February 16-18, 2003, a time series sampling was carried out. During the sampling session, 10 samples were collected from stations SI and GP during a 25- 45 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hour continuous sampling. In the mean time, 4 samples (one duplicate) were taken at stations HIE and HIW. All samples were not spiked on site so that both phases (particulate and dissolved) can be measured for radionuclide activities. A sediment core was taken at station TJR2, which located in a bushy lowland inside the Border Field State Park. Station TJR2 was less than one mile to the north o f station T JR l. 3.2 Sample Processing and Analyses Sediment cores were sliced into 0.5-cm or 1-cm intervals after being extruded by a extruder designed and manufactured by Dr. Will Berelson and Gerry Smith. Each sample was weighed, dried at 60°C, and weighed again to calculate water content and porosity. Dry sediments were then pulverized using a jade mortar, and y-counted using a high-resolution, well-type intrinsic germanium detectors for determination of ^’°Pb, ^^^Ra, ^'"^Pb, and '^^Cs aetivities. ^^^Ra were used to ealculate excess oin o o A oin Pb activities beeause Ra represents the source o f Pb other than atmospheric 9in • • 91 0 «« input (sediment Pb activity is meant to indicate excess Pb activity for simplicity 9 3 S • 9 9 ^ > • 91 4 thereafter). Due to the interferenee o f U with Ra in the y-spectrogram, Pb peaks were often used to calculate supported activities o f ^'°Pb. Atmospheric fallout o f ^’°Pb was also measured using y-counting. The method of sample preparation was largely adopted from that used by Huh et al (1999). Briefly, fallout collectors were washed alternately with IN HNO3 and DDIW, at least 5 times. 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. into a 1-L glass beaker. The washout in the glass beaker was then condensed by evaporation to under 50ml, the remainder was transferred to a 150ml Teflon beaker with 5N HNO3. Then the eontent in Teflon beaker was digested overnight with eoncentrated HNO3, HCIO4, and H2O2, condensed further to near dryness, then transferred to 5-ml plastic tube for gamma counting by washing with 5N HNO3. W ater samples in plastic jars were weighed (to a precision of 5g) before processing. If only total (particle + dissolved phases) activities o f radionuclides were to be analyzed, samples were acidified and spiked in the field, as pre-filtration was not needed. For those samples that radionuclides in both phases were to be analyzed, they were pre- filtered via vacuum-filtration with 0.47-pm Millipore® membrane filters in a system powered by a Welch 1908 vacuum pump. Pre-filtration was carried out immediately upon sample reception (<24 hours) at the laboratory. Filter-retained particles (together with filter membrane) were completely dissolved by eoncentrated FINO3, HCIO4 and HF in Teflon beakers. Spikes o f ^^°Th, ^®^Po, common lead (^**^Pb) were added to the digested solutions and dried for subsequent analysis. The water samples (both unfiltered samples for total radionuclide analysis and the filtered samples for analysis of dissolved radionuclides) were acidified, yield tracers o r iQ - tq ^ __ Po and Th as well as co-precipitant F eC f were added, and samples were mixed by vigorous shaking and allowed to equilibrate for 12-24 h. Excess NH4OH was 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. added to the samples to adjust pH to -8-10, and the samples were mixed well and then kept still for at least 24 h, allowing radionuclides to co-precipitate with Fe(0 H)3. Water samples were checked closely to make sure that the precipitate was well formed and completely settled before vacuum-filtration was carried out. When the filtration was finished, the precipitate was washed by DDIW to remove salt, and then the filter membranes were washed with 4N HCl to collect precipitates. For unfiltered water samples, there would be a small amount o f insoluble materials. In this case, the sample was dried and concentrated HNO3, HCIO4 and HF were added to digest the insoluble materials. After digestion, the sample solution was dried completely. Concentrated HCl was usually added to fume off remaining HNO3 or HF that could jeopardize subsequent analyses. ^'°Po was plated onto a silver planchet following a similar procedure similar to that described by Flynn (1969). Dried sample was dissolved on hotplate by IN HCl, ascorbic acid was added until the yellowish color o f the solution disappeared to prevent Fe^^ interference. The silver planchets used in this study were purchased from Southwest Jewelry with 0.9cm diameter and 0.2mm thickness. Each planchet was polished on one side with extra-fine sandpaper to remove the oxidized surface, marked at the other side, and placed in the sample breaker with the polished side up. A cover glass was placed on top o f each beaker to prevent vapor loss. 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. * * 2 1 0 * • An improved procedure over the conventional Po-plating method (Santschi et al, 1979, Carpenter et al, 1985) was developed in this study after a number o f tests. It was found that conventional method using hot plate tends to fail due to small temperature fluctuations because ascorbic acid was found to be easily degraded at higher temperature. In this study, a water bath with a temperature precision within ± 1.0 was used to set the plating temperature at 60 ® C , and plating time was controlled by a self-timer, usually set to be 6 hours. It was found that lower temperature and longer plating time in the new method produced an optimal result, which was characterized by sharp, well separated ^*^^Po and ^^®Po peaks with very low background. It should be noted that the development o f this new method benefited from the method development by Benoit (1988), whose method involved room- temperature plating for 7-10 days. However, his method has an important problem, i.e. it neglected the short half-life o f ^^'^Po (138 days). During the period o f 7-10 days, an additional 3.5%-5.0% decay o f ^***P o will take place. The result will be further 210* 21 0 complicated depending on whether Pb is separated from Po or not. So its 210 210* * application is limited to the measurement o f Pb after Po is removed by excessive silver plate/scraps. Comparatively, the new method in the present study is superior in plating time (due to higher plating temperature) and simplicity. All the Po-plating was carried out in a fume hood, which provided vibration so that the plating is no longer controlled by diffusion. The silver planchet was then counted 49 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 0 9 1 0 9 1 0 by a-spectrometry to obtain Po activity. Due to low activities o f Po and Pb in water samples, conventional procedure dictated that large quantities o f sample be collected (19 liters). By improving the sample processing method as described above, sample size can be reduced to less than 50% o f original (l-2.5gallon) with comparable or better precision. At times, the a-spectrum o f some samples would give broad peaks for both ^’^Po and ^®^Po, presumably due to the interference by other metal ions in the solution. The resulting silver planchet would appear yellowish or brownish, and the peaks in the a-speetrom eter would have tails toward lower energy. It was found that this problem can be solved satisfactorily by boiling the planchet for about 30 minutes in concentrated hydrochloric acid until the color o f the silver planchet faded. ^^^Pb was measured over a year after the measurement o f ^**^Po, when the original ^'®Pb had produced sufficient ^'**Po, which was measured by the same procedure as described above. The remaining solution after Po-plating was treated with HNO3 to destroy ascorbic acid, dried and re-dissolved with 1% ITNO3 to a volume o f 500ml. The resulting solution was then measured in Jobin Yvon/Horiba inductively coupled plasma-atomic emission spectrometry (ICP-AES) system at wavelength of 283.306nm, which was shown to give the best sensitivity. Three duplicates were measured for each sample with standard deviation o f usually less than 5%, and 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • 2 ~ ^ • * * standard solution o f Pb gave a correlation coefficient o f 1.00, suggesting a satisfactory precision. Water samples collected on December 25, 2002 and during February 16-18, 2003 were also measured for trace metals Pb, Cr, Mn, Zn and Cu using Jobin Yvon/Horiba ICP-AES system using similar methods. The eoncentrations o f Cu and Pb were mostly below the detection limit o f the instrument (15.0 ppb) and were not reported. Thorium in the solution after polonium plating was purified by anion exchange, extracted by 0.4N thenoyltrifluoroacetone-benzene solution, and deposited onto stainless steel plates for simultaneous a~P counting. Concentrations o f TSS were measured by weighing filter-retained particles before and after filtration. Filter membranes were rinsed with DDIW multiple times after filtration to remove salt. • 71 0 7 T 4 • • Particulate phase Pb and Th were administered by total dissolution o f particles together with the membrane filters with HNO3, HCIO4, and HF. A blank sample was 7 1 0 7T4. processed following the same procedure, and no Pb and Th were detected. Samples collected by in-situ pumps during the June 1999 and February 2000 sampling seasons were processed and analyzed for TOC/TON and PCBs in the chemistry lab o f SCCWRP. The procedures are described in Zeng et al (2002). 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 4 Results 4.1 R e s u lts o f J u n e -J u ly 1 9 9 9 s a m p lin g c ru is e s Due to the large number o f sampling cruises/trips, and each cruise/trip produced a different set o f data, the data will be summarized in a chronological sequence except for tidal phase samplings and atmospheric fallout samplings. A complete list of sampling trips and samples collected is given in table 4-1. 4.1.1 C TD Profiles Two CTD cast results are shown in figure 4-1, measured on July 22 aboard the RV Seawatch. Both stations had salinity o f 33.5%o, pH o f 8.0, and oxygen level o f 6-9 mg/L, similar to Chadwick et al (1999)’s reported data. It was evident that the water columns in these stations were very well mixed. a 20 b 20 : — Temp 1 5 ..................... 15 i ♦ ♦ * ♦ ' * ♦ .......................................... — *— Salinity/10 in 1 0 i < * t • 5 i - ^ 0 2 % 5 ■ ■■ ■ ■ ■ 0 , ----------- : 0 i -------- ----- , ------ ■ . 0 2 4 6 8 10 * 0 2 4 6 8 10 D e p t h ( m ) D e p t h Figure 4-1. CTD data for station 5 (a) and station 3 (b) 52 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.1.2 A ctivities o f T otal ^'°Po and 1 A 9 1 0 The results o f total (dissolved + particulate) Po and Pb activities in water columns in the summer 1999 sampling cruises were shown in Figure 4-2. Usually the relative errors are ahout 5%, but there are some samples having estimated errors of 30-40% (as marked by * ), primarily due to poor recovery o f Po during plating, which was shown in the a spectrum as broad peaks. 0.300 0.250 0.200 E 0.150 0.100 0.050 0.000 ■ Total Po-210 □ Total Pb-210 al 2 a2 3 a3 4 a4 5 a5 6 a6 7 a? 8 9 Station (Bay Mouth—> South Bay) Figure 4-2. Activities (dpm/L) for total ^ 'V b and ^’°Po for stations across SDB. Uncertainties were about 5% but could reach 30% for some samples. Refer to figure 3-1 for station locations. 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It should be noted that this batch o f samples were analyzed using the old method of Po-plating (refer to chapter 3), which has higher uncertainties due to incomplete 1 n 0 1 n plating, interferences from organics and other metals. When Po growing from Pb was measured after one year o f storage, less than optimal amount o f ^°^Po spike was added. These factors result in larger uncertainties and could slightly overestimate 0 1 n 1 n Pb activities relative to Po. As a result, the data should be interpreted with caution. Despite the relatively large uncertainties with the sample analysis, the data in general show some clear patterns that cannot be attributed to analytical errors. One o f the 910 9 1 0 striking features is that Po activities exceed those o f Pb in all samples with an 910 91 0 • average Po/ Pb ratio o f 2.56. At the same time, the geographic differences between ^'®Pb and ^'°Po activities were obvious. Central Bay from station scS to station scl (also including stations 4,5,6) had relatively low ^'**Pb aetivities with an average o f 0.0123 dpm/L, while the stations in the North Bay (stations l,sc7, 2,se6 ’ 2 1 0 1 O and station 3) and South Bay (stations 8 and 9) had much higher Pb and Po activities than the central SDB. A detailed discussion about this unexpected pattern will be given in chapter 6. Briefly, large excess ^'**Po over its grandparent ^'®Pb can • • • • 2 1 0 * * * • only be explained by intensive recycling o f Po-rich biogenie particles in the shallow 9 1 n water. In comparison, Pb seemed to preferentially attach to lithogenic, heavier 54 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. particles and is relatively more difficult to resuspend or recycle in the water column. 210 2 10 Unfortunately, the activities o f Po and Pb in particulate and dissolved phases were not measured separately and it is impossible to attribute the excess total ^'*^Po over ^'^Pb to either phase. But based on the data accumulated to date, it is inferred that the particulate phase contributed most o f the excess total ^'°Po in these samples. It should also be noted that these samples were collected at different times with similar but not identical tidal patterns. All the samples were collected from 6am to about 10am (sometimes as late as 12pm), during which there were rising lower high tides for the study area. Since the sampling lasted for over three weeks, the tidal pattern changed to some extent, albeit not much. As suggested by more recent data (shown below), tidal pattern at the sampling time has a profound effect on the radioisotopic pattern. Thus the data should be interpreted with caution. 4.1.3 Water Contents of Sediment Cores Seven sediment cores from stations 1-6 and station 8 were also collected on June 22, 1999 during the summer 1999 sampling cruises. The water content in sediment columns is shown in figure 4-3. The irregularities in the water content in many cores suggested that the sediment columns were vertically heterogeneous, perhaps indicating a rather dynamic sedimentation environment, especially for station 8. At other stations there were layers with water content higher or lower than adjacent 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. layers, indicating changes in sediment composition and/or events that disturbed the normal sedimentation process such as dredging. Water Content {%) 20 40 60 20 40 20 40 60 20 40 60 E n 4-4 Q . < D a 4-4 c 0 ) E 0 ) C O 1 — ' ^ — 0 - ----- 1 -------------1 ^ ■ ------ 0 ■------^ -------------1 - ^ -------0 - 5 - 1 10 - / / 1 I 1 0 - 10 - • / 10 - A 20 - } • • f • • # i 2 0 - 15 - • 15 - • 30 - 1 • I • 1 20 - 20 - 4 0 ^ • ' 30 -J (a) Station 1 20 40 60 (b) Station 2 20 40 60 (c) Station 3 20 4 0 60 (d) station 4 0 - ^ — ^ ‘ '— 0 - 5 - \ 10 - T 10 - 10 - f 20 - • i • # t 20 - 15 - • 30 - 1 20 - 4 0 - 30 - j (e) Station 5 (f) Station 6 (g) Station 8 Figure 4-3. Water content (%) o f sediment cores collected on June 22, 1999 4.1.4 TOC-TON Profiles of Sediment Cores The depth profiles o f total organic carbon (TOC), total organic nitrogen (TON), and the ratios o f TOC/TON in sediment columns are shown in figures 4-4, 4-5 and 4-6, respectively. 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TOC for sediment samples averaged about 0.8% for all stations (figure 4-5H) except for station 1, which has a much lower TOC o f about 0.3 on average, presumably due to strong tidal currents that prevented deposition o f organie-rich fine partieles. For each station, there is a general trend o f deereasing TOC toward deeper samples, indicating the effect o f post-depositional degradation o f organie matters. Again, the irregularities in the profiles o f station 4, 6 and station 8 indicate sediment heterogeneity in these locations. As for TON, the shape of the depth profiles were similar to those o f TOC with TON range o f 0.05-0.15%, and there are large differences among different stations in TON. TOC/TON ratios have a range o f 6-15, with a generally increasing trend toward deeper samples. The irregularities in station 4 and station 8 also indicate heterogeneity. Compared to previous studies for TOC-TON and their ratios for surface sediments in an area close to the Point Loma Ocean Outfall (Zeng and Khan, 1994), which yielded values o f 0.53%, 0.038% and 13.95, respectively, SDB has higher TOC and TON but lower TOC/TON ratios. Suspended partieles in Southern California Bight in general also showed similar patterns (Williams, 1986), where TOC/TON ratio for surface water was close to 6.5, but increase significantly to 14.6 in deep waters. This suggests that 1) SDB water has higher productivity so that TOC and TON is relatively higher than sediment from outside; 2). Organic matter in surface sediments within SDB undergoes less degradation compared to sediments in the outer sea either because o f higher productivity or shallower water depth, or both. 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1.4 1.2 1 0.8 ^ 0.6 i 0.4 I 0.2 I 0 ■ ♦ ♦ ♦ ♦ 0 10 15 20 1.4 1.2 1 0.8 0.6 0.4 0.2 0 10 15 20 1.4 1.2 1 0.8 0.6 0.4 0.2 0 c 10 20 30 40 1.4 1.2 1 0.8 0.6 0.4 0.2 0 10 20 30 0.8 0.6 0.4 0.2 15 20 5 10 0 25 1.4 1.2 1 i 0.8 I 0.6 ^ 0.4 0.2 0 0 10 20 30 1.4 1.2 1 0.8 0.6 0.4 0.2 10 20 30 Figure 4-4. TOC% (w/w) profiles o f sediment cores, a to e: profiles for stations 1-6 and 8, respectively. 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.2 , 0.15 0.1 I 0.05 ! 0 0 10 15 20 0.2 0.2 0.15 0.1 0.05 0 0.2 0.15 5 10 15 20 d 0.15 0.05 0.2 0.15 0.1 0.05 0 10 20 30 40 0 0.2 0.15 0.1 0.05 10 20 30 0 5 10 15 20 25 0 [ 0 10 20 30 0.2 0.15 0.1 0.05 0 20 30 0 10 Figure 4-5. TON% (w/w) profiles of sediment cores in sediment cores collected on June 22, 1999, from stations 1-6 and station 8, respectively. 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 18 16 14 12 10 8 6 0 y = 0.0476X + 6.8963 R^ = 0.4157 5 10 15 20 18 16 14 12 10 8 6 y = 0.056x + 8.716 = 0.6559 20 10 15 20 y = 0.2149x + 8.244 ♦ R'' = 0.6645 ♦ ♦ 40 20 1 18 y = 0.1355x + 9.85 R^ = 0.0433 16 14 12 10 8 0 10 20 30 20 18 16 14 12 10 8 y = -0.0356x + 9.8115 R^ = 0.5282 10 20 30 20 18 16 14 12 10 f y = 0.0288x+ 10.317 3 2 . R^ = 0.7246 0 10 20 30 20 18 16 14 12 10 y = 0.1078x + 9.1503 R^ = 0.3523 10 20 30 Figure 4-6, TOC/TON ratios (w/w) o f sediment cores collected on June 22, 1999, from stations 1-6 and station 8, respectively. 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.1.5 Profiles o f A ctivities o f E xcess and '^^Cs in Sedim ent C ores Excess Pb, Th and Cs for stations 1-6 and station 8 were shown in figure 4-7. O f them, excess ^'‘ ’ Pb activities were calculated based on ^"^Pb activities measured at 295.2 and 242.0 kev, respectively, instead o f activities o f ^^^Ra (Ku and Luo, personal communications). The reason was that we have found that the peak o f ^^^Ra at 186.1 kev was often indistinguishable from the U peak at 185.7 kev. At the same time, it 222 226 is assumed here that the loss o f Rn (produced from Ra in the sediment) is • • 137 • • negligible. It was surprising that Cs peaks, which represent the time horizon o f year 1963 A.D., were absent in all sediment cores. There are two possible reasons for their absence. One is that the sediment horizons representing the year 1963 were dredged away during one o f the numerous dredging events (refer to chapter 2); another possibility is that the sedimentation rate is so high that all the peaks were buried beyond the sampled depth; or, 137Cs is rather mobile in SDB sediment due to high salinity and/or bioturbation. 21 A Based on the profile of excess Pb, the sedimentation rate o f each station was calculated and listed on table 4-1, together with the calculated enrichment factor based on the atmospheric fallout measurement and soil ^^*^Pb inventory calculation as described later in this chapter. The sedimentation rate (s; in cm/yr) was derived from 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Activity (dpm/g dry wt) S. ( U Q 0 2 4 6 0 5 20 - 10 - 10 30 - 40 (a) Station 1 (b) Station 2 (c) Station 3 30 -I 210| 234. 226| 137, Pb Th Ra Cs (d) Station 4 0 2 4 6 0 10 - 5 10 - 10 20 - 20 - 30 4 15 (f) Station 6 (g) Station 8 (e) Station 5 Figure 4-7. Profiles o f ^"^Pb, and in sediment cores collected on June 22, 1999, from stations 1-6 and station 8, respeetively. Uneertainties o f all data points are 10%-15% except for '^^Cs, whieh often has uneertainties of up to 50% due to low aetivities. Refer to figure 3-1 for station loeations. the profiles o f ^'®Pb in the sediment column if a constant input o f ^^®Pb and a constant sedimentation rate were assumed, and eompaetion neglected, using the following equation (Robbins, 1978): e = = Co*e (4-1) 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where Till /in c is the Pb activity (dpm/g) at depth z; co is the Pb activity at z = 0 (i.e., at the water-sediment interface); and X is the decay constant o f Pb with a value o f 0.031 yr-\ 91 0 The accuracy o f this method depends on the constancy o f the Pb input and sedimentation rate, as well as the extent o f mixing and other post-depositional processes. From table 4-1 it is seen that there are large uncertainties in the sedimentation rate calculation, and sedimentation rates at some stations (stations 2,4 and 8) could not be calculated based on equation (4-1) because o f apparently decreasing ^'^Pb aetivities with depth. As shown in figure 4-3 to 4-6, there are strong vertical heterogeneities in many sediment cores, especially cores at station 4 and 8, suggesting that there might be connections between sediment heterogeneity and irregular sedimentation rate or strong post-depositional processes that prevent calculation of sedimentation rate based on equation 4-1. The enrichment factors (EF) shown in table 4-1 were estimated from the sediment • 9 1 0 column inventory and atmospheric fallout rate o f Pb: E F = — \ A * d z (4-2) d D '''■atm 0 63 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Where X is the decay constant o f with a value o f 0.031 y f 910 9 7 ?atm is the atmospheric Pb fallout rate (dpm/cm /yr); ^ A , * d z is the sediment column ^^°Pb inventory integrated down to depth z 0 (dpm/cm^); A : is the ^'^Pb activity (dpm/cm^) at depth z. d was chosen for each stations such that 25 years o f sedimentation is accounted for in order to make different stations comparable. The result is listed in the table 4-1 below. Table 4-1 Sedimentation Rates and Enrichment Factors for ^'*^Pb for Summer 1999 Sediment Cores Station Spb210 Scsl37 Def EF 1 0.39±0.12 0.49 12.3 3.21 2 NC 0.40 >10.0 >6.5 3 1.54±0.29 1.00 38.5 12.6 4 NC 0.60 >15.0 >2.70 5 0.34±0.11 0.35 8.8 1.02 6 0.92±0.38 0.81 23.0 2.91 8 NC 0.70 >17.5 >1.50 64 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 n Note that in the above table, Spbiio is estimated from Pb profile according to equation (4-1). The values might be overestimates due to sediment mixing. Scso? is the lower limit derived from absence o f '^’Cs peak marking the 1963 A.D. horizon. Def is the depth (cm) representing recent 25 years that is considered for calculation o f enrichment factors(EF). A striking feature o f the enrichment factors is the large enrichment factor for stations 1- 3, especially for stations 2 and 3 with enrichment factors o f 6.5 and 12.6, respectively. It should be noted that these values are much higher than previously reported (up to 3, Appleby and Oldfield, 1992). This indicates that factors other than atmospheric input and river input have contributed to the overall inventory o f in the sediment columns in the north SDB. This important feature will be discussed in detail in chapter 6. 4.1.6 In-Situ Pum p Sam pling o f Suspended Partieles and PC B s During summer 1999 sampling cruises, in-situ pumps were deployed at stations 1-9, which were mainly aimed for analyses o f polychlorinated biphenyls (PCBs) in the water columns. PCB data were reported elsewhere (Zeng et al, 2002) and will not be described here. Figure 4-8 a and b show TSS levels at stations during summer 1999 as 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. well as winter 2000 cruises. As described in chapter 3, TSS was calculated by adding up the dry weight o f filter-retained particles (>0.8pm) in all 16 filter cartridges divided by the volume o f wafer being filtered, which was calculated from the recordings made by a microchip on the pump. TSS levels in two seasons were comparable, and showed similar pattern. There were higher TSS levels at both bay mouth and south bay, and lower levels in central bay. Higher TSS levels at both ends o f SDB were due to different mechanisms. In the bay mouth, it is related to heavy shipping traffic and stronger tidal current, both o f which cause sediment resuspension that caused elevated TSS level. However, in shallow south bay, where the average depth is about Im, wave action together with the wetting-drying o f bottom sediment maintain high TSS here. This pattern provided important insights into the particle dynamics o f the entire SDB and will be discussed in more detail in chapter 5 and 6. 15.0 10.0 5.0 0.0 ■ TSS-1.5m □ T S S -5m 1 1 1 1 J ] i fi l l I 1 2 3 4 5 6 7 8 9 15.0 10.0 5.0 0.0 ■ TSS-1.5m □ TSS-5m [l 1 2 3 4 5 6 7 8 9 Figure 4-8. TSS in mg/L measured by in-situ pumps during summer 1999 (a) and winter 2000 (b) cruises. Refer to figure 3-1 for station locations. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Compared to TSS, total organic carbon (TOC) in water columns across the bay did not have apparent geographic differentiation. M ost o f the TOC levels clustered around 3%, with winter 2000 data showing a larger variation. The exception at station #1 at 5m from the bottom was probably due to the disturbance o f the mooring after the deployment, as evidenced by the movement o f the mooring by 50m from its original place, and an abnormally high TSS value (see figure 4-8b). A comparison between figure 4-9a and figure 4-4 indicates that particles in water column have much higher TOC than the underlying sediment(>3% vs. <1.5%, for suspended particles and sediment, respectively), presumably due to the fact that suspended particles contains more fresh organic matter that undergoes little or no oxidation compared sediment column. 12 10 ■ TOC-1.5m □ T 0C -5m t l H I l D .biI 15 10 5 0 ■ TOC-1.5m □ TOC-5m 2 3 4 5 2 3 4 5 6 7 I Figure 4-9. TOC (%, w/w) o f in-situ pump samples collected in June-July 1999(a) and January-March 2000(b) cruises. Refer to figure 3-1 for station locations. Total organic nitrogen was also measured for the same batch o f filter-retained particle samples collected in summer 1999 season and the result is shown in figure 4-10. As easily seen, TON% o f these samples have very similar distributions as TOC%, and 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. TOC-TON relationship is shown in figure 4 - lla and b (for summer 1999 cruises only). The trend lines in figure 4-11 were forced through (0,0) for simplicity. If they were not forced, both trend lines would have a small positive intercept, suggesting contributions from “detrital or resuspended sediment” (Hammond, personal communications). TON-1.5m DTG N-Sm 2 3 4 5 6 7 Figure 4-10. Total Organic Nitrogen (TON%, w/w) for SPM collected by in-situ pump during summer 1999 cruises. Refer to figure 3-1 for station locations. y = 5.7946X R ^ = 0.8016 0.0 0.3 0.6 0.9 y = 5.9245x = 0.9202 6 4 2 0 0.3 0.6 0.9 1.2 Figure 4 - lla and b. TOC-TON relationship for SPM collected by in-situ pump during summer 1999 sampling cruises. A: each data point represents an average value for 16 filters in each filter cartridge; b; each data point represents a single filter. 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The average TOC/TON ratio o f 5.79 (w/w) in figure 4 - lla eorresponds to a molar ratio of 6.75 (if not specified, C/N ratio represent molar ratio hereafter), which is consistent with the ‘Redfield ratio’ o f 6.63 for fresh phytoplankton. Our in-situ pump has a filter cartridge consisting o f 16 filter paper that is analyzed separately. For single-filter samples, the data points have wider range than the combined samples (0.1-1.2% for individual samples vs. 0.3-0.8% for whole samples). It is interesting to note that samples with smaller TOC and TON (at TON~0.4%) have higher TOC/TON ratio (7.33, not shown, vs. 6.79 for other samples) if plotted separately, suggestive o f a possible source o f older sediments that have higher TOC/TON ratio. For example, the average TOC/TON ratio for surfaee sediment o f SDB is 9.42 (refer to figure 4-6). Comparatively, samples o f higher TON, higher TOC and lower TOC/TON ratio (5.82) might have more eontribution from fresh plankton. Compared to the data for suspended particles in Southern California Bight (Williams, 1986), in whieh the TOC/TON ratio for surfaee water was close to 6.5, but increase to 14.6 in deep waters. However, the sediment trap data showed a relatively consistent value o f 8.9. 4.2 Results o f winter 2000 sampling cruises 4.2.1 Aetivities o f Dissolved and Particulate ^'V o and ^'^Pb 210 210 • » » Figure 4-12 shows Po and Pb activities for both unfiltered and filtered water samples collected in February 2000. There were drastic differences between stations 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1,2 and sc7 and other stations, with activities o f both filtered and unfiltered ^'**Po and ^’°Pb 2-5 times higher than those o f other stations. The cause o f the large difference was probably due to the storms with 52mm total precipitation prior to sampling o f stations 1, 2 and sc7 (Michelle Chambers, Western Regional Climate Center o f the Desert Research Institute, personal communications). Samples from stations 3,4,5 were collected on February 17 before the storm, and stations 6,8,9 were sampled a few days after the storm, when tidal exchange should have diluted the effect o f the storm runoff. 0.30 0.25 0.20 0.15 0.10 0.05 0.00 0.53 □ Unfiltered Po-210 □ Filtered Po-210 O Unfiltered Pb-210 ■ Filtered Pb-210 al Lilji 2 Figure 4-12. ^'‘ ^Pb and ^*^Po (dpm/L) in unfiltered (total) and filtered (dissolved) samples collected during winter 2000 cruises. Refer to figure 3-1 for station loeations. Data bar for unfiltered ^'^Po activity (0.53dpm/L) for station 1 is truncated for clarity. 21 0 Particulate phase Po activities in these samples were much higher than those o f the dissolved phase, with an average particulate/dissolved ratio o f 5.74; correspondingly, the ratio for ^'V b is only 2.03, indicating a stronger particle affinity for ^'V o. During 70 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • 910 910 • • • • a storm period, both Pb and Po had higher activities in particulate phase relative to dissolved phase, which was manifested by high particulate/dissolved ratio for stations 1, 2, and a l. It is thus inferred that during a major precipitation event, high concentrations o f particulate matter in the water column will effectively scavenge ^in 9 10 9 1 0 both Po and Pb from water column, with Po having a higher efficiency due to stronger particle affinity. The coneentrations of SPM were not measured for these samples, but sediment traps were deployed (data shown below) so that the roles of 91 0 91 0 SPM can be inferred. At the same time, the average ratio o f Po/ Pb activities in particulate phase (5.11) is higher than those in dissolved phase (1.56). However, if the 910 91 0 samples from stations 1, 2, and al are excluded, dissolved Po and Pb activities are essentially in equilibrium (^'^Po/^’°P b= l.l). In summary, the data show that ^'*^Po • 910 910* has stronger particle affinity than Pb, and excess o f total Po in water columns at many stations may indicate intensive recycling o f ^’*^Po-rich particles in the water column. 4.2.2 Result of Three Sediment Traps Deployed during Feb.17-Feb.29 Table 4-2 summarized the data for three sediment traps that were deployed and retrieved in February 2000 sampling cruises. There were a number o f storms during the sampling cruise, the major one was on February 20 and 21. The bay water was turbid on February 22 during a sampling cruise. Immediately after a major storm. 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediment accumulation rate (SAR) was the highest (4.10g/cm /yr for station 4 deployed during February 17-22), and it decreased quickly with time (1.44 g/cm^/yr for station 2 deployed during February 25-29). This indicated that the storm runoff was very important on the particle inventory o f SDB. Table 4-2. Summary o f Sediment Trap Results (SAR: Sediment Accumulation Rate) Station Depth Collection Particle SAR* Excess ^'°Pb Excess ^ ^ '*T h Number (m) Period Weight (g) (g/cm 7y) (dpm/g) (dpm/g) 4 9.5 Feb. 17-22 3.98 4.10 3.41 ± 0.11 5.09 ± 0.14 3 11.3 Feb.22-25 1.24 2.12 12.18 ± 0.33 16.43 ± 0.39 2 11.6 Feb.25-29 1.12 1.44 13.39 ± 0.36 18.48 ± 0.44 Upon retrieval o f the sediment traps, the collected particles were weighed and excess ^'^Pb and activities measured using gamma counting. Unfortunately, ^'**Po activities were not measured. Activities o f excess Pb and Th in the collected particles had two intriguing features. First, despite the diverse particle settling rate, the overall radioisotope scavenging rate was rather uniform. For example, for individual • • 9 1 0 trap, the daily scavenging rates o f Pb were 2.71, 3.01 and 2.99 dpm for station 4, 3 and 2 with depths of 9.5, 11.3 and 11.6m, respectively. For ^^"^Th, the daily scavenging rates were 4.05, 4.06 and 4.12dpm, respectively, close to its production rate in the overlying water column. Apparently the radionuclides being scavenged 72 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. onto particles are ‘diluted’ when there are excessive amount o f particles. This showed a typical ‘particle concentration effect’, or PCE, which will be discussed in detail in chapter 5. Secondly, It is interesting to note that there was a very good correlation 9 9 1 n 9 9 id (R ==1.00) between the activities o f Pb and Th (in dpm/g) in the sediment trap particles (see figure 4-15). The effect o f the depth o f each station over the accumulation rates o f radionuclide is not obvious; however, the shallowest station 4 did collect less radionuclides than the other two deeper stations. H 25 20 15 1 0 5 0 y = 1.32x4-0.54 R2= l.O i 5 Pb 10 15 o, •a 20 15 1 0 5 0 1 3 SAR ^ Figure 4-13. a: correlation between ^*^Pb and ^‘ * '‘Th activities in dpm/g in sediment 9 1 0 99 4 trap data; b: particle dilution effect for Pb and Th in sediment trap data. Top line: ^^'‘Th, with a fitting function ^^“ ^Th—5.19*SAR+26.585 (r^=0.99); bottom: ^'*^Pb, with a fitting function o f ^'*’Pb=-3.90*SAR+19.63 (r^=0.98). 234n As shown in the previous results (see table 4-1), the upper limit o f the sedimentation rate in SDB are 0.70 cm/yr or 1.4g/cm /yr on average. The sediment accumulation rate was 4.10g/cm /yr calculated from the February 17-22 sampling period. This is converted to a much higher (~ 3 times) sedimentation rate than the upper limits o f the 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ♦ • • 9in i’ 5 7 average sedimentation rate estimated based on sediment Pb and Cs profiles (table 4-1). After the major storm, the sedimentation accumulation rate quickly dropped to the average accumulation rate. 4.3 Results o f Samplings around SDB in June-December 2002 The radioisotopic and salinity data for samples collected from stations around SDB during June and December 2002 (including rivers and outer sea samples) were listed in Appendix 1. Refer to figure 3-2 for station locations. Two sampling trips (July 6-7, September 6) that collected samples during tidal cycles will be discussed later. O f the listed parameters, not all were measured due to the availability o f time, especially for ^^“ ^Th, which has a short half-life. ^^''Th data would be virtually unusable if the sample was not processed and analyzed within two weeks because o f the complicated » * 234 ingrowth and decay correction o f Th. The data were grouped together in figures 4- 14, 4-15 and 4-17, representing sampling trips on June 6-11, 2002; November 9, 2002; December 25-26, 2002, respectively. ^ * * ’Pb data from December 2002 sampling was not available yet, therefore only ^*‘ ’Po data is reported. As discussed in chapter 2, rivers contribute only a negligible volume to San Diego Bay (Chadwick et al, 1999) except for the period o f major winter storms. The sampling covered all major rivers draining into SDB, including Sweetwater River, 74 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Otay River and Chollas Creek. San Diego River and Tijuana River, two major rivers in the vicinity not draining into SDB were also sampled for reference. From figures 4- 16 and 4-18 it is seen that rivers usually have higher as well as than SDB, but there are large variations. In the summer 2002 sampling period (figure 4-16), San Diego River and Otay River (during the period o f a low tide) had very high ^'*^Pb and activities. San Diego River had a total ^'®Po and ^’°Pb aetivities o f 0.61±0.02 and 0.39±0.01dpm/L, respectively; Otay River had values o f 0.41+0.02 and 0.27±0.01 dpm/L, respectively, during a low tide period. These values were 3-6 times higher than those o f the open ocean (stations SMB 1, SMB2, IP and OBP) and even rainwater, and much higher than stations inside SDB. These unusually high aetivities ■ y 1 A have not been reported before. Hence, resuspension o f Pb-rieh surfaee sediment was suspected to be the cause o f these high values. It was observed that the sampling loeations for San Diego River and Otay River were both characterized by shallow depth, organic-rich surfaee sediment and high turbidity. Unfortunately no separation o f dissolved and particulate phases were carried out. Low activities o f Pb and Po in Otay River (0.04 and 0.11 dpm/L, respeetively) during a high tide and Sweetwater River (0.024 and 0.027dpm/L, respeetively) during the summer 2002 sampling season were probably due to different reasons. It was observed that during the sampling o f Otay River the water was flushing back into the 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. a 0.15 i Q ]Q ' □ Po-uf BPo-f 0.05 0.00 llii. n , 1 , 1 0.41 0.61 0.15 0.10 0.05 0.27 0.39 2 3 4 5 6 7 8 9 10 11 12 0.00 □ Pb-uf BPb-f il,D .1 1 1 ,! ! ,n ,1 ,n 0 1 2 3 4 5 6 7 8 9 10 11 12 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 □ Po-uf B Pb-uf I I 1 , n , Di on, O B U . I 1 2 3 4 5 6 7 8 9 10 11 12 I 0.06 L 11 li □ Po-f BPb-f 1 ....... 1 1 2 3 4 5 6 7 8 9 10 11 12 1 .2 1 0.8 0.6 0.4 0.2 0 □ Th-uf BTh-f 1.1.,I . " 1 .. .. ... ... . 1 'T - 2 3 4 5 6 7 8 9 10 11 12 Figure 4-14 a-e: Radioisotopes in water samples collected in summer 2002 at stations around SDB. Refer to figure 3-2 for station locations. Numbers 1-12 on the horizontal axis represents stations SM B l, SMB2, OBP, IP, SI, HIW, HIE, CW, SW, OTR (low tide), OTR (high tide), and SDR, representatively, a: ^'*^Po in unfiltered and filtered water samples; b: Pb in unfiltered and filtered water samples; c: Po and Pb in oin oin» 0 ” 3 4 unfiltered water samples; d: Po and Pb in filtered water samples; e: Th in unfiltered and filtered samples. Note that figures c and d are different presentations of the same batch o f data as figures a and b for clarity. Otay River from SDB because of water level in the bay during high tide was higher than that o f Otay River, so the data might represent bay water instead o f river water. 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Sweetwater River sample was eolleeted about 5 miles upstream o f its estuary, so it was not affected by tides. The flow was negligible during sampling, suggesting that the water sample collected might not be representative o f Sweetwater River. In fact, the data from Sweetwater Authority (Michael Garrod o f the Sweetwater Authority, personal correspondence) showed that there has been no flow over the Sweetwater Dam since 1998. As a result, the trickling flow at Sweetwater may be either groundwater or sewage water and it should not be representative o f the regular Sweetwater River water. For the same reason, the source o f Otay River water during dry season should also hardly be river water in the common sense. These characteristics o f river water flows have important implications on the water budget and radionuclide inventory o f the entire SDB. However, as will become clear in ^ 1 n 01A subsequent chapters, the forms and concentrations o f Pb and Po in rivers have little effects on the overall pattern and mass balance o f this radionuclide duo in SDB due to the very low flux o f rivers as a whole. This issue will be addressed in more detail in chapter 6. Samples were collected from San Diego River and from stations SI and GP during a minor precipitation event on November 9, 2002 (cumulated rainfall during November 8-9 was 5.0 mm, Michelle Chambers o f the West Region Climate Center, personal • » " 2 1 A " 2 ^ 4 • • • • communications). Po and Th activities in unfiltered, dissolved and particulate 77 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 n 9 10 phases were measured, as shown in figure 4-15. Rainwater Po and Pb were 0.24±0.01 and 0.57±0.02 dpm/L, respectively (not shown). The ^’^Po activity o f the San Diego River water sample (unfiltered) was 0.24± 0.08 dpm/L, very close to that o f the rain water o f 0.23±0.07dpm/L, suggesting that rain water might be the chief 0 1 n source o f Po in loeal rivers during a storm event. This possibility is based on the faet that the normal flow o f the San Diego River, although a major river in the entire region, is negligible compared to the flow during a storm. 0.5 «4 0.3 0.2 ■ total-Po ■ particulate Po ■ Part.Th 1 dissolved Po 1 Diss.Th ITSS/50(mg/L) SDR SI GP 9 1 0 9 T 4 Figure 4-15. Po and Th activities (dpm/L) and SPM eoncentrations (mg/L, scaled down by a factor o f 50 in the graph) in water samples collected on November 9, 2002, during a minor precipitation event. Refer to figure 3-2 for station locations. 910 910 • Unlike commonly observed rainwater Po/ Pb ratio o f 0.1 to 0.2 (Appleby and Oldfield, 1992), the ratio obtained in this study was 0.43. This may be indicative o f a long residence time o f ^'®Pb in the air for a dry region sueh as San Diego Bay. 78 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. However, this data might not be representative because the rainwater sample was eolleeted near the end o f the precipitation event. The initial precipitation is generally 1 n 0 1 n considered to contain higher levels o f both Po and Pb than subsequent precipitation due to a ‘first flush effect’ (Fuller, 1982). This notion is supported by the 9 1 n fact that if this data is used to extrapolate annual atmospheric fallout o f Pb, there will be a significant underestimate (-40% ), which cannot be accounted for by dry fallout. Obviously, more sampling o f rainwater is needed to confirm its radioisotope concentrations. Wet season sampling was carried out on December 25 2002, which included two additional stations at the western and eastern shore o f south SDB (WSB and ESB). ESB is a shallow intertidal region covered almost completely by floating macroalgae, and the water was extremely turbid. Floating macroalgae was removed prior to filtration, but the filtration was stalled completely due to high turbidity and perhaps large amount o f colloidal organic matter that clogged the pores o f the filter paper. As a result, only the total activity o f ^'^Po could be obtained. Even with limited data, it is 9 1 0 clear that the abnormally high Po activity (0.233±0.09dpm/L) was largely due to the resuspension o f bottom sediments. Comparatively, the station WSB is a sandy shore with moderate amount o f suspended inorganic particles, and total ^''’Po activity 21 0 should be a result o f high particulate-phase Po contribution. Considering a major 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. precipitation event (31.3mm, from Michelle Chambers o f the West Region Climate \ Center, personal communications) 4 days prior to the sampling, high ^'®Po activities in stations at inner bay (G-Pier, WSB and ESB) could be the after effect o f the precipitation. As will be shown in modeling results, precipitation has major effects on the levels o f ^'°Po, especially for inner bay (i.e. central and south SDB) immediately after the precipitation events. 0.08 0.06 0.04 0.02 0 2 1 0 t IT otal □ dissolved □ particle 0.23 I 2 3 4 5 6 7 9 10 Figure 4-16. Po activities (dpm/L) in water samples collected on December 25, 2002. Refer to figure 3-2 for station locations. Numbers on the horizontal axis correspond to sampling stations as follows. 1:TJR; 2: OTR; 3: SWR; 4: CC; 5: SI; 6: HIW; 7: HIE; 8: GP; 9: WSB; 10: ESB. Data bar for sample #10 is truncated for clarity. A common feature shared by patterns in figure 4-14, 4-15 and 4-16 is that there is a clear gradient o f Po and Pb (and their activities in dissolved and particulate phases, if measured) from near the bay mouth to the region further inside the bay. 910 910 • • Shelter Island (station SI) always has higher Po and Pb activities than the stations further inside the bay (HIW, HIE and GP). There is only one exception for station GP in figure 4-16, which was higher than station HIE. ^'°Pb and "^'"Po activities at station 2 1 0 t 80 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. SI are often comparable to those o f the open ocean stations (SM B l, SMB2, IP and OBP), as shown in figure 4-16 and 4-18. As will be discussed in chapter 6, this similarity between station SI and those o f the open ocean must have been a result o f more efficient exchange with open ocean due to its closeness to the open ocean, and the gradient toward the inner bay is due to stronger particle scavenging that brings 1 n 7 1 n • down the Pb and Po in the water column. Otay River, Sweetwater River, and Chollas Creek all had relatively high ^'"^Po activities, averaging about 0.05dpm/L. Tijuana River had a low ^*^Po activity of 0.029+0.005dpm/L, but that might not be representative for Tijuana River because o f a large farmland upstream from the sampling site. Extremely high turbidity and unnatural coloration o f the water suggested that the water might have been contaminated by farm land effluent. • »« Tin Tin The effect of rain can be recognized if the patterns o f Pb and Po are compared for samples collected during dry season (figure 4-14) and during wet seasons (figure 4-16). The result is shown in figure 4-17. It is clear that there is essentially no significant difference between the two seasons. This suggested that the effect o f rain is 21 0 insignificant on the activities o f Po, or at least the effect could be largely dampened rather quickly shortly after a precipitation event (the December sampling was done 4 81 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. days after a major precipitation event o f 31.3mm). The reasons for the lack o f difference could be 1) a fast exchange with outer sea for north SDB, and 2) efficient particle scavenging driven by higher particulate matter supplied by surface runoff during storm events; or 3) tidal patterns, which affect transient Pb and Po activities for north SDB, as shown in section 4-5. These important factors will be modeled and diseussed in chapter 5 and 6 subsequently. 0-10 ■ Dry Season □ Wet Season 0.08 0.06 0.04 0.02 - 0.00 SI HIW HIE GP jn c 9 1 0 Figure 4-17. Comparison between activities o f unfiltered Po for samples collected in dry and wet seasons o f 2002 from stations across the north SDB. Refer to figure 3-2 for station locations. Vertical axis is activities in dpm/L. 4.4 Results o f September 6, 2002 Sample Cruise Figure 4-18 a-f show the variations o f water column TSS, ^"^Po, ^''^Pb, and ^^"^Th concentrations across SDB and adjacent outer sea. For a detailed description o f the suspended particulate matters o f each water sample, please refer to table 3-3. The 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 12 l.l 1 1 1 - 0.12 b 0.08 □ d -P o -2 1 0 ■ d -P b -210 0 71.3 0.16 □p-Po-210(dpm /L ) 0.12 0.08 0.04 0 ilp-Pb-210(dpm /L ) J □ p-Po(dpm/g) p-Pb(dpm/g) 10 20 30 40 50 60 70 80 90 100 3.0 - j 2 Q j n p -T h -234 .0 0.0 ' ■ d-Th-234 I jijJl j i J B J i k J i i j 10 20 30 40 50 60 70 80 90 100 Figure 4-18 a-e: Concentrations o f chemical species for water samples collected on September. 9, 2002. Refer to figure 3-3 for station locations, a: SPM in mg/L; b: ” 2 1 0 91 0 » activities o f dissolved Po and Pb (dpm/L); c: volume-based particulate phase ^'°Po and ^'V b (dpm/L); d: weight-based particulate phase ^’'’Po and ^'‘ ^Pb (dpm/g SPM); e: dissolved and particulate ^^^^Th (dpm/L). error bars o f data points were not shown for clarity. For ^'*’Pb and ^'°Po data the relative errors are constantly less than 5%; ^^"^Th activities had large errors (up to 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 30%) due to the uncertainties in decay and ingrowth correction, as well as the system errors from the instrument (a-(3 counter). 0.4 0.3 0.2 0.1 i 0 □ d-Po-210 ■ d-Pb-210 Hp-Po-210 ■ p-Pb-210 IhVlh 60b 70b 80b 12.626 90b 100b ■ TSS □ p-Th-234 0 d-Th-234 50b 60b 70b 80b 90b 100b Figure 4-19 a and b: Activities o f radionuclides in deep water samples at stations 50- 100 collected on September 6, 2002. All stations are located in the open ocean except for stations 50. Refer to figure 3-3 for station locations, a: ^'‘ ^Po and ^'°Pb in dpm/L; b: SPM concentration (mg/L; value for sample 70b truncated for elarity), particulate ')'1A and dissolved Th in dpm/L. The profiles o f SPM, weight-based particulate ^'°Po, ^'°Pb activities, dissolved V o, Pb activities, and Th activities were clearly different between two water bodies inside and outside SDB. From inside SDB toward outer sea, the activities of dissolved ^'V o, ^'°Pb and weight-based particulate ^’V o, ^’'^Pb (in dpm/g) increased (Figures 4-20b, d, e) while TSS concentrations decreased (Figure 4-20a). There was 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • • 21 0 210 no obvious trend for volume-based particulate Po and Pb activities (in dpm/L) (see figure 4c) across the bay mouth because o f the conflicting effects o f particle abundance and availability of dissolved radionuclides. There was also no obvious trend for Th (in dpm/L) (figure 4e) across the bay mouth. TSS data here is compatible with the tindings o f Chadwick et al (1999) and the measurements by in- situ pumps deployed during both summer 1999 and winter 2000 sampling cruises, which yielded an average TSS concentration o f 4.2 mg/L inside SDB. The results from non-parametric (rank sum) t tests on the data shown in Figure 4 showed that waters inside and outside SDB were significantly different {p < 0.05) with respect to TSS (p=0.047), dissolved ^'‘ ^Po (p=0.009), dissolved ^'^Pb (p=0.016), weight-based particulate ^*^Po (p=0.009) and ^'*^Pb (p=0.016). No significant differences were found for the volume-based particulate ^'^Po(p=0.60) and ^*Vb(p=0.25), particulate ^^^Th(p=0.96) and dissolved ^^^Th(p=0.31). The residence time o f Th can be calculated with dissolved Th activities using the following equation, assuming unidirectional uptake (Cochran, 1992): 34.8 ,. ,, ■ C 2 3 4 - ~ 2-------- (4-3) ^238 _ 2 ^234 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where ')'\A T 234 is the residenee time (days) o f dissolved Th ^238 is the aetivity in dpm/L for (~2.4dpm/L for SDB estimated from salinity relationship) 'J'lA A 234 is the aetivity for Th in dpm/L. The residence time calculated from dissolved Th aetivities using equation 4-3 increased from 1-8 days inside SDB to over a year outside SDB. Since residence times o f particle-reactive species are negatively proportional to the strength o f particle scavenging, the drastically different residenee times for Th from inside SDB to outer sea again underscores the distinct oceanographic processes occurring inside and outside SDB. Lastly, to investigate the possible particle concentration effect (PCE) on the relative 9 10 9 10 • aetivities o f Pb and Po m dissolved and particulate phases, the logarithmic of SPM concentration and distribution coefficients (Kd) are plotted in figure 4-20, as commonly done in previous studies (Turner and Millward, 2002, and references therein). Kd is indicative o f the particle affinity o f a chemical species and is defined as follows; 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. K d = | (4-4) Where P is the concentration o f the sorbed radionuclide (w/w), C is the dissolved-phase concentration o f the radionuclide (w/v). As suggested by PCE, SPM concentration should be inversely related to Kd, and thus their logarithmic values should be negatively proportional to each other. The resulting plot is shown in figure 4-20. The data in figure 4-20 has special significance for the subsequent modeling o f the particle-radionuclide interaction (see chapter 5). For the same reason, figure 4-20d is shown despite that it includes data points from February 16, 2003 and November 9, 2002 samplings. It is seen from the above figures that T T d . Tin Tin Th, Pb and Po all have appreciable particle concentration effect (PCE) as manifested by negative relationship between Kd and SPM concentrations. The difference in the slopes o f the trend line o f each radionuclide may be caused by its availability as well as its affinity to different components o f the particulate matter. For example, as mentioned in chapter 1, ^'**Po may have the highest affinity to organic particles among the three radionuclides. As a result, if the differences in the SPM 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 14 13 1 2 1 1 1 0 9 a;Th y = -1.3063x+ 13.903 = 0.2763 1.5 2.5 14 : 0 0.5 1.5 b:Pb 1 I y =-0.4619 x + 12.931 = 0.4869 1 2 - i ! ♦- 1 1 r 2.5 14 13 1 1 2 1 1 ♦ 1.5 y = -0.2229x+ 13.092 R^ = 0.1009 2.5 c: Po(l) 14 13 I ] 1 2 : 1 1 i d; Po(II) 0 1 >v * - A A 1 A O ,, _ 1 _ 1 A OAA ♦ ♦ ♦ y = -0.2128x+ 12.972 R^ = 0.0683 Figure 4-20a-d: Relationship between In(SPM), in the x-axis, and In(Kd) in the y-axis for ^^'‘Th(a), ^'°Pb(b), ^*“Po(c) for surface samples from September 6, 2002 cruise, and ^^°Po for all available dataset (d). Particle concentration effect (PCE) is indicated by the negative relationship in all cases. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • • • • 0 1 n concentrations are ehiefly eaused by biologieal produetivities, PCE for Po may be the smallest among the three. On the other hand, if most of the dissolved ^'^Po has already heen consumed, the composition o f SPM would have no effects on PCE, i.e. the slope o f the ln(Kd)-ln(SPM) trend line heeause o f a simple dilution effect. From the limited data, it ean be inferred that ^^"^Th has the strongest PCE as shown by the largest eoefficient (-1.31) and ^'°Po has the weakest PCE (-0.21). At the same time, within the median range o f SPM eoneentration o f about 3mg/L in SDB, partiele affinities for these radionuelides follow the sequence o f ^'*’Po>^^'*Th>^'^Pb, different from the sequenee o f ^^'^Th >^’Vo>^'*^Pb proposed by Sarin et al (1994). However, Sarin et al (1994)’s proposal was based on the open ocean (Arabian Sea), which had much lower SPM concentrations. If such low SPM concentrations were to be applied in the ahove dataset, a sequence similar to that o f Sarin et al (1994) would have resulted due to a minimum PCE for the SPM-depleted open oeean. During September 6, 2002 cruise, three gravity eores were also eolleeted at stations 10, 30 in the central and north SDB, as well as at station 90 outside SDB (figure 3-1). The profiles o f excess Pb and Cs o f these cores were shown in figures 4-21 a-e. The sedimentation rate at station 10 could not he caleulated due to negative exeess ^'°Ph aetivity at a shallow depth, but if that point was eliminated, a sedimentation rate o f 0.46+/- 0.25 cm/yr was obtained, and this result is eonsistent with the '^’Cs profile. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. which had a peak at 19.5cm depth. Sedimentation rate at station 30 was calculated to be 0.30+/- 0.10 cm/yr. Comparatively, the ^^^Pbex inventory for station 30 was again significantly higher (45%) than that o f station 10, similar to the pattern observed for summer 1999 sediment cores but with a smaller contrast (note that the sampling locations are different). As for station 90 outside SDB, the excess was extremely » » • • 910 • • » high (5.7-1 Idpm/g). This is the highest Pb activities we have ever observed at any stations, suggesting the situation at station 90 is a special case. The surface sediment o f station 90 was found to be very fine, well-sorted and had high water content. It is suggested that the region around station 90 was a preferential settling site for fine 91 0 particles that are enriched with Pb. Perhaps intensive resuspension facilitated a more effective scavenging o f ^’^'Pb from the overlying water column. This finding is 9 1 0 supported by extremely high particulate Pb concentration (both volume- and weight-based concentrations) in the surface water at this station (figure 4-18). During September 6, 2002 cruise, surface sediments at each station were also collected by a van Veen grab and their excess ^'°Pb and ^^"*Th activities were measured. Due to the difficulty o f keeping sediment-water interface intact in the grab during sampling, these samples were not necessarily representative o f the patterns o f the very surface sediment. There was no apparent trend o f either ^'^Pb or ^^"^Th from inner SDB to outer sea (Figure 4-22). But ^^°Pb and ^^'^Th had good correlation with 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 3 2 0 0 10 20 30 40 y = 2.9693e r L= 0.3172 -♦ -P b -2 1 0 e x —A—Csl37 Depth (cm) 5 n .- - 0 0 6 9 9 X 5 4 3 2 1 Q — A ™ -A ~a 0 5 10 depth(cm) 15 I pjj2 1 Oex Csl37 12 Cs-137 Pb-210 9 0.5 6 3 0 0.5 2 4 depth (cm) Figure 4-21 a,b,c: Depth profiles o f excess 210Pb and Cs-137 for stations 10, 30 and 90, respectively, for the September 6, 2002 cruise. Sediment cores were collected with a gravity corer. each other, especially for stations within SDB, or for all stations if station 9 is not considered (Figures 4-23), indicating similar effect o f partiele scavenging for both radionuclides. Again, station 90 has high activities o f both Pb and Th, supporting our suspicion that station 90 may be a depocenter that accumulates fine-grained 91 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediments as well as particle-reaetive radionuclides. If station 90 is treated as a special case and excluded in figure 4-23, the correlation between and would be ^*°Pb =1.18*^^^Th -0.07 (r^=0.84). Compared to the correlation o f ^^°Pb =0.76*^^^Th - 7 « • • 0.41 (r =1.00) for the sediment traps deployed in winter 2000 (refer to figure 4-13a), lower activities o f ^^"^Th relative to ^'*’Pb for sediment surface seem to reflect the age o f the sediment. The good correlation between the two radionuclides in both cases suggests that both are effectively scavenged from water column by settling particles. 12 9 6 3 0 □ Pb-210 ■ Th-234 (llhnn^ rifi 10 20 30 40 50 60 70 80 90 100 Stations Figure 4-22. Activities (in dpm/g dry weight, vertical axis) o f ^'°Pb and ^^"^Th in surface sediments collected by van Veen grab during the September 6, 2002 cruise. y = 0.4872x+ 1.5703 ♦ = 0.4367 6 4 2 ♦ 0 T ' o (N I -O C u 0 2 4 6 8 10 Th-234 Figure 4-23. Relationship between ^ * * ’Pb and ^^"^Th (dpm/g) for the dataset shown in figure 4-22. If station 9 is excluded, the correlation would be y=1.18x-0.07 (r^=0.84). 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4.5 Results o f Tidal Cycle Samplings 4.5.1 Result from July 6-7 Sampling Trip Figure 4-24 summarized the result o f the July 6-7 sampling trip, during whieh a series o f samples were collected at station SI (Shelter Island) at 6 different times to cover an entire tidal cycle over a 13-hour period. O f the 6 samples, 3 of them were analyzed for both particulate and dissolved phases for Po and Pb, as shown in figure 4-25a 7 1 n 7 1 0 and 4-25b. The eovariance o f Pb and Po with tidal magnitude was significant , 7 1 n 7 1 n and a fairly close covariance between Pb and Po supported the validity o f the 7 1 n 710 tidal influence. Both Pb and Po had higher activities during high tides when the water at this station was mostly occupied by water from outer sea, which had higher 210 210 • concentrations o f both Pb and Po. When tide was low, the water at this station 710 7 1 0 was replaced by water from inner bay, which had lower Pb and Po activities due to stronger particle scavenging (see figure 4-2, 4-12, 4-16 and 4-18). The variation of ^'®Pb and ^’®Po seemed to be leading the tidal phase change. Comparatively, ^^“ ^Th variance with tidal phases was more irregular and seemed to lead the tidal phase. • 71 0 From figure 4-24b and c it was caleulated that the portion of dissolved phase Po vs. total ^'°Po (averaging 22%) is lower than that o f ^'°Pb (averaging 30.3%), indicating 2 10 21 0 that Po should be more particle-reactive than Pb. This finding is eonsistent with 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.10 0.09 0.08 j s 0.06 3 0.05 0.04 ^ 0.03 0.02 0.01 0.00 Total Po-210 ■ ♦ - - Total Pb-210 - A — Th-234 -)K- - Tide(m) j * ■ 'iK -. - 4 -- 0:00 2:17 4:45 6:47 11:52 13:32 time (hh:mm) 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 - 0.2 E o - ~ a S' H Figure 4-24. Variations o f activities o f total ^’^Po and with tidal phase for water samples collected on July 6-7, 2002. Errors for both ^^*^Pb and ^‘^'Po are ~5%, and 10-30% for (not shovra for clarity). X-axis is the time o f sampling in hh:mm counting from 11:03pm o f July 6, 2002. 2 1 0 t 0.08 0.06 0.04 0.02 0.00 0.06 0.04 0.02 0.00 iD iD ln total ^ dissolved □ particulate T1 r i ” 71 f t Figure 4-25 a.b: Activities o f total, particulate and dissolved Po (a) and Pb (b) for samples 1, 4 and 6 (numbering o f stations is shown in figure 4-24). All activities are in the unit o f dpm/L, with relative errors about 5%. the result shown in figure 4-20 and related discussions. The data also show that during low tides, portions of dissolved phase vs. total activities were lower for both dissolved 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ^'®Pb and This is in line with the relatively stronger particle scavenging for inner bay. 4.5.2. Variation of Water Column TSS, ^'Vo, and with Tidal Phase. Figure 4-26 a and b are the result o f a major sampling session that took place at February 16-18, 2003. Stations SI and GP were both sampled 10 times over a 25-hour period that spanned two complete tidal cycles. The parameters measured included the following: suspended particulate matter (SPM; grain size<0.47|am), dissolved-phase 01ft • 0 1 f t « Po, volume-based particulate-phase Po (in dpm/L), weight-based particulate- phase ^*® Po (in dpm/g particulate matter). On the figures the error bars were not shown for clarity, and the relative error were consistently below 5% thanks to the 0 1 ft 0 1 ft improved analytical method for Po as described in chapter 3. Pb data is not available at the time when this manuscript was prepared. Variation o f ^’°Po and TSS with tidal phases at station SI exhibited several intriguing features (Figure 4-26a). First, dissolved ^^°Po activity varied closely with the tidal 0 1 n pattern, i.e., high tide occurred concurrently with high Po activity and vice versa. Second, TSS moved oppositely with dissolved ^“^Po activities. Finally, the tidal phase correlated poorly with volume-based particulate Po aetivity, but strongly with 91 0 weight-based particulate Po aetivity. These features are similar to what we 95 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. — • — d-Po - - - -p-Po (dpm/lOOL) 5 K p-Po(dpm/g SPM) — — tide(m) 0 SPM(mg/L) tr . - - 4:48 9:36 14:24 19:12 0:00 •2 -♦— d-Po — - tide(m) ■ - - h e - -p-Po(dpm/100L) — * — p-Po(dpm/g SPM) 4 SPM(mg/L) 8 6 4 2 r W- 0 4:48 14:24 0 :36 19:12 0:00 Figure 4-26 a, b: Variations o f 210Po, SPM concentrations with tidal phases for water samples collected at stations SI (for figure 4-26a) and GP (figure 4-26b) during February 16-18. Trend lines are smoothed for clarity. observed before (figure 4-24) and are also consistent with the static geographic patterns shown in Figures 4-2,4-12, 4-16 and 4-23 a-e. They were clearly resulted from 96 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 0 exchange o f two waters with distinct TSS and dissolved/particulate Po activities. For example, the close negative co-variance o f dissolved ^’*^P o with tidal phase showed that during high tides, water at station SI was replaeed by outer sea water with high concentrations o f dissolved ^'*^Po and vice versa during low tides, when the water at station SI came mostly from inner SDB with lower dissolved concentration of ^’®Po. The opposite pattern is observed for TSS for the same reason, but TSS is higher in bay water and lower in outer sea water so TSS show positive co-variance with tidal 9 1 n • . phase. The vagueness o f volume-based particulate phase Po variations with tidal phase may possibly due to the conflicting effect o f particle scavenging and availability ^ 1 n ...w " 7 1 D o f dissolved Po for scavenging. The outer sea has high dissolved Po but low TSS, so volume-based particulate-phase ^’^Po activity is constrained by the • • 9 1 f ) • availability o f TSS; conversely for bay water, TSS is high but dissolved Po is low. 91 0 These two conflicting factors reduce the contrast o f volume-based particulate Po activities between waters inside and outside SDB, and as a result tidal phase has 9 1 0 limited effect on the volume-based particulate Po. However, for weight-based particulate phase ^’®Po activities, for the same reason as stated above, the contrast between waters inside and outside SDB is strengthened by the differences in TSS and dissolved ^'*^Po. Compared to station SI, which is closer to the bay mouth, no consistent patterns were observed at station G (Figure 4-25b) loeated further inside SDB. However, good 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • 910 positive correlations can be found between TSS and volume-based particulate Po, 91 0 and there is a negative correlation between TSS and weight-based Po. The latter correlation was more ambiguous than the former one, and seemed to have lags and leads with TSS. Interestingly, these species (TSS, weight and volume-based particulate ^*^Po) seemed to have a frequency o f twice o f that o f the tidal phase. If this pattern was not an artifact, it might have reflected a delayed oscillation effect caused by tidal exchange. It should also be pointed out that the range o f both dissolved/particulate ^‘‘ ’Po variations in station G was 2-3 times smaller than that at station SI o f north SDB (Figure 4-25 a and b). This important pattern will be addressed in chapter 5. Statistical analyses using Statistical Analysis Software (SAS) confirmed the above observations. A simple multivariate linear correlation analyses were conducted to determine the variability o f the ^*°Po o f different phases and TSS with tidal phase. A confidence interval o f 95% was used in all the statistical analyses. At station 4 (Figure 4-25a), TSS and dissolved and weight-based particulate ^'‘ ’Po strongly correlated with the tidal phase with the correlation coefficients o f -0.81 {p = 0.0004), 0.90 ip = 0.0042), 0.88 {p = 0.0007), respectively. Volume-based particulate ^'*’Po correlated weakly with the tidal phase with a correlation coefficient o f 0.38 ip = 0.27). Comparatively, little or no correlation was observed at station 4 ’ (Figure 4-25b) 98 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. between the tidal phase and TSS, dissolved and weight-based particulate ^'^Po and volume-based particulate ^''^Po with correlation coefficients o f 0.17 {p = 0.96), -0.02 ip = 0.22), 0.00 ip = 0.64), and 0.43 (p = 1.0), respectively. ^1 2 3 ^ Po-d (dpm/lOOL) 4 1 2 A Po-p(dpm/100L) O "sPM(mgA.) X ^o-p (dpm/g) Figure 4-27 a-c: Geographic trends o f SPM, dissolved and particulate ^'*^Po activities for stations SI, HIW, HIE and GP (marked as 1, 2, 3, 4 on the figures) Besides temporal variations, spatial variations o f SPM and radionuclide activities were also obtained by measuring these parameters along stations SI, HIW, HIE and GP (marked as I, 2, 3, 4 in figure 4-27) along the north SDB at about the same time. Generally there is a decreasing trend for radionuclide activities and an increasing trend for SPM with a few exceptions. This pattern is compatible with our finding before, i.e. higher SPM for inner SDB and lower activities for both dissolved and particulate phase radionuclides. 99 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 210 4.6 Atmospheric fallout o f Pb measured by fallout collectors and salt marsh sediment cores A 13-month deployment o f fallout collector on the roof o f the Microanalysis Laboratory near the Geochemistry Lab located inside USC campus (central Los Angeles) gave an annual fallout rate o f 0.2H-/-0.01dpm/cm^/yr; a parallel deployment of a identical collector on station NM inside the Cabrillo National Monument for a 8- month period gave a fallout rate o f 0.17dpm/cm2. Taking into consideration o f the time (0.75 year) as well as the difference o f wet and dry season (the deployment period covered most o f the wet seasons), a best estimate for the fallout amount during the whole year should be about 0.20dpm/cm2/yr. To confirm the results o f the direct 910 147 atmospheric fallout collectors, Pb and Cs activities in two land sediment cores collected at station TJRl and TJR2 were also measured and the results were shown in figures 4-28a and b. The top layer o f the TJR2 core was organic-rich with some live grass, which was removed before being gamma-counted (other plant debris retained). The collection o f TJRl and TJR2 sediment cores was intended to calculate the 7 1 0 atmospheric fallout o f Pb, in addition to the atmospheric fallout collection. Based 7 1 0 • • on the Pb profile o f each core, the inventory o f atmospheric fallout is calculated as follows; 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ^ f Az*W z , Fatm= A . 2io*J dz (4-4) 0 Where 'y Fatm is the atmospheric fallout rate in dpm/cm /yr > .210 is the decay constant o f ^'°Pb in yr’' Az is the activity o f dry sediment in dpm/g Wz is the weight o f dry sediment at a certain depth in g; D is the depth interval o f each sediment slice in cm S is the cross section area o f the sediment core in cm z is the depth o f each slice o f sediment core in cm The atmospheric fallout rate calculated based on equation 4-4 for core TJRl was 0.27 ±0.07 dpm/cm^/yr. For TJR2, which apparently did not reach the ^'"Pb equilibrium depth, the apparent ^'"Pb fallout rate is 0.16±0.06 dpm/cm^/yr. Since equilibrium of ^'"Pb is not reached at this depth, this fallout rate should be considered as a lower limit. These values are consistent with atmospheric fallout collectors described above. Therefore, we conclude that the fallout rates o f 0.20 dpm/cm^/yr should be the best estimate for the SDB area. 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5 Ex Pb-210 Cs137 4 3 2 1 0 0 10 20 30 40 2.5 Ex Pb-210 Cs137 0.5 0-0'"'*^^^ 5 0 10 15 20 25 Figure 4-28 Depth profiles o f excess and '^^Cs for sediment cores collected from two stations inside the Border Field State Park to the south o f SDB. a: station T JR l; b: station TJR2. No reliable estimate o f sedimentation rate can be obtained from the ^''^Pb profiles. Tentative sedimentation rates estimated from '^^Cs profiles are 0.16±0.03cm/yr and 0.09±0.04cm/yr, respectively for TJRl and TJR2. 4.7 Salinity and Trace Metal Distributions and Variations With Tides Spatial and temporal variations o f salinity were also investigated during the February 16-18 sampling trip, the result is shown in figures 4-29 a-c. Salinity seems to co-vary weakly with tidal phase in both stations SI and G, and the average salinities for station SI and GP (34.15 and 34.01, respectively) suggests that it is the differenee between inner SDB with outer sea that has caused the weak co-variance as shown in figure 4- 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 29a and b. More over, figure 4-28c shows a generally decreasing trend for salinity from station SI to GP in the north SDB. It was not clear why station HIW consistently had the highest salinity. It is suspected that the remnant San Diego River channel, a closed basin nowadays, to the north o f this station is to blame. The results o f ICP-AES measurements o f trace metals Zn, Cu, Cr, Mn and Pb around San Diego Bay are shown in figure 4-30. Except for Zn, trace metals Cu, Cr, Mn and Pb all showed higher concentrations at Sweetwater River, Chollas Creek and Tijuana River, with Otay River showed only elevated concentration o f Cu. This pattern suggests that river runoff perhaps is the major contributor o f contaminant metals to SDB. The exception may be Mn, which should largely come from continental erosion and from estuarine bottom sediments. Since the stations are listed from left to right for locations from north bay to south bay, it can be inferred that these trace metals are fairly well mixed across SDB, no apparent gradient is present as Mahn et al (2002) observed. The pattern o f Zn is rather different from other trace metals (figure 4-3 Oe), indicating either different source o f contamination or different o f locations o f higher concentrations within SDB. Compared to the results o f Mahn et al. (2002), whieh showed rather eonsistent gradients for Cu, Mn and Zn for 28 locations along the central line o f SDB, our results might be affected by the fact that the samplinglocations are close to the shoreline. At the same time, tidal phase might also affect transient distributions o f trace metals, as will be discussed below. 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ■ o H - - D - -Tide ♦ salinity 2.5 9 5 0.5 0 -0.5 0:00 5:26 11:23 15:25 19:25 1 34.4 - 34.2 ._ _ . f 34 i E u S 33.8 33.6 33.4 u •a a - - Tide ♦ salinity 2 .5 - I 1 0.5 0 -0.5 0:00 5:26 11:23 15:25 19:25 34.4 34.2 E 34 U . O . 33.8 .- g * 'c 33.6 33.4 ♦ round 1 O - - round2 34.6 — -A — round3 -A- 33.: 2 3 4 Figure 4-29 a-c: Salinity variations with tidal phase at station SI (a) and GP (b) and variations with locations (c) in north SDB. For figure 4-28e, the label on the horizontal axis 1, 2, 3, 4 represent stations SI, HIW, HIE and GP, respectively, and curves labeled round 1, round 2 and round 3 represents samplings that took place when the second, fourth and tenth samples were taken (shown in a and b). 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 ^ a 20 - 15 ' 10 [Mn] [] n d n n u n 107.63 411.21 20 n 15 10 n Q ^ 1 ‘ “ L -, Q 7 8 9 10 11 [P b ] I 0 , B . ___ 1 2 3 4 5 6 7 8 9 10 11 i 8 c [C r] 6 4 2 Q W 0 , ffl r-®-7___1__ 1 2 3 4 5 6 7 8 9 10 11 8 d 6 4 2 0 -— , ^ [Cu] 2 3 4 387.6 1 2 3 4 5 6 7 8 9 10 11 30 I e 25 20 15 10 [Zn] D U D n U D 0 Q 1 2 3 4 5 6 7 8 9 10 11 Figure 4-30a-e: Concentrations o f trace metals Mn, Pb, Cr, Cu and Zn at stations across SDB. Samples collected in summer 2002. Vertical axes are concentrations in ppb, horizontal axis labels from 1 to 11 represent SI, HIW, HIE, GP, ESB, WSB, Pond, SWR, CC, TJR, OTR, respectively. Some data bars are truncated for clarity. The variations o f trace metals Cr, Mn and Zn with tidal phases for samples collected during February 16-18, 2003 are shown in figure 4-31. Compared to tidal phases, only Zn concentrations at station SI shows an anti-phase variations with tidal height, suggesting higher concentrations in SDB compared to outer sea. Other trace metals do not show appreciable correlation with tidal phase, neither do these metals at station G. At the same time, as shown in figure 4-32, there is no consistent patterns for trace 105 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. X I a . o . u -a c c d •a H Tide - A — Mn -♦— Cr — Zn 2.5 2 5 1 ,5 O'l ■ 0 .5 40 30 & 20 c N •a § 10 I 0-.00 2-28 526 8:36 1123 13:10 15:25 17:50 19:25 23:53 X I o. o. U ■ o X .S P '5 X T3 P -Tide -Mn -♦ Cr — Zn 2 0 1 - 0:00 2:28 5:26 ;:36 11:23 13:10 15:25 17:50 19:25 23:53 14 12 10 J D G , O . 8 N 6 • T S c a 4 2 0 Figure 4-31 Temporal variations o f the concentrations o f trace metals Mn, Cr and Zn with tidal phases for samples colleted during February 16-18, 2003. a: station SI; b: station GP. metal concentrations from station SI, HIW, HIE and G, which are progressively further south inside SDB. The lack o f gradient among these samples may come from the short spatial distances. The fact that the sampling locations are close to shore may 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. also contributed to the lack o f gradient in these samples. 25 -Zn -Mn o . o. G SI HIW HIE C r^ 0.8 0.6 0.4 XI & . Cl ^ 0.2^ Mn Zn 0.6 1 0 0.2 5 0 I SI HIW HIE ■M n Zn ■C r SI HIWl HIE2 G 0.8 0.6 0.4; ( 0.2 0 X & & U Figure 4-32 a-c: Spatial variations o f concentrations o f trace metals Mn, Zn and Cr for stations SI, HIW, HIE and GP in the north SDB. a, b, c represent samples collected during the collection o f the second, fourth and tenth samples for SI and GP. 4.8 Summary o f Geochemical Studies The above data from samplings during June, 1999 to February, 2003 features a wide geographic coverage and different types o f samples over both dry and wet seasons. In 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. fact, the sampling strategy had evolved over time in order to achieve this goal and serve the needs o f the geochemical-hydrodynamic model to be described in the next chapter. In summary, the geochemical studies have yielded the following results; 1) Spatial distributions o f SPM and radionuclide concentrations within and around SDB were obtained. Do achieve this, 36 stations inside and around SDB were selected and SPM and concentrations o f radionuclides measured, with many stations sampled more than once. 2) All input/output fluxes o f chemical species for mass balance calculations, including atmospheric input o f radionuclides, precipitation and riverine inputs of freshwater, SPM and radionuclides, were systematically measured during the 4- year period. 3) Activities o f radionuclides in both particulate and dissolved phases were measured in order to help with the simulation o f particle scavenging in the geochemical-hydrodynamic model. 4) Seasonal variations o f chemical species were also disclosed through the measurements o f spatial variations o f SPM and radionuclides in both dry and wet seasons. 5) Temporal variations o f SPM and radionuclides with tidal phase were measured. Two sets o f such data were obtained by sampling repeatedly at one or more stations over multiple tidal cycles. 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6) Transient effects o f precipitation events in SDB were measured by sampling during precipitation event. In a dry region like SDB, precipitation events are important source o f SPM, and might have significance to the long-term variations and mass balances o f many other chemical species as well. Aside from above, trace metals, TOC, TON as well as PCBs were also measured along with SPM and radionuclides in order to relate the radionuclides as chemical tracers with environmentally important chemicals such that the environmental significance o f our study can be addressed. Finally, when it became increasingly evident that hydrodynamic modeling will be an important part o f this study, the need o f boundary conditions as well as calibration standards for chemical species in the model was addressed. In fact, samplings o f outer sea, rivers, as well as samplings over tidal cycles were planned at least partly for this purpose. As will be discussed in chapters 5 and 6, geochemical studies as outlined by this chapter are an indispensable part o f the geochemical-hydrodynamic model. 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 5 Integrated Geochemical and Hydrodynamic Modeling of San Diego Bay 5.1 Introduction The geochemical studies described in chapter 4 revealed the spatial patterns and '> '1 A 91 0 9 1 0 temporal variations o f the radionuclides ( Th, Pb, and Po) in sediments and water columns. With these data, it becomes possible to quantitatively describe the dynamic behavior o f these radionuclides by a numerical model. Since these radionuclides and SPM exist in the forms o f chemical species in water, the numerical model should be based on the hydrodynamics o f the aquatic system. The numerical model should include governing equations o f hydrodynamics, suitable boundary conditions (BCs) and initial conditions (ICs) imposed on the governing equations. The numerical model o f the system provides a mathematical description, or a c la s s ic a l o r s tr o n g f o r m o f a problem or a model (Kaliakin, 2002). If the governing equations are set up correctly, and BCs and ICs are properly defined, the governing equations can be solved numerically. The numerical solutions provide simulations of the behavior o f the system so that the following can be achieved; 1) the model can be calibrated and verified against the known observational values (excluding BCs and ICs) to ensure that it works satisfactorily; 2) the model results, after calibration and verification, can be used to estimate the parameters at regions not covered by measurements (so-called ‘nowcast’). W ithout numerical simulations, the only ways to fill the gap in the measurements are either making more measurements or using 110 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. interpolation, but the latter is often inaccurate for a dynamic system; 3) the model can predict the behavior o f the system (‘forecast’). Another advantage o f numerical modeling is that a successful model ean be applied to other similar environments with moderate modifications in BCs and bathymetric information. Thus, numerical modeling is a powerful research tool that expands the capabilities o f geochemical studies. Compared to a hydrodynamic model that assumes no steady states, box models, as simple as they are, ean help establish mass balances for conservative and non conservative chemical, species. Our geochemical studies enable calculations of reservoir sizes as well as fluxes in and out o f SDB for water, salt, SPM and radionuelides, thus box models can be established for these species, as described in chapter 6. However, due to large transient variations for chemical species in a short time scales during tidal cycles, steady-state box models have severe limitations. For an energetic, tidally-dominated environment like SDB, a hydrodynamic model based on explicit physical laws is more desirable because only such a dynamic model is able to describe quantitatively the transient variations and flux o f water and ehemieal species within the system and across the boundaries. As mentioned in chapter 1, there are no known studies o f the hydrodynamic modeling o f naturally-occurring, particle-reactive radionuclides in coastal environments despite 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the extensive geochemical studies o f these radionuclides in similar environments and fairly well-developed capabilities o f many hydrodynamic models, many o f which were successfully integrated with studies o f conservative chemical species, especially those with environmental significance (Bormer et al, 1994; Chadwick et al, 1999; Dortch et al, 1998). This is a glaring gap in the field o f coastal biogeochemistry. The reason for this is because o f the complexity o f the problem, which consists o f the biogeochemistry o f the radionuclides, hydrodynamics, numerical modeling and particle dynamics. O f these issues, particle dynamics is a crucial part that relates the behavior o f particle-reactive radionuclides with hydrodynamics. However, particle dynamics itself has many unsolved issues, making it difficult to tackle the problems o f particle-reactive species with hydrodynamic models. This will be the key issues addressed in this chapter. The central piece o f the puzzle is the hydrodynamic model, upon which the geochemical and particle dynamic submodels will be based. Obviously it is impractical to re-invent the wheel by constructing a new model from scratch, in view o f a wide range o f hydrodynamic models available. Rather, an established model will be adapted to serve our needs. The search for a hydrodynamic model was based on the following criteria: 1. The model should be well-established, calibrated and verified to ensure correctness; 2. the computer code o f the model should be accessible in order to add chemical species for this study; 3. the model should be able to run on 112 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. computer systems that are readily accessible, preferably on personal eomputers; 4. the model should be well documented in order to be studied. As a result, commereial software packages are precluded due to their high costs and the inaccessibility o f their programming codes, which are needed to serve our specific needs. The initial plan to re-invent the wheels by establishing a new hydrodynamic model from scratch was given up due to the requirement o f enormous amount o f time for design, calibration and verification. The search for the model eventually came to the TRIM model that was designed by Dr. Ralph Cheng o f USGS. With the kindly assistance from Dr. Cheng, the computer codes o f the model, together with supporting files (e.g. bathymetry, tidal harmonics for SDB) were transferred to me with the help of personnel from US Navy and San Diego Supercomputer Center (SDSC). 5.2 Description o f TRIM model TRIM stands for Tidal, Residual, Intertidal Mudflat model. It is a hydrodynamic model that has established itself as a benehmark model for the simulation o f shallow- water eireulation driven by tides and salinity buoyancy (i.e. residual currents). It has been widely used in many shallow water environments, including SDB (Cheng and Casulli, 1993; McDonald and Cheng, 1994; Wang et al, 1998). The model uses a semi-implicit, time-stepping fmite-difference method for the nonlinear, depth averaged (2-D) shallow water equations (Casulli, 1990 a, b). It takes into consideration Coriolis acceleration, baroclinic forcing, and wetting and drying of 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. computational cells. An Eulerian-Lagrangian method (ELM) was used to treat advection terms in the equation such that the numerical method is unconditionally stable if the baroclinic forcing is neglected. Even with baroclinic forcing considered, the solution is subject to only a weak stability condition, which was often met in natural estuarine environments where the model had applied. Despite being a depth- averaged model, it is capable o f simulating the hydrodynamics o f shallow water accurately and efficiently (Cheng and Casulli, 1983). As a result, the TRIM model is recognized as a benchmark model for 2-D hydrodynamic simulations (Martin and McCutcheon, 1999). Detailed description o f the model and the model verification was done in its application on San Francisco Bay (Cheng and Casulli, 1983) and San Diego Bay (Wang et al, 1998), so only a brief introduction o f the original model is given below. For a detailed description o f the numerical modeling and finite differenee methods, the reader is referred to Kaliakin (2002). 5.2.1 G overning E quations The governing equations for the TRIM model for flows and transport o f scalar variables in a tidal system include the conservation equations o f mass, momentum, conservative scalar variables, and an equation o f state. This scheme is essentially the same for almost all hydrodynamic models, which differs only on the number of dimensions, simplifying assumptions about density and isotropy o f water, and whether or not some parameters o f secondary significance are considered in the 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. momentum equations (e.g. Coriolis acceleration, baroclinic situation, wind stress, to name a few). For a detailed derivation o f these equations, please refer to M artin and M cCutcheon (1999). In fact, what distinguish one numerical model from another is primarily the numerical schemes, i.e. how these equations are solved numerically. The numerical scheme will be discussed in the next section. The governing hydrodynamic equations for the original TRIM model are listed below (Cheng and Casulli, 1993). Continuity equation (i.e. conservation o f mass): a ; _ g[(h+QU] a[(h+QV] a ax ay The x-momentum equation (conservation o f momentum at x-direction): W f C '+ o i (5.2) The y-momentum equation (conservation o f momentum at y-direction): )+A„vV - | ; : ( h + o | (5-3) ap ap . . If ^ and ^ are not zero m equations 5-2 and 5-3, i.e. in the baroclinic condition, an additional salt transport equation is needed: D? = ^ V [K ,(h + O V p ] (5-4) 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Within the range o f 10-15 '’C it was found that there is a simple density-salinity relationship (Cheng and Casulli, 1993): p=p„[l+(0.78s-0.75)/1000] (5-5) where in equations 5-1 to 5-5, (x,y,z) are the Cartesian coordinates; (U,V) are the depth averaged velocity components i p is the depth averaged water density Po is a reference density o f water ^ .d .8 . . V = i^ IS a vector operator in x-y plane f is the Coriolis parameter C , is the free-surface above a reference horizontal plane (MLLW) h(x,y) is the instantaneous water depth including C , g is the gravitational aeceleration Ah and K h are the horizontal eddy viscosity and diffusivity coefficients (tj^, ty ) are the specified wind stress; t^ , ty by are the bottom stresses 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Y) d d d . Dt ^ ^ + U ^ +V ^ is the substantial derivative in the x-y plane, which is essentially the time rate o f change (TROC) o f a property following a fluid particle (i.e. a Lagrangian concept) in an Eulerian coordinate system. In momentum equations 5-2 and 5-3, the terms besides the substantial derivatives of U and V (the first term on the left), represent the force due to Coriolis effect, barotropic effect (difference in water surface elevation), wind and bottom shear stress, viscosity, and baroclinic condition. In most estuaries, the ratio o f vertical length scale to horizontal length scale is very small, and the horizontal mixing terms are o f smaller order o f magnitude compared with the remaining terms in the governing equations. For practical purposes, the thickness o f the lateral boundary layer is thin and can be neglected. Thus the horizontal eddy viscosity is set to zero and the eddy diffusivity coefficient is treated as constant. This issue is in fact much more complicated than it seems when eddy viscosity and diffusivity are considered in the hydrodynamic models. In some cases, the correct treatment o f them determines the success or failure o f the entire model. The viscosity is relatively constant at about 10'^ m^/s. In contrast, diffusivity (including both molecular diffusivity and eddy diffusivity) varies with the hydraulic condition. The eddy diffusivity for most surface waters is in a range o f 10'*-10^m^/s, with an average value o f 10®m^/s; molecular diffusivity is fairly constant at 10'*m^/s, 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. thus it is negligible. Thus the treatment o f TRIM does not introduce significant errors into the system. In the governing equations, equations 5-1 to 5-4, the water is assumed to be incompressible. The pressure distribution in the vertical is assumed to be hydrostatic, so that the barotropic pressure gradients have been replaced by the gradients o f water free-surface. The velocity distribution in the vertical is assumed to be the nearly uniform. The salinity is assumed to be well mixed in the vertical water column, while the longitudinal salinity gradients (baroclinic pressure gradients) are included in the formulation. The Boussinesq approximation, which assumes the density to be a constant except in the baroclinic forcing terms is used. The basis o f this approximation is that the temperature and density variations in the water body are so small that they are negligible, yet the buoyancy drives the motion. Thus the variation in density is neglected everywhere except in the buoyancy term (i.e. the last terms in equations 5-2 and 5-3). The bottom stress terms are related to frictional dissipation o f momentum at the sediment-water interface. In depth-averaged models, the bottom stresses are often approximated by a form o f quadratic drag law (Dronkers, 1969). In the presence of Coriolis acceleration, the bottom stress deviates from the direction o f depth-averaged velocity (Moljeld, 1988). However, in relatively shallow water (with depth less than 20 meters), this deviation is estimated to be less than 10®(MoijeId, 1988). For lacking o f three-dimensional velocity properties to define the bottom stress and for shallow 118 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. water applications, the conventional Manning-Chezy type o f formulation for the bottom stress is used in this model as described in the following equations (Sabersky etal, 1999): 1 u gV u^+v^ u ^ --------- — =yU (5-6a) p„(h+Q C(h+(!^) 1 u gV uV v^ V —7rTTTr^ = ^ 3 =yV (5-6b) p„(h+Q > ' C(h+(^) Cz is the Chezy coefficient, which is related to a M anning’s constant n in the following equation: (h+C)‘^ ^ C z = ^ ^ - ^ (5-7) With appropriate boundary and initial conditions, the system o f equations (5-1) -(5- 4), and equations (5-5) and (5-6), constitute a well-formed initial-boundary value problem whose solution describes the depth-averaged circulation in a tidal basin. 5.2.2 T he N um erical Schem e Once the governing partial differential equations are derived and the boundary conditions and initial equations prescribed, the mathematical statement o f a given physical problem is complete. Then, the governing equations must next be solved. Unfortunately, there is no analytical solutions to the equations 5-1 to 5-4 due to their 119 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. complexity, and the solution to these equations can only be obtained through approximation procedures, or numerical schemes. Naturally these approximation brings errors, defined as the difference between the actual response (e.g. measured experimentally) and that predicted by the approximate solution. Although such errors are unavoidable, it is imperative that they be assessed and minimized in order to economically obtain solutions within a given accuracy (Kaliakin, 2002). The numerical method that TRIM model adopted is the finite difference method, which is one o f the most widely used discrete approximate solution technique (another one being finite element method). This method essentially generates approximations only at a speeifie number o f locations within the model domain. In the approximations, the partial derivatives are replaced by suitable difference formulas such that partial differential equations 5-1 to 5-4 are converted into simple algebraic equations, which can be solved through matrix manipulation. Boundary conditions and sometimes initial conditions are needed in the solution. To do so, the first step is to divide the domain into a set o f discrete grid points. It is at these grid points that the approximate values o f the primary dependent variables will be solved for. For TRIM model, the domain (which can be visualized to be a region including San Diego Bay and adjacent open ocean) is discretized as follows: 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ax Figure 5-1. Computation mesh o f the TRIM model. Water elevation ^ij and concentration o f chemical species Sij are defined at the center; velocities U and V are defined on the boundaries. The computational mesh consists o f rectangular cells o f Ax an Ay, as shown in figure 5-1. Each cell is indexed at its center with and index (i, j). However, the discrete U velocity is defined at (i+1/2, j), and V velocity is defined at i,j+l/2. All other scalar variables, including the water elevation water density p, and salinity are defined at (i, j). The basin bathymetry is defined at U and V points. This computational mesh gives the best representation o f the basic bathymetry on a fmite-difference grid-mesh, which results in a 30% higher resolution o f the basin stencil. This computational stencil gives the best representation o f the basin bathymetry compared to defining water depth at the comers o f the computational stencil. Typically, if a variable is not defined at a spatial location, an arithmetic average o f the surrounding values is used (Cheng and Casulli, 1993; Wang et al, 1998). 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Another important issue in the numerical scheme is the choice o f explicit or implicit method when the variables are discretized. The explicit method, simply put, is one that determines the value o f the unknowns at the current time step from known values calculated from the previous step Thus the computing is simple and efficient. However, the numerical artifacts can be severe in an advection-dominant environment like San Diego Bay (Cheng and Casulli, 1993; M artin and McCutcheon, 1999; Kaliakin, 2002). In contrast, the implicit method involves more than one unknown for the future time step, so a series o f coupled equations are produced, which can only be solved by matrix or iterative methods. As a result, implicit scheme often involves more computation. But the implicit method has the advantage o f higher stability, so it is often used in the advection-dominant environments. On the basis of the characteristic analysis o f the shallow-water equations o f Casulli (1990 a, b), the gradients o f free-surface elevation in the momentum equations, equations (5-2 and 5-3), and the velocity divergence in the continuity equation (equation 5-1), must be discretized implicitly to order to achieve unconditional stability. Casulli’s (1990 a, b) semi-implicit fmite-difference algorithm is adopted and extended in this model. The continuity equation is treated in its conservative form in which U and V are discretized implicitly, while the total water depth (h +n) is discretized explicitly. The bottom stress terms are evaluated at a split-time level with 122 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. g in Equation 5-6 evaluated explicitly, and the velocity components evaluated implicitly. All remaining terms in equations 5-1 to 5-3 are evaluated explieitly. Finally, the transport equation is solved by an explicit ELM. Following the convention o f using superscript for time and subscript for location indices, the semi-implicit, finite difference method can be applied to equations 5-1 to 5-4 to give the continuity equation: «u' = ^ u -i ; W / 2 J u-;,j i 1 (s-s) the x-momentum equation u - g ^ K >-At(T « r,k rw x w ,2 .j) (5-9) where W x ",.^ - [ FV+ } , - x E ( h + C ) ^ P a i/t ‘+1/2.J L p „ ( h + Q a: 2 p „ '^ -I1 + I /2 .J and the y-momentum equation -e ^ ( t > - 4t(Ty,„ v w yjj.,,,) (s-io) where W y ; ' „ „ = [ F V + ^ ^ / - ^ ( h + o | ^ fj+ ,/2 and the transport equation. 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. P u ' =Fp"i + A t [ ^ V [K ,(h+ gV p] (5-11) The finite difference expressions for Wx and Wy can be written explicitly, and they are not spelled out here. The horizontal eddy viscosity terms in the momentum equations have been neglected (Ah=0). The explicit treatment o f the baroclinic forcing has a weak coupling to the momentum equations. Equation 5-11 is solved independently by an ELM (Cheng et al, 1984), and must satisfy the following necessary and sufficient condition for stability: In practice, explicit treatment o f equation 5-11 does not affect the stability o f the semi-implicit fmite-difference scheme because in convection dominated problems, inequality 5-12 is much less restrictive than the Courant-Friederichs-Lewy (CEL) condition (Cheng and Casulli, 1993). In the special case when Kh is zero, the method is unconditionally stable. The notation F is an explicit, nonlinear finite-difference g operator, corresponding to the spatial discretization o f the substantial derivative ^ g g + U ^ ^ specific form for F directly affects the stability o f the method; and ELM is used to define F whose details are given next. 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Bring equations 5-9,5-10 into equation 8 and simplify for water elevation (^, we have the following finite difference equation: n+1 .n + 1 „ n+1 . n+1 1 At^ [ r ^ n+1 n+1 i-,|-l/2 '+‘^^ r \ 1 =• r '^ 1 ^ i.j ‘ ’ i.j+ P + i , n . t ^ i.j S i .j .i ) J - S .j l+y^ At l+v^ At + ^ ‘'^ ’‘1.1/20 ■ ^ Y i + l / 2 , i ' ^ ‘- ^ 'i+ l/2 .j'^ ^ A t r C i.j+ i/2 + h i.j.i/2 C i.j + i/2 + h j .j + /2 'n g .„ 2 A t ^ " W y , A tW y ^ ,,,)] (5-13) Obviously there should be i x j equations taking the form of 5-13. Conversely there are i X j unknowns (i.e. < 1 ^ " ^ ,’ to ) thus they can be solved. To do this, equation 5- 13 is rearranged in the following matrix form: A-C"^‘ = B (5-13’) 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where all three terms are in matrix form. A is an (ixj) by (ixj) matrix, and B are both (ixj) by 1 matrices. This equation constitutes a linear 5-diagonal matrix o f equations for (^. In other words, there are 5 ‘unknown term s’ or those with The parameter matrix A o f equation 5-13’ is symmetric and strictly diagonally dominant with positive elements on the main diagonal and non-positive elements off the diagonals. Through matrix transformations, equation 5-13’ can be solved to compute H “ t " 1 the unknown matrix, ^ (‘future’ water elevation). Since all terms in matrix A are non-negative, the system 5-13 has a unique solution and can be solved efficiently by a preconditioned conjugate gradient method (Cheng and Casulli, 1993). 5.2.3 ELM for treating the convective terms The nonlinear advective terms are often the source o f difficulties in numerical solutions since they give rise to nonlinear instabilities. The nonlinear terms are also the production terms for residual currents. Most tidal hydrodynamics processes are strongly convection-dominant. Therefore, a Lagrangian treatment o f the advection terms is favored. For logistical simplicity, this Lagrangian treatment is superimposed upon an Eulerian computational mesh resulting in an effective Eulerian-Lagrangian method (ELM) for convection dominant problems (Cheng, 1983; Casulli, 1990 a, b). 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Based on the results o f a von Neumann stability analysis, Casulli (1990 a, b) examined the properties o f the difference operator F which determines the stability properties o f the semi-implicit method. In this model, an ELM representation o f F is used beeause the resulting difference operator gives a relatively accurate solution, and is unconditionally stable if the dispersion in the salt transport equation is neglected. To implement the ELM for the convective terms, the substantial derivative o f a general variable w is written as Where D/Dt represents the time rate o f change (TROC) along the streamline, which is defined by * = U a n d f = V (5-15) By definition, grid courant numbers are defined as: At At a = U - j;,a n d b = V ^ So the substantial derivative can be expressed as n+1 „ n Dw Wjj -FWjj n n where Fwij =Wi.a.j.b (5-16) 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. i- lj bAv Streamline i-a,j-b aAx Figure 5-2. The Eulerian-Lagrangian Scheme for Convection The scheme is shown in figure 5-2. Since point (i-a, j-b) is not necessary on a grid point (p and q are its distance to vertical and horizontal boundaries, respectively). To make the method more accurate, an interpolation is needed to estimate the location o f this point by 4 grid points surrounding it by the following bilinear interpolation: F w"j =(l-p)[(l-q)w "nj.m + qw"„j.n,.i ]+p[(l-q)w "n.i j.m + qw"n.l.j-m-l ] (5-17) Since U and V are not constant, a and b must be obtained from integrating the ordinary differential equation 5-15. If the grid courant numbers are estimated to be greater than 1, then the time integration along the streamline can be divided into N equal sub-intervals At’=At/N. The value o f N is chosen such that At’|U"| or At|V"| are 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. always less than 1. This condition ensures consistency, so that the streamline swill never cross a solid boundary (Casulli, 1990 a,b). The bottom friction terms in equations 5-9 and 5-10 are calculated by the following: i+l/2-a.i-b- I L i+l/2-a.j-b- Y.. ,,0 =— r— (5 -isa) n g^[U"a,.,/2-b]'+[VL.i.l/2-b]' Ti u ,/ =-------- 2 --------------ii— ---------- ( 5 - 18b) Since all U and V terms in equations 5-18a,b are calculated in equations 5-9 and 5-10, no additional calculation is needed. Finally, both boundary and initial conditions are provided by the tidal functions: ^ (t)= C o + X { FiAiCos[0jt-^i] } (5-19) ; = 1 where (!^(t) is the tidal height above the reference water depth Subscript i labels different harmonic constituents (16 are considered in TRIM model); Fi is the nodal factor reciprocal, which is a slow varying function with a period of 18.6 years; 129 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ai is the mean tidal amplitude; o o i is the tidal period; ( |) is the phase lag, which is independent with time and can be calculated by ( j) ; =Kj-Ei. Here k; is the local epoch, Ej is the local equilibrium argument (Cheng and Casulli, 1993). O f the above, A and k are known as the harmonic constants. Therefore, once the harmonic constants are determined, the water level can be calculated for other times using equation 5-19. The harmonic constants can be obtained through harmonic analysis by best-fitting the observational data with equation 5-19. The result o f harmonic analyses for San Diego Bay is listed in table 5-1 below (Cheng and Casulli, 1993). 5.3 Modifications o f the Model Parameters Without major changes in the hydrodynamic portion o f the original model, the grid size was changed in order to achieve higher efficiency so that a simulation o f years of real time can be done in a reasonable amount o f time. The original model (Cheng and Casulli, 1993; Wang et al, 1998) was designed to be executed on high-end computers or super computers, which are not readily available for this study, so it is necessary to sacrifice spatial resolution to accommodate higher efficiency. A FORTRAN program was designed to read the original GIS file o f 50mx50m grid into 150mx200m grid. As 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 5-1 Harmonic constants for the 16 major harmonic constituents in San Diego Bay Constituent i A © E K F Q 1 3.924 0.234 306.283 74.042 0.953 oi 21.475 0.243 13.828 79.781 0.953 M l 0.757 0.253 192.168 125.137 0.702 PI 10.632 0.261 128.087 82.709 1.000 K1 33.578 0.263 223.313 86.485 0.971 J1 1.986 0.272 286.959 91.835 0.951 MU2 1.561 0.488 108.481 226.175 1.005 N2 11.828 0.496 165.696 252.626 1.005 NU2 2.286 0.498 176.126 255.673 1.005 M2 49.895 0.506 233.239 269.924 1.005 L2 1.392 0.515 132.986 286.791 0.893 T2 1.159 0.523 140.787 241.967 1.000 S2 20.350 0.524 0.000 256.752 1.000 K2 5.879 0.525 266.127 251.990 0.947 M4 0.023 1.012 106.479 75.065 1.010 MK3 0.014 0.768 96.553 115.460 0.976 a result, the number o f cells decreased from 308x400 to 77x134. The grid size increased by 12 times, and a calibration run showed that the actual increase of efficiency was about 45 times due to a combined effect o f decrease o f spatial 131 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resolution and simplified computation. Eventually the model was able to simulate one year o f real time in less than one day in a Pentium III 500 MHz personal computer. The effect o f a decreased special resolution is shown in figure 5-3a and 5-3b. It can be seen that the low resolution map (figure 5-4b) maintains most o f the geographical features o f the high resolution one (figure 5-4a). As a result, a correct hydrodynamic simulation is assured for the new scheme. E ast-; Figure 5-3 a: Low resolution computational grid used in the new model; b: high resolution computational grid used in the original TRIM model. Contour lines at 1, 5, 1 0 ,2 0 ,4 0 ,6 0 ,8 0 ,1 0 0 m . As stated above, a number o f modules for geochemical modeling were added onto the original hydrodynamic model, and in each step more calculation is needed. However, after comparison it was found that the addition o f geochemical modeling modules did not affect the efficiency o f the overall simulation appreciably. The main reason for this is because the most computing-intensive part o f the model is the calculation of 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. hydrodynamic parameters, which generally involves solution o f large 134x134 Jacobian matrices for over 20 times in each step because o f the non-linear terms in the ELM scheme (please refer to the earlier portion o f this chapter). In comparison, the chemical species can be computed conveniently at the end o f each step after the hydrodynamic parameters, including all the non-linear terms, are computed. 1200 1000 ■ B 800 600 i E 400 200 0.1 0.2 Time Step 0.3 Figure 5-4. Calibration result showing the relationship between the length o f time step in simulation (horizontal axis) with the ratio o f real-time (RT) vs. simulation time (ST). The highest efficieney is achieved when time step is 0.08 hour. 0.05 (i.e. 3 minutes) is used for simplicity o f calculation The selection o f the optimal length o f time step was carried out by calibration runs. As shown in figure 5-5, the computing efficiency increases with longer time steps but reaches a maximum value at around O.OShour (i.e. 4.8minutes). The efficiency then decreases and levels off with increasingly long time steps. However, when the time step exceeds 0.25 hours (15 minutes), CFL conditions are usually not satisfied (see previous portion o f this chapter for reference) and the model becomes unstable. So 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.08 hour seems to be the optimum length o f time step for simulation. However, 0.05 hour was eventually used for simplicity. From figure 5-5 it is seen that there is only small difference between the computing efficiencies between these two choices of time steps. Additionally, the use o f smaller time step makes grid courant number smaller, which again makes the model more stable. 5.4 Schemes o f Geochemical Modeling o f Non-Conservative Radionuclides The above hydrodynamic model provides a platform for a geochemical modeling of particle-reactive radionuclides for our study. Based on hydrodynamic computations, geochemical modeling o f non-conservative chemicals, including particulate matter as well as radioisotopes can be established by adding computations o f non-conservative terms describing the sources and sinks o f these species within each time step. Each o f these terms should be based on thermodynamic or chemical laws. As subsequent sections indicate, the original TRIM model has been extensively modified in order to accommodate the simulation o f radionuclides and particle dynamics. The added modules significantly enhanced the functionalities o f the original TRIM model, which was designed largely for tidal hydrodynamic forecast and nowcast for practical and research needs (Cheng and Casulli, 1993). The new model will be the first coupled geochemieal-hydrodynamic model in which the 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. geochemistry o f partiele-reactive radionuclides is driven by the hydrodynamics o f the oceanic system, with particle dynamic model acting as the connection between the geochemical and hydrodynamic models. To distinguish this new model with the original TRIM model, it is termed as “RS-TRIM ”, with ‘R ’ representing radionuclides, and ‘S’ representing sediments or particles. 5.4.1 Governing Equations for Geochemical Modeling Time rate o f change (TROC) for each chemical species constitutes an independent equation consisting o f terms deseribing diffusion, advection, decay, particle scavenging, river runoff, atmospheric fallout, and in-situ production from parent radionuclides. Since the computing-intensive part o f these equations (i.e. diffusion and advection) has already been calculated in the hydrodynamic modules, the addition o f these species to the model does not affect the efficiency o f the model. Goveming equations for dissolved concentrations o f radioisotopes are listed below in equations 5-19 to 5-21: DC™ 6^C™ 5^C™ F o r ^ 3 4 T ^ h : ^ ^ K h ( ^ + ^ ) - X ™ C ™ -k™+P™ (5-19) DC'i'’ 5^C f T 7 210^1 . _________________ ^ I ^ \ Pb^Pb i.Pb I A rb I T^Pb / - c For Pb: - Kh( + g^2 ) -A C^ - k +A + R^ (5-20) 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. F or2‘% : - 5 ^ D s ( - ^ + - ^ ) - ^ " ° C ; ° -k'’° + + Apo+r J° (5-21) Correspondingly, the governing equations for concentration (dpm/1) o f particulate radioisotopes are listed below (equations 5-22 to 5-24): F n r P V ' ( -------- ^ ? r - 4 - ----------- § — ^ T h p l h i u T h , _ 9 . , | s _ L , i c p T h For Th. -K h ( g^2 + gy2 y k Cp +k + g^ q C p (5-22) DC’ ’’ ’ 5a 1 F r > t - ^ ^ 7 P b ^ P b , i P b , _ 9 . ^ A s | ! p P b / c For Pb. -K h( g^2 + gy2 )-^ Cp +k +g^ q Cp (5-23) g2^Po g2^Po For ^'"Po: K » < - ^ + +k'’ « (5-24) Where DC is the substantial derivative defined previously Kh is the horizontal diffusivity described before C is the concentration o f the radionuclides Superscripts o f Th, Pb and Po indicate the species, and subscripts o f d or p indicate dissolved or particulate phases X is the decay constant k is the scavenging constant. It is expressed as such here for simplicity. A detailed discussion o f scavenging scheme is at the subsequent sections o f this chapter. A and R represent atmospheric and riverine input terms q is the amount o f SPM. 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In the above equations 5-19 to 5-24, the left side o f equation and the first term at the right side o f each equation represent advection and diffusion terms for all chemical species implying that they behave in the same way as conservative tracers in terms of advection and diffusion. These terms are essentially similar to equation 5-4, the salt transport equation, with the exception that the concentrations o f these chemical species are too low to affect the density, as salinity does. The other terms in these equations are calculated at the end o f each step after the advection and diffusion terms are calculated. The differences between equations 5-4 and the above equations (equations 5-19 to 5-24) lie in the subsequent source and sink terms (non-conservative § 1 terms) in equations 5-19 to 5-24. Among them, the term calculates the amount o f increase or decrease o f particulate phase concentration due to particle settling or sediment resuspension. To be more precise, C here should aetually be Cp only in case o f settling. In the event o f sediment resuspension, C should be the activity o f the surface sediment for a particular radionuclide. Here they are not distinguished for clarity, but in the actual model they are treated differently. Also not shown in the above equations are the variations o f the concentrations o f both particulate and dissolved phase radionuclides due to precipitation, river input, and evaporation. These processes are dealt with in the model based on the mass balance calculations described in chapter 6. 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.4.2 Q uantification o f N on-C onservative T erm s for R adionuelides It is clear that what makes this model different from the original hydrodynamic model are the non-conservative terms in equations 5-19 to 5-24. Among them, the radioactive decay and production from parent radionuclides are relatively easy to define and they are not spelled out here except for ^^^Th, which is modeled as being directly related to salinity according to the following U-salinity relation (Ku et al, 1977). Since the water in SDB and vicinity is well mixed, the salinity is thus largely constant and varies only at a small area close to freshwater input during major storm events (refer to figure 4-31). The salinity range (33.5 to 35 per mil) o f San Diego Bay represents an error in ^^"^Th estimation much smaller than the analytical error (-10- 20%) and thus is not a concern. production from ^^^Ra decay in the water column is not considered due to its negligible contribution to the whole inventory of Tin Pb within the bay, as in other similar environments (Appleby and Oldfield, 1992). For physical input and output terms such as river input, precipitation, evaporation and atmospheric input, the -modeling is relatively straightforward as long as these fluxes are defined by previous geochemical studies. They will be discussed in chapter 6 and will not be detailed here. The remaining terms in equations 5-19 to 5-24 are sediment resuspension/settling 6 q (i.e."^ in those equations), and particle scavenging (represented by the parameter k in the equations). It should be noted that sediment sources such as man-made sediment 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. resuspension and riverine input o f sediment are external discrete processes that affect only a few model grids directly. Their randomness precludes them from equations 5- 19 to 5-24, and they will be treated as stochastic processes described below. Since the radiotracers used in this study are all particle-reactive, their spatial and temporal variations are all dictated by particle dynamics and the related particle-radionuclide interactions. Therefore, sediment dynamics and particle scavenging o f radionuclides are crucial parts o f the new model and will be treated carefully. 5.5 Particle Dynamics Modeling The strength o f particle dynamics in SDB is shown in the aerial photo (figure 5-5). Consistent with our observations (chapter 4), SPM concentrations seem to be higher at north SDB and the south end o f SDB. In view o f the importance o f particulate matter on the geochemistry o f the radionuclides in our study, it is necessary to simulate the particle dynamic carefully. To treat particle dynamics mathematically, the governing equation takes a relatively simple form as shown below: Dq 8^q 8^q D t“ ° " < 8 x ''" 6 y ' where q is the concentration o f SPM 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure 5-5 Aerial photo o f San Diego Bay, which shows particle plumes (light- colored) at south bay, bay mouth, as well as parts o f inner bay. Source: www.swrcb.ca.gov. The two arrows were drawn in the original image for other purposes. s represents the amount o f sediment resuspension during time interval At Sd represents the amount o f sediment settling during time interval At Equation 5-25 indicates that particulate matter is simulated as a non-conservative chemical species that participates in adveetion and diffusion. Non-conservative processes include resuspension and settling, which can be estimated using semi- empirical schemes. Other non-conservative processes such as shipping-induced resuspension and river input are o f stochastic nature and have to be modeled as such, they are not in the numerieal model. Finally, biological resuspension and atmospheric fallout are not considered to be significant sources o f particles so they are not 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. considered. In the subsequent sections, storm runoff, shipping-induced resuspension, tidal resuspension and settling will be discussed in detail. 5.5.1 Stochastic Processes as Sources o f Suspended Sedim ents Storm runoffs are potentially important sources o f particulate matter to SDB. San Diego region has a typical Mediterranean climate with annual precipitation o f about 25em (Chadwick et al, 1999). A probability model was constructed that allows 2 major rainfalls (5cm), 5 average ones (2em eaeh), and 5 minor ones (1cm each) for each calendar year. Random-numbers were generated in each step to be converted to probabilities o f eaeh event, during which fresh water loaded with particulate matter is input from 4 major rivers during a time period o f 12 hours. According to Chadwick (1999), 120mg/L is used as the average value for SPM concentration in all rivers during storm events. The value is used with an understanding that there are large variations related to the duration o f rainfall, the time o f measurement, and recent precipitation history. The exact amounts o f freshwater and SPM from rivers are calculated in Appendix 1 1. It was found (Wang, personal communication) that shipping activities, especially those related to large vessels, were the most important sources o f particulate matters into SDB. Therefore, it is crucial to simulate the production o f SPM by shipping activities in the model. O f the over 9000 vessels harbored in SDB (Chadwick et al. 141 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 1999), only a handful o f large vessels have significant impact on the particulate matter production in SDB. They include Navy cruisers, aircraft carriers, cruise liners that harbor at Cruise Terminals near the Broadway Pier, and about 50 large commercial (>10,000 ton total weight) vessels that call SDB harbor every year. According to Chadwick et aTs (1999) coordinated monitoring during 1995 to 1996, there were 4.76±0.5 ship movements per day, which produced 41.7±22.3 xlO^ Kg of sediment per day. Since this calculation only covered Naval Stations located in Central SDB, a slight different set o f data was used in the model that considered all Naval cruiser docks throughout SDB (including a few submarine docks), and each dock is considered individually in a probabilistic manner. For Cruise Liners, a two- year dataset (source: www.portofsandiego.com) showed that there is a cruise liner coming to SDB every 2.5 days. There is no direct observation data for the resuspension related to them, so we presume that each docking produces the same amount o f sediment as that for Navy cruise. For aircraft carriers, the movements in and out o f SDB are far less frequent, but dredging activities are frequent at the Navy Aircraft Carrier Base (figure 5-6a); besides, each docking o f aircraft carriers creates far more suspended materials than cruisers or cruise liner does (figure 5-6b). In the model, aircraft carrier base was modeled as producing suspended matter in an amount equivalent to that from two Naval cruisers. 142 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.5.2 S edim ent R esuspension D ue to T idal C urrents It is well established that (Krone, 1962; Wang, 2002; van Rijn, 1989; van de Kreeke, 1988; Torfs et al, 2001) the resuspension o f surface sediments due to current is related to the shear stress exerted on the sediment-water interface that overcomes gravity and Figure 5-6. Sediment resuspension aroimd the Aircraft Carrier Base in SDB. a: sediment resuspension due to frequent dredging; b: sediment resuspension due to docking o f aircraft carrier. attraction from underlying sediments. In m ost cases, a critical shear stress is defined, which is a threshold o f shear stress that resuspension only takes place when bottom shear stress exceeds it. Unlike bottom shear stress, which can be calculate relatively straightforwardly from current velocity and bottom roughness, critical shear stress is often designated empirically to fit observational data. The same approach was used in this study, where critical shear stress is set to be 0.5 N/m'^ throughout SDB. Bottom shear stress is calculated by the following equation (Sabersky et al, 1999) if metric units are used and SDB is treated as an open channel: 143 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2 r * = 1 9 .6 1 * ^ 7 1 * (5-26) where Tb is the bottom shear stress in n W n is the bottom roughness coefficient, which is 0.025 based on the characteristics of SDB (Sabersky et al, 1999) R is the ratio o f cross-sectional area to the wetted perimeter of the open channel V is the velocity o f currents in m/s This scheme produced identical results with those based on M cDonald and Cheng’s (1994) scheme, which used a different empirical approach. At the same time, the result from this calculation was in an apparently reasonable range for SDB. When bottom shear stress exceeded the critical shear stress, the amount of resuspension was calculated using a simplified scheme (Wang, 2002); e = M ( - A ) (5-27) ^ c e where s is the amount of resuspended sediment in mg/L 144 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. M is an empirical constant, 10’^ is used in the model % is the bottom shear stress as calculated in equation 5-26 T e e is the critical shear stress, which is 0.5 as stated previously 5.5.3 Settling o f Suspended Particulate M aterials The last term in equation 5-25 is the loss o f SPM due to settling. The effect o f settling on the concentration o f SPM is related to three factors: settling speed, water depth, and SPM concentration itself (McDonald and Cheng, 1994). Settling speed Ws is calculated by Stokes Law: Ws=^^ 1 (5-28) Where Ws = velocity o f fall (cm sec'^), g = acceleration o f gravity (cm sec'^), d = "equivalent" radius o f particle (cm), p s = density o f particle (g cm' ), ■ 2 p„ = density o f medium (g c m ') A p = viscosity o f medium (dyne sec c m '). But simplified versions o f Stokes’ Law were often used based on empirical observations (Sternberg et al, 1999), which is used in this study: 145 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ws=0.0002D"' (5-29) Where D is the diameter o f the typical particle size in micrometer Ws is particle settling velocity in mm/s. It was shown that settling speeds calculated by equations 5-28 and 5-29 are very close, lending credibility to the scheme o f particle settling. Subsequently, decrease o f SPM concentration due to settling is calculated based upon McDonald and Cheng’s (1994) scheme: 2Ws Tb Sd=-“^ C ( 1 - — ) whenC<Cc (5-30a) n X iX Sd= - ) when C>Cc (5-30b) Where Sd is the deposition rate in kg/m /s Ws is settling speed as calculated above but in a unit o f m/s C is the sediment concentration in kg/m • * • 3 Cc is the critical concentration o f particles in kg/ m T b is bottom shear stress 146 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Id is the depositional shear stress. Deposition takes place only when % < t^. The selection o f ta is summarized in table 5-1. Table 5-2 List o f model parameters used by other researches Parameter Values Source M 8x10'^ kg/m^/s 1 lO''* kg/m^/s 2,5 T ee (N/m") 0.5 1 0.02 2 1.0 5 0.2-1.5, density-dependent 3 0.06-0.78 6 0.18-1.1 7 0.35 8 T ed (N/m^) 0.001 8 0.01 5 Ws 5x10’" * m/s 1 1x10'"^ m/s 2 4.8x10'^ m/s when C<1000mg/L 4.8x10'^C'^^ m/s when 4 1000<C<10,000mg/L 3x10-5 m/s when C<6mg/L 3 3.96xlO-6xC1.19 m/S when 5 6<C<400mg/L 8 Use equation 5-29 for 2 grain sizes 3x10'^ m/s 147 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The sources of the empirical parameters in table 5-1 are listed below: 1: Lang et al, 1989 2: Wang, 2002 3: Lumborg and Windelin, 2003 4: Normant et al, 1998 5; This study 6. Krone, 1962 7. Mehta and Partbeniades, 1975 8. Johnson and Fassardi 5.6 Scheme o f Simulation o f Particle Scavenging o f Radionuclides Radionuclides in the model are modeled dynamically according to equations 5-19 to 5-24 after the particulate matter concentration is calculated in each step based on equation 5-25. The challenge now is to simulate the scavenging o f radionuclides by SPM from dissolved forms and possibly the desorption o f radionuclides from SPM. Remineralization o f particulate matter is not considered for simplicity, but it can be treated to be equivalent to reversible scavenging in terms o f radionuclide mass balance. There are two different approaches to treat the interactions between SPM and 148 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radionuclides (or other particle reactive chemicals). In the first one, as proposed by Nozaki et al (1987) and Cochran (1992), the scavenging process was conceptualized as a first-order reaction that is related only to the concentrations o f dissolved radionuclides. In some cases (Cochran, 1992, and references therein), particles of different grain sizes were treated differently, each with unique reaction constants. In the scheme o f Clegg and Whitfield (1990), particles o f different grain sizes can decompose to release radionuclides into dissolved phase or smaller particles can aggregate to form larger particles. The second approach to the scavenging process involves partitioning coefficient Kd between particles and solution in the context o f particle surface chemistry (Honeyman et al, 1988; Turner and Millward, 2002). The introduction of Kd essentially implies an equilibrium between radionuelides o f dissolved and adsorbed phases, and that the scavenging process is a reversible process, which is defined by the equation: Kd (5-31) Where [MP] is the concentration o f metal-particle complex or that o f the sorbed metal (w/w), [M] is the dissolved-phase concentration o f the metal ion. 149 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In theoretical studies (Hummel, 1997), the binding o f metal ions by particulate matter exemplified by bumic substances were studied by a number o f models, where Kc, the ‘binding reaction constant’ are defined by the following equation: [MP] where [P] is the concentration o f un-complexed particles. It can be seen that Kc resembles the stability constant K for the metal-particle complex MP for the reaction [MP] [M]+[P] if particulate matter is treated as an ideal ligand (Hummel, 1997). However, for very low concentrations o f metal ions, no matter what models are used, the relationship between the logarithmic o f dissolved phase concentration o f metal ion M with the logarithmic o f [M] could largely be described simply by a linear relationship (Hummel, 1997): ln(Kc)=-B’Hn[M ]+A’ (5-33) where A ’ and B ’ are positive empirical constants. Similarly, it was observed that there is an inverse dependency o f distribution coefficient Kd on particle concentration (so called particle concentration effect, PCE). The processes related to particle-particle interactions, reaction kinetics, sorption 150 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. reversibility, or a eombination of several processes are proposed to be responsible for PCE (Turner and Millward, 2002). A linear-logarithmic relationship appears to define the PCE for many chemical constituents as shown below (Sanudo-W ilhelmy et al, 1996; Turner and Millward, 2002): ln(Kd)=-B*ln(SPM)+ln(A) (5-34) In equation 5-34, A and B are empirical constants. Equations 5-33 and 5-34 are essentially the result from equation 5-32 when constant concentrations o f particulate matter (for 5-33) or metal coneentrations(for 5-34) are assumed. As a result, it is reasonable to use equation 5-34 in our model to relate Kd with SPM for its simplicity and effectiveness. Based on the previous studies from which we have enough data on both SPM and "21 n "210 activities o f Pb and Po in both phases (see chapter 4), as shown in figures 4-20a- d, it is estimated that for ^*°Po, lnA=12.97, B=0.21; for ^'*’Pb, lnA=12.93, B=0.46; for ^^'^Th, lnA= 13.90, B=1.31. These values are roughly in line with other published data (Rawling et al, 1998; Turner and Millward, 2002; Cochran, 1992, and references therein) and they will be used in the model. As shown in the subsequent sections in this chapter, this scheme enables a fairly close simulation of both spatial and temporal variations of all radionuclides. 151 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 n 1 n Based on SPM concentrations calculated in each step, Kd is calculated for Pb, Po and based on equation 5-34. Then the dissolved phase concentration o f eaeh of the radionuclide is calculated based on the following equation: fc SPM (5-35) 1+Kd* ^q6 where fc is the fraction o f dissolved phase radionuclide (in dpm/L) with regard to the total activity o f the radionuclide Kd is the distribution coefficient SPM is the concentration o f SPM in the unit o f mg/L Finally, the particulate phase radionuclide activities are calculated based on equation 5-36: fp=l-fc (5-36) where fp is the fraction o f particulate phase radionuclide (in dpm/L). 152 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.7 Coupling of Sediment and Water Columns A realistic sediment column is not needed but desired if the observation data on sediment cores and sediment traps are to be compared to the modeling result. The sources and sinks o f particles in the model are modeled in self-sustained ways with no requirement for a sediment column. However, a good coupling between sediment and water column, and a subsequent simulated sediment column will provide a good calibration criteria for the model in a longer time scale. The implementation o f a sediment column is done by tracking the settling o f particles and recording the activities o f the particulate phase radionuclides, which could be either ^ * * ’Pb or ^^"^Th, or both, depending on the research needs. The amount o f settling particles can be calculated from equations 5-3Oa and b with a series o f conversions as outlined below. First of all, the decrease o f SPM concentrations, as calculated by Sd in equations 5- 30a and b, can be converted to weight-based particulate phase activity as follows: As=C*Ap/SPM*Sd*H (5-36) where As is the accumulation rate o f radionuclide (dpm/cm^) C is the conversion factor specific for the model without units 153 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ap is the modeled volume-based particulate phase radionuclide concentration (dpm/L) SPM is the concentration o f suspended particles in mg/L Sd is the settling rate o f a grid cell in mg/L; H is the water depth (m). It is well established (Broecker and Peng, 1982; Robbins, 1978) that the ideal sediment profile o f a radionuclide, assuming constant input and sedimentation rate, can be described by the equation below: A=As*e'^^^® (5-37) where A is the activity o f the radionuclide at depth d from sediment-water interface; X is the decay constant o f the radionuclide; s is the sedimentation rate. Two relevant scenarios need to be considered in the model: sediment mixing and pulsing input from major storms. In the event o f sediment mixing near the sediment surface, the profile below the mixed layer is given as below (Robbins et al, 1977): A=Am*e'^^‘ ‘-"’^ ^ ^ (5-38) where 154 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Am is the average activity of the radionuclide in the mixed layer; m is the thickness o f the mixed layer. The schemes shown in equations 5-37 and 5-38 can be simulated by the model because all the variables can be calculated or assumed by the model at each time step. Caution should be called upon when the modeling result o f sediment profile of radionuclides is compared with the observation because o f the random human- induced disturbances (such as dredging and shipping) that cannot be modeled determini stically. 5.8 Results o f Geochemical Modeling o f SPM and Particle-reactive Radionuclides ^’^Pb, ^'^Po, and The finalized computer program o f RS-TRIM model has many added modules that simulate precipitation, river input o f water and radionuclides, shipping events, sediment resuspension, scavenging, and settling. M ost of them were designed as separate subroutines for ease o f handling, and the global variables are stored in a file that is shared among all programs. The flowchart o f the program and a more detailed flowchart for the particle dynamics and geochemical part are shown in Appendix II. The computer program eodes ineluding GIS files and Matlab programs for result visualization are too large to be shown. They can be obtained from the author directly. 155 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.8.1 B ottom Shear Stress Sim ulation R esults The modeling of SPM in RS-TRlM is based on the empirical equations involving bottom shear stress (equation 5-27) and simulation o f random events such as shipping and storms, all leading to sediment resuspension. At the same time, bottom shear stress prevents particles from settling due to the turbulence-induced buoyancy, which is represented by a threshold shear stress for settling, or X d . The significance o f SPM simulation to the well-being o f the model as well as other empirical parameters in the model necessitate a series o f calibrations and sensitivity tests. In fact, these were done well before the geochemical model for radionuclides was finalized. For these calibrations and tests, the key issues are to find optimum values for critical shear stress for erosion, X c e and that for deposition, X d . Other important parameters include the erosion constant M in equation 5-27, which was set to be 10'"* mg/L/s at the end of the calibration. The results presented in this section are essentially results of calibrations carried out for the subsequent simulation o f SPM. The result o f a short simulation is shown in figures 5-8. There is a constant phase difference (90°) between shear stress/velocity and tidal height. Conversely, shear stress and velocity are in-phase and positively proportional to each other. Resuspension o f sediment due to large shear stress is shown by slightly higher SPM concentrations after each jum p o f bottom shear stress. 156 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. tide ■SPM/5(mg/L) -tau - - - velocity 0.8 0.4 - 200 100 150 0 50 Figure 5-8. Relationsh M concentration, bottom shear stress, and velocity at the moi ; relationship between velocity/shear stress and tidal phase. PM after each increase o f bottom shear stress. 0.30 0.25 0.20 0.15 0.10 0.05 0.00 1400 2000 Figure 5-9. Result o f a calibration test on the bottom shear stress at a grid cell near the mouth o f SDB. Horizontal axis indicates time steps, which covers 6 spring-neap cycles; vertical axis represents the shear stress in N/m^. 157 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The model shows that the bottom shear stress at some region near the bay mouth can exceed 0.5N/m^ (figure 5-9), which is set as the critical shear stress for erosion. As a result, current-induced sediment resuspension is significant at the bay mouth. A large area in south SDB also has high bottom shear stress, also resulting in intensive sediment resuspension. Compared to south SDB and the bay mouth, central bay usually sees no erosion o f bottom sediment due to tidal current. Rather, shipping events are the dominant sources o f resuspended sediments, as will be shown later. The pattern o f tidal resuspension rate can be clearly seen in figure 5-10. Note that figure 5-10 illustrates the sediment resuspension rate, not SPM concentration itself. 20 40 60 80 100 120 Figure 5-10. Result o f a calibration run for the simulation o f sediment resuspension based on equation 5-27. The gray scale in the graph represents resuspension rates in each time step o f 0 mg/L (light) to O.lmg/L (dark) with contours o f O.Olmg/L interval. 158 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.8.2 S im ulation R esults o f SPM C oncentrations Two types o f particulate matter with grain sizes o f 2pm and 20pm are included in the particle dynamic modeling, representing medium clay and medium silt, the dominant suspended particles in SDB (Chadwick et al, 1999). Coarse-grained particles are also significant, but their distribution is limited to the source of production (either river mouth or location o f shipping-induced resuspension) hence they have no significant impact on the geochemical sub-model. This scheme o f simplification is workable because o f the use of some empirical coefficients that can dampen the errors caused by the simplification. For simplicity o f presentation, only the results o f the fine particles (i.e. 2pm grain size), which are found to be the dominant SPM in SDB (>75%), are presented. It is also assumed that there is no particle-particle interactions, and the particles o f different grain sizes behave independently with each other and with other particles o f the same grain size. Usually 4 representative locations are monitored for the simulation results o f SPM and radionuclide. They are: station 8 at the south end of SDB (labeled as Sbay), station 5 at central SDP(labeled InBay), station 3 (labeled Brdwy) and station 1 (labeled Bmouth, see figure 3-1). All the subsequent simulation results o f the radionuclides will have the same labeling for consistency. These locations are selected because they represent the pattern o f the entire SDB, and there are sufficient observational data to be compared with the modeling results. Among them, station 8 159 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (i.e. Sbay) is sometimes omitted for elarity o f the graphs. At the same time, for clarity o f the graphs, a 15-step (equivalent to 15-hour in length) running average is taken for the simulation result, which smoothes out the high frequency ‘spikes’ caused by each individual tide that make the graphs unreadable (refer to figure 5-13 below). Tidal variations are still kept after the running averaging because the length o f time is short relative to tidal period. 120 tsslnbay 'tssSbay tssBrdwy 'tssBmouth 100 500 1000 1500 2000 2500 0 Figure 5-11. Part o f a 3-year simulation result o f SPM (labeled here as tss, or total suspended solids) for stations 8, 5, 3, 1, labeled as Sbay, Inbay, Brdwy and Bmouth, respectively. The vertical axis has a lOx factor. For station locations, refer to figure 3- 1. Two spikes for station 8 and 5 at step number 1100 and 1600 are caused by major precipitation events. Figure 5-11 shows part o f a 3-year simulation o f SPM. A general pattern o f SPM is that there are more SPM in the bay water than in the outer sea with a factor o f up to 10 (Chadwick et al, 1999; also refer to figure 5-5). However we have also observed 160 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. previously that there are considerable geographic differentiation within SDB(see also figures 4-8 and 5-5). For south SDB, high SPM come from the shallow region where tidal flooding and ebbing create large bottom shear stress. The simulated values are usually around 8-10mg/L(average 8mg/L) during steady-state condition without the influence o f major precipitation. At the same time, due to the tidal pumping, suspended particles in the south bay are not easily transported outside with tidal currents, so SPM level at south bay is constantly high. At the bay mouth, the effects o f strong resuspension (see figure 5-10) is largely diminished by dilution with clean water o f the outer sea. Similarly, north SDB constantly has lower SPM concentrations than inner SDB due to tidal exchange with outer sea, as observed before (see chapter 4). As will be seen later, the distribution o f SPM in SDB has profound effects on the patterns and behavior o f the particle-reactive radionuelides. 5.8.3 Sim ulation R esults o f R adionuclides Following are the simulation results o f the radionuclides ^'°Po, ^'‘ ’Pb and After a large number of simulations, it was found that quasi-steady state is reached after 2-3 days (figure 5-12), similar to length o f time to reach steady state for the original hydrodynamic model (Cheng and Casulli, 1993). The reason for the quick steady state for long-lived radionuclides (esp. ^'°Pb) is because o f the following reasons. First, the previous geochemical studies have provided good boundary and initial conditions (BC and IC) that shortened the length o f time needed for reaching a steady state. 161 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Particulate Pb-2I0 (dpm/m ) D issolved Pb-210 (dpm/m ) 40 35 30 25 20 15 10 5 0 0 500 1000 1500 2000 2500 3000 3500 4000 Figure 5-12. Illustration o f the short period o f time for the model to reach quasi- 9 1 n steady state. Particulate and dissolved phase Pb concentrations at station 3 (dpm/L, with a lOOOx factor, as all subsequent figures) are shown. Simulation time is about 21 days, so less than a week is needed to reach a quasi-steady state. Note that the first ‘hum p’ is caused by spring-neap cycle, not the instability o f the numerical model. Second, rapid processes such as tidal exchange, scavenging and settling contribute much more than radioactive decay to their patterns and variations, and these processes have essentially wiped out the effect o f the radioactive decay. Only processes such as 21 0 sedimentation and ‘ Pb stripping’ (to be discussed in chapter 6) need long-term simulations. The simulation results to be shown below representing a nearly 3-year simulation, which should be representative o f steady-state situation o f SDB. For illustration 162 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. purposes, figure 5-13 below shows the simulated tidal phase in a entire 3-year period. The spring-neap tidal cycle represents most o f the variance in the maximum tidal range o f 1.4 meter. Not all the data is shown for the purpose o f clarity, because too many data points make all graphs difficult to read and data files difficult to handle due to large file sizes (e.g. figure 5-13 alone represents 500,000 data points and a 40 megabyte file). Instead, simulation results o f about 8 spring-neap eycles are shown. 0.0 500 1000 1500 2000 2500 Figure 5-13. Tidal height (in meters on the vertical axis) simulated by RS-TRIM for 500,000 steps (2.85 years). Zero in vertical axis indicates MLLW. 9 1 A Part of the results o f the 3-year simulation o f particulate phase Pb activities is 91 0 shown in figure 5-14. For particulate phase Pb activities, there is a clear trend 163 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. within San Diego Bay o f station l>station 3>station 5, as expected based on our observation. Similar trend is found for the simulated dissolved phase ^'V b, as shown in figure 5-15. It is easily inferred that the trend is related to the increasing trend o f SPM from the bay mouth to inner bay, as shown in figure 5-11. From the particle dynamics point o f view, higher SPM concentrations in a steady state scenario correspond to higher settling rates o f SPM (refer to equation 5-30), which bring down the concentrations o f the dissolved phase radionuclide and eventually result in lower concentrations o f particulate phase radionuclide as well. Figure 5-14 and 5-15 both show clear trends o f dissolved and particulate phase ^'V b activities from inside SDB towards the bay mouth (note that for plots o f ^''^Pb and ^’^Po there is a 1000-time factor, for example, the number 20 on the vertical axis indicates 0.02dpm/L). This is consistent with the results o f the previous geochemical studies that showed similar trends. Considering the clear trend o f decreasing SPM concentrations from south SDB towards the bay mouth, as figure 5-11 indicates, the trends in figure 5-14 and 5-14 are expected. As discussed previously in this chapter, the concentrations o f radionuclides at a particular location should eventually be determined by the local particle dynamics when a quasi-steady state is reached. As shown in equations 5-30a and 5-30b, high concentrations o f SPM result in high settling rate, which in turn brings down the concentrations o f both dissolved and particulate phase activities o f ^’®Pb through scavenging. Compared with our 164 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 45 ■Pbplnbay ■PbpBrdwy PbpBmouth H i l l 15 500 1000 1500 2000 2500 Figure 5-14 Part o f the 3-year simulation results for particulate V b activities (xlOOO dpm/L on vertical axis). Horizontal axis unit: hour (real time). 40 30 20 ■Pbdlnbay “PbdBrdwy Pbdmouth m 10 500 1000 1500 2000 2500 Figure 5-15. Part o f the 3-year simulation results for dissolved ^'^Pb activities (xlOOO dpm/L on vertical axis). Horizontal axis unit: hour (real time). observation, the pattern is not always as regular as shown in figure 5-14 and 5-15, but with certain variations (refer to figure 4-25a and b) for the following reasons. First of 165 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. all, the scavenging process is modeled as an equilibrium process that is controlled by Kd, as a result, particulate and dissolved phase ^**^Pb co-vary closely. In fact, scavenging is a kinetic process that takes time to reach equilibrium. It was estimated that the time could be days or even longer (Baskaran and Santschi, 1993; Santschi et al, 1979). Second reason for the difference between modeled result and observation is some random processes such as rain, shipping-induced resuspension and the randomness o f the tidal flow itself, which are not modeled in a deterministic way. As already shown in figures 5-15, precipitation events have major impact on some parts o f SDB (the effects o f precipitation on chemical species will be discussed later in this chapter). So the modeling o f ^*°Pb in both particulate and dissolved phases is deemed satisfactory. 55 45 35 25 15 ■Poplnbay ■PopBrdwy PopBmouth “4 ’4 4 ..... I 500 1000 1500 2000 2500 Figure 5-16. Part o f the 3-year simulation results for particulate ^'°Po activities (xlOOO dpm/L on vertical axis). Horizontal axis unit: hour (real time). 1 6 6 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 15 10 Podlnbay ■PodBrdwy Podmouth % A II 500 1000 1500 2000 2500 Figure 5-17. Part o f the 3-year simulation results for dissolved ^^Vo aetivities (xlOOO dpm/L on vertical axis). Horizontal axis unit: hour (real time). The simulation result for ^^®Po is shown in figures 5-16 to 5-17. The geographic trends and particulate-dissolved phase distributions are similar to those o f ^*®Pb except that ^^*^Po has lower activities in dissolved phase than ^^®Pb, and the opposite is 0 1 A T 1 A observed for particulate phase Po and Pb activities. Besides, the differentiation 0 1 A 0 1 A o f dissolved activities o f Po is clearer than that for Pb (compare figure 5-15 and 5-17). The differences between the pattems o f these two radionuclides are clearly due to the difference in their particle affinity, as revealed by previous geochemical studies. A comparison between particulate phase activities o f ^^*^Po and ^'°Pb (figure 5-18) showed a higher activity for ^*®Po than ^'*^Pb, which is consistent with the notion that 167 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PopBrdwy PbpBrdwy m 2500 Figure 5-18. Comparison o f particulate and ^*°Po activities at station 3 (xlOOO dpm/L on vertical axis). Horizontal axis unit: hour (real time). ^'®Po has higher particle affinity that On the other hand, the magnitude o f the difference between these two radionuclides is determined by SPM concentration as well as the activities o f the dissolved phase ^^®Ph. For example, figure 5-19 shows a larger difference between the activities o f particulate phase at station 1 near bay mouth than that at station 3 toward the inner bay (figure 5-18). If figure 5-18 and 5-20 are eompared, they showed that at station 3, dissolved phase ^^°Pb activity exceeds that o f ^‘V o , and vice versa for the particulate phase activities o f the two radionuclides. This is consistent with the observation, where ^*^Po shows clearly higher particle affinity. 1 6 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 ■ PbpBmouth PopBmouth ' I 500 1000 1500 2000 2500 Figure 5-19. Comparison o f particulate and ^^°Po activities at station 1 (xlOOO dpm/L on vertical axis). Horizontal axis unit; hour (real time). 30 20 10 -PbdBrdwy ■PodBrdwy 500 1000 1500 2000 2500 Figure 5-20. Comparison o f dissolved ^^°Pb and ^^Vo activities at station 3 (xlOOO dpm/L on vertical axis). Horizontal axis imit: hour (real time). 169 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Due to the scarcity o f data and larger analytical uncertainties in the present studies, the parameters related to the scavenging o f ^^'*Th by particles could not be 01 n ‘ 7 1 n estimated with the same confidence level and precision as those for Pb and Po. However, the geochemical studies do support the general notion that ^^'^Th has higher particle affinity than both ^*®Pb and ^**^Po, as can be seen from the o f logKd- logSPM relationship (equation A=1.09xl0^ and B=1.31. The simulation results were shown in figures 5-21 and 5-22. For dissolved phase the sequence o f activities at different stations is as expected; Station l>Station 3>Station 5, with ranges o f 0.45- 0.7dpm/L for bay mouth, 0.35-0.65dpm/L for Station 3, and 0.3-0.4dpm/L for Station 5, corresponding to residence time o f about lld ay s, 8.8days, and 5.8days, respectively for three stations, consistent w ith the range o f 1-16 days as calculated from the previous geochemical studies. As for particulate phase Th, the sequence o f activities for the three stations are the same as for ^^‘ ^Pb and ^*‘ ^Po(staionl>station 3>station 5) but the ranges are narrower, as shown in figure 5-22. The range o f variation within each station has the following sequence: station l>station 3>station 5, and the variations follows the tidal phase closely. This pattern is clearly due to the differences in the strength o f tidal exchange between SDB and outer sea, as well as the differences in SPM concentrations in the two locations. If figure 5-21 and figure 5-22 are eompared, it can be seen that particulate phase Th activities are much higher than dissolved phase. Compared to 170 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. •Thplnbay ■ThpBrdwy ThpBmouth 500 1000 1500 2000 2500 Figure 5-21. Dissolved phase activities o f at stations 1, 3 and 4 (dpm/L, with a factor o f lOx). Horizontal axis unit: hour (real time). 6 Thdlnbay ThdBrdwy Thdmouth 5 4 3 2 0 1000 500 1500 2000 2500 Figure 5-22. Particulate phase activities o f ^^"^Th at stations 1, 3 and 4 (dpm/L, with a factor o f lOx). Horizontal axis unit: hour (real time). 171 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 210 210 • * 234 Po and Pb, the differences between the activities o f Th in the dissolved and particulate phases are the largest as expected for ^^"*Th due to its stronger affinity for particles than both ^'°Po and ^’* ’Pb. 5.8.4 Simulation Results of Precipitation Effects Precipitation events are modeled to be stochastic processes that take place with a prescribed probability, so the simulations need to be run for a relatively long time in order to encounter precipitation events and to evaluate their effect on the patterns of chemical species. It is assumed that there would be 2 major rains o f 5cm, 5 medium I rains o f 2cm, and 5 minor rains o f 1cm. The simulated precipitation events may have large variations if the time o f simulation is not long enough. For a 3-year simulation, as presented here, 6 major rains, 8 minor ones and 26 minor ones were recorded. This translates into 24cm annual rainfall on average, consistent with the preset rainfall pattem. In the 2500-hour simulation result as shown in the above figures (figure 5-14 to 5-22), there are two precipitations at aboutl200th and 1600th hour, respectively. As described earlier in this chapter, the effect o f precipitation on the chemical species o f the model domain is two-fold. First, precipitation creates river flow (modeled to be • t • • •• 9 1 0 9 1 0 dry if there is no rain), which carries dissolved Pb, Po and SPM into SDB. The 91 0 9 10 values for Pb, Po and SPM were all set according to observational data. 172 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Secondly, precipitation event brings fresh water directly to the bay water itself and dilutes concentrations o f all chemicals except for those o f ^ 'V b and ^'‘ ’Po in both phases because they have higher activities in rainwater than in bay water. For both river input and direct atmospheric input dilute its concentration. The effect o f precipitation events on SPM and radionuclides are visible in figure 5-11, 5-14 to 5-22 but not in every station. The effects o f added SPM, ^''’ Po and ^*^Pb are clear, but largely only on the stations close to major rivers (i.e. station 5). The effects become much weaker at station 8 at south SDB, and virtually indistinguishable at stations 1 and 3. Unfortunately there is not enough data to calibrate the modeling results. The closest one has only two stations at SI and GP, with higher activities o f ^'°Po for station SI in both particulate and dissolved phases than station GP. For ^^"^Th, the opposite was found, with GP had higher activities o f ^^"^Th in both ')'1A particulate and dissolved phases. The Th pattem is consistent with the modeling result, and the ^'°Po pattern also supports the model. In the model the effects o f rain • 910 « • • • are not important to affect Po activities at stations SI and GP, so the original gradient is preserved, i.e. ^“’Po is higher in both phases at station SI than at station GP. • • 9 T 4 • • • On the other hand, the dilution effect for Th due to rain is relatively small, because the inputs o f fresh water from rainwater and river are not important on the volumetric 173 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. basis. On the contrary, it seems that the precipitation events had drawn down the concentration o f both dissolved and particulate phase perhaps due to the quick 'y'iA scavenging o f Th by the high level o f SPM supplied by the storm runoff (in the model, the SPM level in storm water is 6 times more than the background SPM level in SDB). 5.8.5 Simulation Result of Particle Scavenging and Deposition One of the important findings o f our previous geochemical studies o f SDB was that there are very large enrichment factors in the north SDB sediment column (up to 12 near station #3, refer to table 4-1). This finding can be explained in principle by the 9 1 n fact that SPM remains high at north SDB and dissolved Pb is supplied continuously from outer sea through tidal exchange. The continuous stripping o f dissolved ^’°Pb from water column down to the surface sediment should result in higher sediment 91D • 910 • • inventory o f Pb. However, the unusually high Pb inventory in sediment column 9 1 A is not readily calculated from geochemical data because the ratio o f 3-5 between Pb inventories o f station 3 and those at south SDB is not justified by the differences in the measured water column data. It was suspected that a repeated tidal exchange needs to play a role in the process, and RS-TRIM provides the only tool to justify the 9 1 0 observational data. As shown by equation 5-30a and 5-30b, the amount o f Pb falling onto the sediment surface is determined by the following variables: particulate 174 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. * • 9 1 0 • phase activity o f Pb, SPM concentration, and settling rate o f SPM. By calculating every variable in each time step, the model should be able to produce a comparable 0.8 - depoBrdwy ■ depolnbay depoBmth 0.6 0.4 500 1000 1500 2000 2500 0 210 2 Figure 5-23. Simulated accumulation rate o f Pb in dpm/em /step for stations 1, 3 and 5. The vertical axis has a factor o f 1000. Florizontal axis unit is hour (real time). sediment column inventory o f ^'*’ Pb if the geochemical modeling scheme is « 91 0 conceptually correct. In fact, the sediment column inventory o f Pb is a stringent calibration standard because it provides a strong constraint on the performance o f the model over a long period o f time. As a result, even a small glitch in the geoehemieal model can lead to exorbitant simulation result. The good fit o f the modeled result as shown in figure 5-23 is a convincing testimonial that the coupled geoehemieal- hydrodynamic model is running correctly. More convincing evidence to justify the 175 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • 9 1 n • ♦ high Pb inventory in north SDB sediment is provided by a mass balance calculation to be diseussed in the next chapter. 5.9 Concluding Remarks on the Results o f Geochemical and Hydrodynamic Modeling In summary, the simulation results o f SPM, activities o f ^'°Pb, ^“’Po and ^^"^Th in both dissolved and particulate phases at steady state during a long-term simulation fitted well with the observational data during the previous geochemical studies. The agreement exists at both spatial and temporal domains in that the gradient in the activities at different stations is observed consistently in the geochernical studies and was successfully simulated by the coupled RS-TRIM model. At the same time, the 2 1 0 sediment inventory o f Pb provides a strong evidence that the model is running correctly. However, care should be taken on the following issues. First o f all, there are some parameters in the model that are not very well defined. They include the exact volume o f annual river flow and average SPM concentration in the river water, rain water 9 1 n 9 1 n Pb and Po concentrations. These parameters are not crucial to the well-being o f the model (as to be shown in the subsequent chapter, they are only minor contributors 9 1 D o f Pb in the sediments), but they play important roles in transient variations o f the 176 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. radionuclides during storms. Secondly, the use o f the distribution coefficient to simulate the scavenging process has advantages and disadvantages. The advantage of this scheme is safe, simple and efficient if the selection o f the distribution coefficients is based on sufficient observational data. The disadvantage is that the scheme may distort some transient pattems. For example, the model could smooth out jum ps in Tin Tin dissolved Pb and Po aetivities immediately after a precipitation event due to concurrent high SPM concentration, although the actual process may takes days to take effect. This could be the reason o f the absence o f precipitation effects for the stations in the north SDB (figures 5-14 to 5-22). Thirdly, there are large uncertainties in the empirical approximation o f distribution coefficients (figure 4-20), so the accurate simulation o f scavenging process is not guaranteed. Lastly, there is good agreement between the observational data and the modeled result in both temporal and spatial domains, it should be cautioned that the observational data might have been affected by transient processes including tidal phase, recent precipitation events, man-made processes such dredging and passage of large boats shortly before sampling, etc., exactly as shown by the model. For example, the water column data for the summer 1999 sampling were all from samples collected during a flood tide. As shown in the subsequent data, tidal phases have a major influence on the transient variations o f the aetivities o f radionuclides. In this sense, samples collected simultaneously at a known tidal phase, or those collected at the same location over a 177 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. period o f time were more valuable. However, the effects of some transient processes still cannot be eliminated. Despite the cautions listed above, there is enough confidence that the coupled geochemical-hydrodynamic model is functioning correctly. The model has been able to reproduce all the observed spatial and temporal variations/patterns with satisfactory agreement. This good agreement between the simulated and observed sediment inventory o f ^’*^P b is evidence that the long-term performance o f the coupled model is satisfactory. 178 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 6 Discussions 6.1 General Statements Our study is the first hydrodynamic modeling o f the biogeochemistry o f naturally- occurring, particle-reactive radionuclides in a tide-dominated, dynamic coastal environment. The long-term, continuous sampling in and around San Diego Bay, including two time-series measurements, gave a picture o f spatial and temporal 9 1 n 9 1 A 9 T 4 variations o f the radionuclides Pb, Po, Th, and suspended particles. The development o f RS-TRIM, a geochemical-hydrodynamic model, provides an unprecedented link between coastal ocean hydrodynamics and biogeochemistry of particle-reactive radionuclides. In view o f the extensive application o f particle- reactive radionuclides as tracers on the studies o f ocean mixing, particle dynamics, ocean biogeochemistry, and environmental issues in the coastal oceans, this study makes a significant contribution by linking the behavior and spatial/temporal patterns o f radionuclides with ocean hydrodynamics. The link between the radionuclides and hydrodynamics in RS-TRIM is established through particle dynamics, which is also an extension o f the original TRIM model. Particle dynamics controls the behavior and patterns o f the radionuclides through scavenging, resuspension, settling and tidal exchange, with the hydrodynamics acting as the underlying driving force for all these processes. 179 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In this study, hydrodynamic modeling and biogeochemical study rely on and benefit from each other. The biogeochemieal studies provide data that set boundary and initial conditions for the numerical model and provide criteria o f both transient, long term and steady-state calibration and verification for the new geoehemieal- hydrodynamie model. On the other hand, the numerical model provides a spatial and temporal domain in which biogeochemieal processes are driven by explicit hydrodynamic laws that are solved by well-established numerical schemes. In this chapter, major issues such as mass balances o f water and radionuclides in SDB, residence times o f the radionuclides, will be diseussed. Other issues such as the recycling o f 210Po in the water column, environmental significance o f this study, as well as the desired future research along the trail that is blazed by this study will also be diseussed relatively briefly. In the end, a summary and conclusion will be presented. 6.2 Mass Balances o f Water, and in San Diego Bay For a semi-elosed system like San Diego Bay, mass balance calculations can provide a valuable assessment o f the measurements o f geoehemieal species in terms o f the input, output and the fluxes between different parts o f the system. Based on the box model, it is easier to calculate the residence time o f each species in the system. 180 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. During the construction o f the box-model, some unknown input/output terms can be derived from mass balance considerations. Sometimes these terms can only be calculated this way. The mass balance o f geochemical species requires that the time rate o f change (TROC) o f a reservoir is the sum o f all input terms (sources) minus the sum o f all output terms (sinks), which can be expressed as follows: A reservoir = S input - Z output (6-1) Where A reservoir is time rate o f change (TROC) o f a reservoir Z input is sum o f all input terms Z output is sum o f all output terms In many cases, steady state is assumed. Thus the right hand side o f equation 6-1 is zero, or equation 6-1 becomes: Z input= Z output (6-1') In the following sections equation 6-T will be applied for water and radionuclides of San Diego Bay. 181 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.2.1 W ater Balance o f San Diego Bay The mass balance diagram o f water in SDB is shown in figure 6-1; X < Ir Figure 6-1. Mass balance o f water o f San Diego Bay In steady state, the mass balance equation for water within SDB can be written as: AV^=0 = lr+lp-Oe+X (6-2) where AVw is the change o f the volume o f water in SDB Ir is the input o f fresh water from rivers Ip is the input o f fresh water from direct precipitation into SDB Oe is the output o f fresh water due to evaporation X is the imbalance o f water that makes up the non-zero sum o f Ir, Ip and Oe that makes the right side o f equation 6-2 to be zero. This can be groundwater input or exchange with outer sea. 182 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5.E+04 . 4.E+04 3.E+04 2.E+04 1.E+04 O.E+00 1980 1985 h 1990 1995 2000 Figure 6-2. Record o f flow in acre-feet over Sweetwater Dam since 1980. Data courtesy o f Michael Garrod o f Sweetwater Authority. O f the above terms, Ir, the amount o f water from river is the most difficult to define due to the sporadic nature o f the surface runoff. SDB watershed is comprised o f Sweetwater, Otay and the Pueblo San Diego hydrologic units. The corresponding rivers flowing into San Diego Bay are Sweetwater River, Otay River, and Chollas 9 9 9 Creek, with drainage area o f 540 km , 360km , and 70km , respectively (Chadwick et al, 1999). For Sweetwater River, the surface flow is largely trapped by the Sweetwater Authority's two dams on the Sweetwater River. As a result, surface flow from the drainage area upstream o f the Sweetwater Lake, which is 76% o f the total drainage area, is impounded by the reservoir and no constant flow exists in the riverbed except during storm events, making it impossible to measure (see figure 6-2). The drainage area is composed o f 29% urban area, 22% open space/agriculture and 49% undeveloped area (forest/grassland), the retention factors (the percentage of precipitation retained by soil that do not reach the rivers) of these are estimated to be 30%, 90% and 90%, respectively (www.portofsandiego.com). The contribution of 183 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. freshwater from Sweetwater River to San Diego Bay with an average annual rainfall o f 25cm can be calculated as follows: h i= A * P * S fa* (l-R i) (6-3) Where Iri is the riverine input o f freshwater into San Diego Bay(m /yr) A is the drainage area (m ) P is the precipitation in m/yr Ri is the retention factor for type i land surface fa is the fraction o f type i land surface. The resulting V is 2.18 x lO^m^. Similarly, the annual input o f freshwater from Otay River (effective drainage area: 90km^, 90% undeveloped, 10% open space) and Chollas Creek (90% urban, 10% open space) are 1.4xl0^m^/yr and 2.1xl0^m^/yr. As a result, the total annual input from the major rivers into SDB is 3.23 x lO^m^ or 3.23 x lO^L/yr. This translates into about 1.2% o f the volume o f SDB at MLLW, which is 2.8x1 0 "L . Ip, the input rate o f fresh water from direct atmospheric precipitation (calculated by multiplying area o f SDB with annual rain fall) is 1.08*10'*’ L/Yr. The annual 184 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. evaporation Oe is calculated by multiplying annual evaporation by the area o f SDB, resulting a rate o f loss for freshwater in SDB o f 6.88xlO*°L/yr. Bringing all variables into equation 6-2, the only unknown, X, the imbalance in the water budget is 5.48*10*°L/yr, or 2.55% o f the total volume o f SDB. This should be the amount o f net input from outer sea in the form o f ‘residue currents’. The significance o f this will be discussed later. 6.2.2 M ass B alance o f A good understanding on the biogeoehemistry o f ^'*^Pb must be based on solid quantifications of all source and sink terms o f both radionuclides. Based on the extensive sampling and measurements made in the past 4 years, the mass balances o f ^’®Pb for San Diego Bay can be established. As for ^'^Po, no separate mass balance calculation will be performed since it is expected to largely follow the pattern of Concentrations o f ^'°Pb in river waters were measured for samples collected during July 6 and November 6, 2002 sampling trips (refer to chapter 4). Despite the large variations of the data due to the sporadic nature of the flow, the first-degree estimate can be made. 185 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 n • For Pb, there are 4 data points for Sweetwater River low tide, Otay River low tide, and Otay River high tide, and San Diego River. Only total was measured for these samples, with values of 0.027, 0.41, 0.11 and 0.61 dpm/L, respectively. O f these, the San Diego River sample was only for reference because it does not drain to San Diego Bay, and Otay River high tide sample was essentially the back-flushed San Diego Bay water. So only the Otay River low tide sample and Sweetwater River • 9 1 n • » sample data are used here. The resulting average total Pb in river water is estimated to be 0.15dpm/L. Thus the amount o f ^’^Pb from river into SDB is 3.26x1 O^dpm/yr. 9 1 0 Assume 50% o f the riverine Pb is in particulate phase, 50% in dissolved phase (only total activities o f ^''^Pb were measured for river water samples, see chapter 4), 1.63xl0^dpm/yr will be the input o f ^'^Pb from river in both phases. The assumed phase distribution o f ^'^Pb is not critical to this study because they are assumed to be exchangeable, which is driven by equilibrium distribution coefficient (see chapter 5). • • • 9 1 0 • Direct atmospheric input o f dissolved phase Pb onto SDB is the product o f the area o f SDB and the atmospheric fallout rate o f ^^®Pb. The result is 8.6x10 ’®dpm/yr. 910 • • • • « 910 The particulate Pb reservoir for SDB, taking the activity o f particulate Pb to be 0.02dpm/L, is 5.6xl0^dpm; similarly for the dissolved ^*°Pb reservoir when dissolved ^’®Pb is taken as O.Oldpm/L, the reservoir size is 2.80xl0^dpm. 186 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The sedimentation sink o f is calculated as follows. The average sedimentation rate is about 0.5cm/yr, or 0.5cm^/cm^/yr. If the average surface sediment density is o 'y assumed to be 1.5g/cm , the sedimentation rate is 0.75g/cm /yr. The summer 1999 sediment core data suggest that the average exeess ^'°Pb activity o f surface sediment is about 2dpm/g, so the sedimentation rate o f ^'*^Pb is 1.5dpm/cm^/yr, and overall annual sink o f sedimentation inside SDB is 6.5x10"dpm /yr. Tidal exehange volume estimation is based on the tidal prism given by Wang et al (1998) for SDB, where on average 24% o f the volume is exehanged during a flooding cycle. However, according to Largier (1995), only half o f this amount or 12% o f the volume o f SDB will be exchanged beeause 50% o f the water flushing baek to SDB is the ‘old’ water flushing out during the previous ebbing tide. For the amount o f dissolved phase ^^°Pb that undergoes exchange, the difference between two waters is 0.03dpm/L, so the amount o f dissolved ^*'’Pb to be provided by exchange should be 3.67*10*’dpm/yr (exchange volume multiplying the difference in activities). Similarly the partieulate ^^°Pb net input due to tidal exehange is 2.46*10"dpm /yr when 0.02dpm/L is used for the difference in partieulate ^*® Pb activities between SDB water and open ocean water. 187 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. To keep dissolved ^’°Pb balanced, the scavenging should be the difference between the radioactive decay and the sum o f tidal exchange and atmospheric input and river input, the result is 4.55*10"dpm /yr. The mass balance diagram is shown in figure 6-3, with each input-output terms explained as follows. Atmospheric input 8.6xlO'°dpm/y Excess: ^ _ 4.3*10"’dpm/y River input 1.63xl0^dpm/y Particulate 2 I 0 p b 5.6x1 O^dpm Concentration: 0.02dpm/L ----------------------i ^ scavenging Dissolved ^'“Pb pool 2.80x1 O^dpm Concentration: O.Oldpm/L i River input: 1.63xl0^dpm/y 4.55*1 o ’'dpm/y ^................ w Decay ^ 1.74xl0*dpm/y Decay ^ 0.87xl0*dpm/y Sedimentation 6.5xlO''dpm/y 1 Exchange:2.46 ^ *10"dpm/y Exchange:3.67 ^ *10” dpm/y Open ocean particle ^'®Pb: 0.04dpm/L Open ocean dissolved "'“Pb: Figure 6-3. Mass balance chart for ^'°Pb in San Diego Bay 188 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. The apparent imbalance for the particulate phase is 4.3*10'®dpm/yr, indicating the overall sink o f particulate is slightly smaller than the combined source. This accounts for about 6% o f the overall input amount and is comparable to the analytical uncertainty o f ^'*^Pb. Uncertainties in the above diagram could also come from the following sources: 1. Vigorous tidal exchange creates a ‘buffer zone’ between SDB water and outer 9 1 0 seawater such that the net gain o f both particulate and dissolved Pb could be smaller than estimated. This may be the most significant source o f uncertainty because it is at least an order o f magnitude larger than all other terms except for scavenging and sedimentation; 2. Remineralization o f particulate-phase ^’* ’Pb back to dissolved phase is possible but difficult to define. If ‘scavenging’ is taken as ‘net scavenging’ that includes both scavenging and remineralization, no additional consideration of remineralization is needed. 3. The sedimentation sink has relatively large uncertainties. For example, excessive sediment mixing/bioturbation in surface sediment will result in underestimation o f surface sediment ^'^Pb activity. Combining the mass balance calculation o f water and o f Pb it is interesting to note that the net input o f ^'®Pb from outer sea into SDB is 2.23*10^dpm/yr due to a net 189 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. input o f water from outer sea with higher concentrations. However, this amount is negligible ( i.e. a factor o f 275 ) compared to the net input o f ^'®Pb o f 6.13xl0' ’dpm/yr in combined dissolved and particulate phases from tidal exchange. 6.2.3 M ass B alance o f ')'1 A The calculation o f the mass balance o f Th suffers from the large analytical uncertainties in the previous geochemical studies, so it should not be treated with the same confidence as with ^'^Pb. Due to the short half-live of the mass balance is calculated on a daily basis. For dissolved Th, the sources include production from dissolved U and tidal exchange with outer sea, which has a higher concentration o f dissolved ^^'^Th. The production from is calculated as follows: Vsdb*A238*A ,234=2 .80x 10*'L* 2.4dpm/L*0.029 day'* =1.95*10’°dpm/d; the input from tidal exchange should be (volume o f SDB)*(portion o f SDB volume being exchanged daily)*(concentration difference between SDB water and outer sea water), or 2.8*10'*L*12%/day*(0.8- 0.3)dpm/L=1.68*10*°dpm/d. The decay o f 234Th is 4.03*10^dpm/day, so another sink term, scavenging, should be the difference between decay o f ^^"*Th and the sum o f production from and tidal exchange, or 1.68x10*^+8.1x10^- 4.03x10‘ ^=2.09* 10"*dpni/d. 190 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Similar calculations are carried out for the particulate phase ^^^Th, but the sedimentation sink needs some careful evaluation. The sediment accumulation rate is 'y T • > 'y'XA about 0.75g/em /yr, or 2.05*10' g/cm /d. However, the activities o f excess Th at the surface sediment averaged about only 2dpm/g (refer to chapter 4), compared to the estimated particulate phase Th activity o f 50-100dpm/g from September 6, 2002 cruise data. The apparent 2dpm/g surface sediment activity o f ^^"^Th is certainly an underestimate due to sediment mixing. The closest reference for the activity of 234Th in particulate matter is the winter 2000 sediment trap data (see table 4-2), which suggested that the activity should be at least 20dpm/g dry weight under average conditions (table 4-2 data were obtained after several storm events, which might have diluted the activity o f 234Th in the trapped sediment). If we assume that 25dpm/g represents the average activity for the particles falling to the sediment-water interface, the sedimentation sink o f ^^“ ^Th should be 2.20*10'®dpm/d. 9 T 4 The mass balance o f Th can be schematically shown below in figure 6-4. It is seen that particulate ^^'^Th is not balanced. The unaccounted amount is 1.40xl0’**dpm/d, which is comparable with the size o f the sink o f sedimentation. This relatively large error could come from the following sources: 994 • • • 1. Large analytical errors for Th. This leads to significant uncertainties o f the amount o f input for both dissolved and particulate phase ^^"^Th through tidal 191 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. exchange. If the gradient between SDB water and outer sea water is small, as 234 the statistical analysis on Th data suggests, the discrepancy could be much smaller; excess: 1.40*10‘''dpm/d sedimentation 2 .2 0 x 1 o 'V p m /d particle p o o l: 1.4*10"dpm activity: 0.5dpm/L exchange 1.01 *10"’dpm/d open oeean particle 0.8dpm/L scavenge: 3.39*10'® dpm/yr decay, 8.06x10® dpm/d dissolved pool 8.4x10'®dpm activity: (0.3 dpm/L) exchange: 1.68* 10'®dpm/d open ocean dissolved T h^"'': 0.8dpm/L 238 u 1.95*10'® dpm/d decay: 2.44x10® dpm/d Figure 6-4. Mass Balance diagram o f Th in San Diego Bay 2. The amount o f Th depositing onto the sediment surface may be underestimated. As stated above, the particulate-phase ^^"^Th activities are mostly in the range o f 50-100dpm/g dry weight, but the sediment trap data showed a lower activity o f about 25dpm/g. If instead 50dpm/g were used in the above calculation, there will be no imbalance in the box model. 192 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.2.4 M ass B alance o f Suspended Particulate M atter in SDB The mass balance for SPM seems to be fairly simple to construct. Based on the steady state assumption, the addition o f SPM from river input (including storm runoff) should be equal to the loss to the outer sea due to tidal exchange. The corresponding calculations are listed below. For river water, the input o f SPM to SDB should be the average SPM concentration (120mg/L, see chapter 5) multiplied by the volumetric annual flux o f river water (3.23xlO^L/yr), or 3.88xlO*g/yr for the entire SDB. However, even without considering the loss o f SPM through tidal exchange, the riverine input o f SPM translates into an average sediment accumulation rate o f merely 9.02x1 C'^g/cm^/yr for SDB. The estimated linear sedimentation rate, in contrast, is 0.5cm/yr, or roughly 0.75g/cm2/yr if we assume the surface sediment has a density o f 1.5g/cm^. If the loss o f SPM to outer sea via tidal exchange is considered, the discrepancy becomes even larger. For example, the volume o f water being exchanged is roughly 35-45 times of the volume o f SDB based on the water residence time of 11 days as well as the volumetric tidal exchange o f 12% o f SDB volume during each tidal cycle. If the difference in SPM concentrations o f SDB water and outer sea water is taken as Img/L, the loss o f SPM from SDB to outer sea is 6.6xl0'*^g/yr, or more than two orders o f magnitude larger than the riverine input. The loss o f SPM to outer sea via tidal exchange is much less than the apparent amount of SPM that settles to the 193 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. sediment surface, which is reasonable. But the enormous discrepancy between riverine input o f SPM and sediment accumulation rate is troubling. The discrepancy could come from one or more o f the following sources: 1. Riverine concentrations o f SPM is underestimated; 2. Riverine input o f water is underestimated; 3. Storm runoffs via storm drains should be considered; 4. Sedimentation rate o f SDB is overestimated; 5. Sedimentation rate has decreased dramatically in recent years, but the sediment core profiles o f ^'^Pb did not show the decrease because o f mixing; 6. There are extensive sediment redistribution within SDB. Deeper region accumulates sediment at the expense o f the erosion o f shallow regions due to wave and tide induced resuspension. A simple order-of-magnitude analysis precludes the above factors 1, 2 and 4 to be major contributors because the observation data corroborates the previous estimates. Therefore, only factors 3, 5 and 6 are possible reasons for the discrepancy. Since factor 6 has no bearing on the overall mass balance o f SPM for SDB, all that is left is factors 3 and 5 in the above list, which implies that either SDB is not accumulating sediment nowadays, or storm runoffs from storm drains account for most o f the sediments accumulated in SDB, or both factors plays significant part in the scenario. 194 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.3 Residence times of and in San Diego Bay The residence times are good indications o f how fast these radionuclides are removed from dissolved form to particulate form, and settling o f these radionuclides in particulate form to the sediment surface. The box model as shown in figure 6-3 makes it convenient to calculate the residence time by the definition o f residence time (Cochran, 1992): X r e s = I / F ( 6 - 4 ) where F is the removal flux other than radioactive decay I is the size o f the inventory o f the radionuclide for the reservoir Firstly, the time rate o f change (TROC) for ^'°Pb in particulate and dissolved forms • • 9 9 9 . can he expressed as follows, if the production from Rn is assumed to be negligible: (Bacon et al, 1976; Masque et al, 2002): ^ =0=Fpb -^Pblpb^ -A + R p b ^* +Xpb“ (6-5a) SPbP = 0 = - A , p b I p b ’’ + R p b ’^ + X p b ’’ - S ( 6 - 5 b ) 195 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. • • • 9 1 n The corresponding residence time for Pb in dissolved and particulate form is then calculated by the following equations: Tpb‘ ^= Ipb“/A (6-6a) T Pb‘ ’=Ipb’ ’ /S (6-6b) where Fpb is the atmospheric fallout rate o f ^’®Pb 21 0 X is the decay constant for Pb; 9 10 I is the inventory o f Pb in a specific form (superscript d or p indicating dissolved and particulate forms, respectively) obtained by multiplying its concentration by the volume o f SDB R is the river flux o f ^’°Pb 21 0 A is the scavenging (adsorption) flux o f dissolved Pb 9 1 0 X is the flux o f Pb via tidal exchange with outer sea 9 1 0 S is the sedimentation flux o f particulate phase Pb onto sediment surface 91 0 • • T is the residence time o f Pb in a certain phase Since all the terms are given in the box model as shown in figure 6-3, the residence times o f dissolved and particulate ^‘°Pb can be easily calculated to be 2.25 days and 3.14 days, respectively. The differences between the two are mainly from the 196 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. difference in the reservoir sizes for particulate and dissolved These values are eonsistent with that o f Narragansett Bay o f 1-5 days (Santschi et al, 1979). Sueh short residence times underline the intensity o f particle scavenging and settling. It should be noted that tidal exehange are the driving force behind the short residence times by 9 1 A supplying large amount o f Pb, which is scavenged and brought to the sediment surface quickly. 234 Similarly for Th, we have the following governing equations for the mass balances: 5Xli^ =0= Xjhlu - ^Thljh^ - Axh + X xh^* (6-7a) 3Xli^ =0= - X.Thlih’ ’ +Rtk'’ + XTh’ ’-S (6-7b) 9 1 A The corresponding residence time for Th in dissolved and partieulate form is then calculated by the following equations: tTh^- Ixh^/A (6-8a) Txh'’= W /S (6-8b) Where 9 T 8 9 T A lu is the reservoir size o f U, which produces Th in dissolved form. Other terms are the same as for ^*^Pb. 197 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. From equations 6-8a and 6-8b, the residence times for dissolved and particulate are easily calculated to be 2.5 days and 3.9-6.4 days (depending whether the imbalance o f 1.4xl0’° dpm/day is included in the calculation or not), respectively. Based on the data o f total, dissolved and partieulate phase ^^"^Th from numerous sampling trips and cruises (see chapter 4), the residence times o f dissolved are found to range from 2-16 days (mostly 2-6 days) for most stations inside SDB. For particulate-phase the range is larger (1-15 days, mostly 1-7 days), primarily due to the complex hydrodynamics within SDB that could affect the production and settling o f particulate matters. Residence times o f total accordingly, range 5-30 days, and average about 7-10 days. Based on the mass balance calculation above, the residenee time o f total Th is 6-10 days, compatible with our geochemical studies. * 210 • • 6.4 Recycling o f Po within Water Columns 9 1 0 Our measurements o f Po activities in water column around SDB have revealed an 210 21 0 almost ubiquitous excess Po over its grandparent Pb. It has been found that in the 91 0 9 1 0 open ocean, surface water has a Pb/ Po ratio o f about 2 due to atmospheric input 21 0 210 o f Pb and a preferential removal o f Po by biogenic particles (Nozaki et al, 1976; Radanovitch et al, 1999). Corresponding to the preferential removal o f ^'*^Po from the 91 0 surface, excess Po was found near the thermocline, where many biogenic particles (up to 95%) are recycled and ^'‘ ^Po is released. Kadko (1993) found that in the 9 1 0 California Coastal Transition Zone, Po had a recycling inventory of 145% 198 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 9 1 n compared to less than 1% for Pb. Since Po is deemed to have much higher affinity for cytoplasm (Friedrich and Rutgers van de Loeff, 2002), biogenic particles usually have higher ^’* ’Po activities (could be higher than lOOdpm/g, Cherry and Heyraud, 1982) and Po/Pb ratios, in some places the ratio can be as high as 5 (Radanovitch et al, 1999). Even higher ratios were found for individual organisms, for example, as high as 12 for zooplankton (Masque et al, 2002). In San Diego Bay, the highest 0 1 A "91 n • Po/ Pb was found during the winter o f 2000 (figure 4-12) after a storm, where the ratio was 4.5. It was not clear how the high ratio was connected to the storm, but the 910 91A 910 widespread excess o f Po over Pb suggests that Po is indeed recycled vigorously within the water column. Compared to the open oeean or coastal ocean, SDB is shallow and has as much as 10 times higher primary productivity than the adjacent open ocean (Chadwick et al, 1999), so there should be plenty o f biogenic particles available for recycling. Given the long (138 days) half life o f ^'°Po compared to the short time scale o f sediment resuspension and settling, recycling o f ^*®Po in the water column is possible, especially in view o f the fact that ^'*’Po largely associates • . •910 with lighter biogenic particles, while Pb seems to have less or no preferences. At the same time, the shallow bottom o f SDB provides a physical lower boundary for organic particle to be recycled than the thermocline in the open ocean. Comparatively, ^'^Pb in SDB is not lost to the water column below thermocline like in the open • 9 1 A 9 1 A • ocean, and this effect will offset the apparent Po/ Pb ratio in SDB. 199 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6.5 Stripping and Trapping of Inside SDB The sediment column profiles indicates EFs in north SDB can be more than 12 times as much as atmospheric input, and 3-5 times higher than those in south SDB (Table 4-1). The differenee cannot be aeeounted for by differenees in riverine runoff 9 1 n because the riverine input o f Pb is negligible (figure 6-3), similar to those observed in similar environments (Appleby and Oldfield, 1992). At the same time, rivers mainly discharge into south and central SDB, thus would not direetly eause elevated levels o f ^^’ ^Pb in north SDB. Rather, major eontributions o f both dissolved and particulate ^’®Pb to SDB are from tidal exehange, and the tidal exchange is through north SDB, so tidal exchange coupled with partiele scavenging should be the 9 10 principal cause for the large exeess Pb aetivities in north SDB sediments. This phenomenon has been simulated successfully by the RS-TRIM model described in chapter 5. At the same time, the box model as shown in figure 6-3 showed that • • 210** • • sedimentary sink o f Pb is indeed 6-7 times higher than the input from atmosphere fallout. The mechanism can be visualized through a conceptual model (Figure 6-5) as follows. Water brought by tidal exehange from outer sea with high dissolved ^'‘ ’ Pb and low TSS mixes with SDB water with low dissolved ^'°Pb but high TSS. TSS remains high during the mixing process due to continuous resuspension from ship 200 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. traffic and tidal currents. High SPM in turn scavenges during the tidal exchange cycle, generating a high sedimentary flux o f Pb in north SDB. This process is 2 10 hereby named “ Pb stripping”. The coastal zone and inner SDB have comparatively low fluxes o f ^^*^Pb stripping due to either low SPM (outer sea) or low dissolved ^'^Pb (inner SDB). Inner Bay North Bay Open Ocean > 9 1 0 Figure 6-5 Conceptual model for “ Pb stripping” in north SDB. Horizontal arrows represent tidal exchange, vertical downward arrows represent the amount o f scavenged (stripped) ^'**Pb, and the size o f the arrows indicates the magnitude. 6.6 Environmental Significance o f Particle Scavenging and Tidal Exchange It is intriguing to note the similarities between particle-reactive radionuclides in this study and many organic and inorganic environmental contaminants. Distribution coefficients Kd for ^'*^Po and ^‘"Pb are in the same range as those o f many organic 210 201 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. contaminants such as PCBs (Rawlings et al, 1998; Turner and Millward, 2002; Erickson, 1997; Pederson et al, 1999). In the relationship between Kd and TSS (equation 5-29), the proportional coefficient B for PCB and ^’^Po are very similar (0.61 vs. 0.64 for PCB and ^^*^Po, respectively). With particulate matter dominating the transport and fate o f both PCBs and particle-reactive radionuclides, it can be easily inferred that the radionuclides in this study can be good proxies for many organic contaminants such as PCBs. Therefore, ^'‘ ^Pb stripping due to tidal exchange and particle scavenging bears important implications for the fate o f many particle- reactive contaminants in SDB. Tidal exchange tends to reduce the levels o f historically deposited contaminants within SDB; it could also introduce particle- reactive chemicals into SDB via the stripping process discussed above. This could become significant as the water column concentrations o f organic and metal contaminants from historical and present discharges continue to decline and the inputs o f wastes to the coastal waters adjacent to SDB continue to rise. Additionally, dumping o f dredged materials from SDB to shallow coastal areas near the bay mouth also potentially increases the availability o f contaminants for stripping. On the other hand, the stripping process as discovered by this study may serve as a barrier that prevents transport o f contaminants outside SDB if the source of contamination is inside SDB. Recent investigations found significant levels o f trace metals in storm runoff in several rivers in central SDB (Chadwick et al, 1999), 202 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. together with very high SPM levels (up to 610 mg/L). Over 60% o f SPM settles within 1 km from the source, potentially scavenging a large portion o f particle- reactive contaminants and preventing them from direct tidal exchange. Spatial distributions o f trace metals clearly showed decreasing trends away from these river mouths (Chadwick et al, 1999), indicating that these trace metals came from storm runoff and were scavenged rapidly to the underlying sediment. At the same time, some heavy metals related to boating activities have higher levels in harbors and marinas in SDB (Mahn et al, 2002), presumably due to higher SPM level resulting from heavy traffic in these locations. The trapping o f the contaminant metals should have prevented them from direct tidal exchange with an average residence time of only 11 days (Largier, 1995). Rather, the beneficial dilution effect through tidal exehange is severely limited by the slow process o f sediment-water interaction with much longer time scale (decades). 6.7 Future Studies This study has produced RS-TRIM, the first numerical geochemical-hydrodynamic model that combines hydrodynamic modeling with geochemical studies o f naturally- occurring, particle-reactive radionuclides. It links the sediment dynamics and radionuclide behavior with the underlying hydrodynamics in a typical coastal environment. However, this study is only the first step towards a more complete geochemical-hydrodynamic model in terms o f functionality, performance. 203 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. application, and accuracy. To this end, future studies are desired in the following areas: 1. The model needs to be more user-friendly in order to be readily used by other researchers. The current model has better readability and better documentation than the original TRIM model, but a lot more work needs to be done. The ideal result would be a graphic user interface (GUI) so that any marine geochemist can run the model simply by inputting his/her set o f geochemical parameters, bathymetry o f the study area, and tidal constants. The computer codes should be transparent to most o f the users; 2. Several aspects o f the geochemical studies need to be strengthened or completed in order to provide better constraints for RS-TRIM model. Many key parameters were measured with relatively large uncertainties, especially those related to ^^"^Th. Besides, river water geochemical studies also needs improvements on sampling coverage and analysis of radionuclides o f different phases. More sediment cores are needed to cover more areas within SDB; 3. The sediment column needs to be simulated in greater detail and capacity in RS-TRIM. The current version only has a ‘virtual sediment colum n’; the communication between this sediment column and overlying water column is unidirectional. A fully coupled sediment column is desired; 204 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4. The model should be tested with other organic and inorganic particle-reactive contaminants. This was one o f the goals o f the original study but it is not fully adequate yet for either PCBs or trace metals; 5. Evaluations o f 3-dimensional hydrodynamic models for upgrading RS-TRIM to 3-dimension need to be made. This requires major efforts, and it is expected that a feasible 3-D geochemical-hydrodynamic model is the ultimate goal for this study despite the difficulties o f application and problems in simulation efficiency. With computing power increasing rapidly, simulation efficiency will no longer be a problem. O f the above goals, item 1 and 3 are the natural extensions o f the current study. They will continue to be pursued after the completion o f the thesis. The others need significantly more effort, and these should be the long-term goals o f further study. 6.8 Summary and Conclusions This thesis work is the first interdisciplinary study that integrates geochemical investigation with numerical hydrodynamic modeling for a coastal oceanic environment. The use o f naturally-occurring, particle-reactive radionuclides as chemical tracers not only enables the studies o f particle dynamics and tidal exchange, but it also leads to a successful modeling o f non-conservative chemical species by the original hydrodynamic model. The key elements that make the integrated 205 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. geochemical-hydrodynamic successful includes 1) comprehensive geochemical studies o f these radionuclides as chemical tracers; 2) seamless ineorporation o f geochemical studies with hydrodynamic model and calibration; 3) successful simulation o f sediment dynamics in the model; 4) successful simulation o f particle- radionuelide interaction in the model. The conclusions that can be made from this study are as follows: • 210 210 234 1. The geochemical studies o f Pb, Po and Th o f sediment and water columns o f San Diego Bay provide complete patterns o f spatial and temporal variations o f these radionuclides in the region; this is also the first such studies in the area, and may be one o f the first studies on shallow, semi-closed, low- inflow embayments. It was found that particulate matter and particle scavenging determine the spatial distribution and temporal variation o f the particle-reactive radionuclides. 9 1 0 2. High-enriehment factors o f Pb with regard to its atmospheric input in north San Diego Bay suggest a combined effect o f both tidal exehange and particle 91 n scavenging through a ‘stripping’ process, in which dissolved Pb supplied from outer sea through tidal exchange is quickly scavenged and settled to the sediment in north San Diego Bay, causing abnormally high ^’^ ’ Pb inventory in the sediment column. The enrichment factors are among the highest ever reported. 206 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 91 n 01 n 3. Po was repeatedly found to have higher aetivities than its grandparent Pb in particulate forms and lower activities than Pb in dissolved form. This is most probably due to active regeneration o f ^’°Po-rich organic particles in the lower water column. This pattern supports the notion that ^'^Po has higher 91 0 affinity to planktons and organic particles, while Pb is more affiliated with inorganic particles that settle faster. In general, ^’* ’Po has a higher particle reactivity than ^"’Pb. 4. The use o f a depth-averaged hydrodynamic model in this study is a successful. It is a pioneering approach to the geochemical-hydrodynamic studies o f dynamic aqueous environment. By relating geochemical processes with hydrodynamics o f the aqueous environment, parameters o f chemical equilibrium and reaction kinetics that controls geochemical processes are no longer unknown or ambiguous due to the lack o f hydrodynamic constraints. 5. The combination o f the hydrodynamic and geochemical studies not only confirmed important discoveries made in the geochemical investigation, but also provided long-term calibration for the model in terms o f sediment column profile and inventory o f radionuclides. 207 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Abbott, P.L. 1999, The rise and fall o f San Diego: 150 million years o f history recorded in sedimentary rocks. Sunbelt Publications, Inc., 231 pp Aller, R., and T.K. Cochran, 1976, U/ U disequilibrium in nearshore sediments: particle reworking and diagenetic time scales. Earth and Planetary Science Letters, 29, 37-50 Aller, R., 2001, Transport and Reactions in the Bioirrigated Zone, in: The Benthic Boundary Layer: Transport Processes and Biogeochemistry, Boudreau, B.P. and B.B. Jorgensen (eds.), Oxford University Press, 269-301 Appleby, P.O., and F. Oldfield, 1992, Application o f lead-210 to sedimentation studies, in: Uranium-series Disequilibrium: Applications to Earth, Marine, and Environmental Sciences, Second Edition, M. Ivanovich and R.S. Harmon (eds.), Oxford University Press, 731-778 Bacon, M.P., R.A. Belastock, M. Tecotzky, K.K. Turekian, and K.W. Spencer, 1988, Lead-210 and Polonium-210 in ocean water profiles o f the continental shelf and slope south o f New England, Continental Shelf Research 8, Nos. 5-7, 841-453. Baker, J. E., P. D. Capel and S. J. Eisenreieh. 1986. Influence o f colloids on sediment-water partition coefficients o f polychlorobiphenyl congeners in natural waters. Environmental Science and Technology 20: 1136-1164. Baskaran, M., M. Ravichandran and T.S. Bianchi, 1997, Cycling o f ’Be and ^"^Pb in a high DOC, shallow, turbid estuary o f south-east Texas, Estuarine, Coastal and Shelf Science 1997 (45), 165-176 Baskaran, M., P. H. Santschi, G. Benoit and B. D. Honeyman. 1992. Scavenging of thorium isotopes by colloids in seawater o f the G ulf o f Mexico. Geochimica et Cosmochimica Acta 56: 3375-3388. 208 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Baskaran M., and P.H. Santschi, 1993, The role o f particles and colloids in the transport o f radionuclides in coastal environments o f Texas, Marine Chemistry 43(1993), 95-114 Benninger, L. K. 1976, The uranium-series radionuelides as tracers o f geoehemical proeesses in Long Island Sound. Unpublished Ph.D. Thesis, Yale University, New Heaven, CT. 91 0 Benninger, L.K. 1978, Pb balance in Long Island Sound, Geoehimiea et Cosmoehimiea Aeta, 42, 1165-1174 Benoit, G., 1988, The biogeochemistry o f ^'*’Pb and ^*°Po in fresh waters and sediments, Ph.D thesis. Woods Hole Oceanographic Institution, M assachusetts Institute o f Teehnology, 1988 Bonner, J.S., A.N. Ernest, R.L. Auterieth, and S.L. Dueharme, 1994, Parameterizing Models for Contaminated Sediment Transport, in: Transport and Transformation of Contaminants Near the Sediment-Water Interface, DePinto, J.V et al.(eds.), CRC Press, 281-305 Broecker, W.S., A. Kaufman and R.M. Trier, 1973, The residence time o f thorium in surface sea water and its implication regarding the rate of reaetive pollutants. Earth and Planetary Seienee Letters 20, 35-44 Broeeker, W. S. and T.-H. Peng. 1982. Traeers in the Seas. Eldigio Press. New York, New York, USA. Brownawell, B.J., 1986, The role o f eolloidal organie matter in the marine geochemistry o f PCBs, Ph.D thesis, Woods Hole Oeeanographic Institution, Massaehusetts Institute o f Technology, 1986 Bruland, K.W., M. Koide, and E.D. Goldberg, 1974, The comparative marine geochemistries o f Lead 210 and Radium 226, Journal o f Geophysieal Research, V ol.79,N o.21, 3083-3086 209 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Carlucci, A.F., R.W. Eppley, and J.R. Beers, 1986, Introduction to the Southern California Bight, in: Lecture Notes on Coastal and Estuarine Studies Vol. 15, Plankton Dynamics o f the Southern California Bight, Eppley, R.W. (ed.). Springer-Verlag, ppl-12 9 1 0 Carpenter R., Bennet J.T. and Peterson M.L. 1981, Pb Activities in and fluxes to sediments o f the W ashington continental slope and shelf, Geochimica et Cosmochimica Acta 45, 1155-1172 Carpenter & Capone, eds (1983) Nitrogen in the Marine Environment, Academic Press Casulli, V., 1990a, Semi-implicit finite-difference methods for the two-dimensional shallow water equation. Journal o f Computational Physics 86, 56-74 Casulli, V., 1990b, Numerical simulation o f shallow water flow, in: Computational Methods in Surface W ater Hydrology, Gambolati, G., Rinaldo, A., Brebbia, C.A., Gray, W.G. and Pinder, G.F. (eds.), Springer-Verlag, Berlin, pp 13-22 Chadwick, B., J. Leather, K. Richter, S. Apitz, D. Lapota, D. Duckworth, C. Katz, V. Kirtay, B. Davidson, A. Patterson, P. Wang, S. Curtis, G. Key, S. Steinert, G. Rosen, M. Caballero, J. Groves, G. Koon, A. Yalkirs, K. M eyers-Schulte, M. Stallard, S. Clawson, R. Streib Montee, D. Sutton, L. Skiimer, J. Germano, R. Cheng, 1999, Sediment Quality Characterization, Naval Station San Diego, Final Summary Report, 1999 Chadwick, D.B., J.L. Largier and R.T. Cheng, 1995, San Diego Bay Circulation: Appendix 2, The role o f thermal stratification in tidal exchange at the mouth o f San Diego Bay, prepared for the California State Water Resourees Control Board and the California Regional Water Quality Control Board, San Diego Region, July 1995 Cheng, R.T., 1983, Eulerian-Lagrangian computations in estuarine hydrodynamics. Third International Conference on Numerical Methods in Laminar and Turbulent Flow, Taylor, C., Johnson, J.A. and Smith, W.R. (eds.), Pineridge Press, pp.341-352 210 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Cheng, R.T., Casulli, V. and Milford, S.N., 1984, Eulerian-Lagrangian solution o f the convection-dispersion equation in natural coordinates. Water Resources Research 20, 7, 944-952 Cheng, R. T., V. Casulli, and J. W. Gartner, 1993, Tidal, residual, intertidal mudflat (TRIM) model and its applications to San Francisco Bay, California, Estuarine, Coastal, and Shelf Science, Vol. 36, p. 235-280 Cheng, R.T., V. Casulli, and S.N. Milford, 1984, Eulerian-Lagrangian solution o f the convection-dispersion equation in natural coordinates. Water Resources Research, Yol.20, No.7, 944-952 Cherry, R.D., S.W. Fowler, T.M. Beasley, M. Heyraud, 1975, Polonium-210: its vertical oceanic transport by zooplankton metabolic activity. Marine Chemistry 3, 105-110 Cherry, R.D. and M. Heyraud, 1982, Evidence o f high natural radiation doses in certain mid-water ocean organisms. Science 218, 54-56 Casulli, V. and R.T. Cheng, 1992, A semi-implicit finite difference model for three- dimensional tidal circulation, in: Estuarine and coastal modeling, proceedings o f the 2nd International Conference, M.L. Spaulding et al (eds.), 1992, published by American Society o f Civil Engineers, 620-632 Chow, V.T., 1959, Open Channel Hydraulics, McGraw-Hill, New York, 680pp Cherry, R.D. and M. Heyraud, 1979, Polonium-210 and lead-210 in marine food chains, Marine Biology 52(1979) 227-236 Cochran, J.K., 1992, The oceanic chemistry o f the uranium- and thorium-series nuclides, in: Uranium-series Disequilibrium: Applications to Earth, Marine, and Environmental Sciences, Second Edition, M. Ivanovich and R.S. Harmon (eds.), Oxford University Press, 334— 395 211 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Crean, P.B., T.S. Murty, and J.A. Stronach, 1988, Mathematical M odelling o f Tides and Estuarine Circulation, in: Lecture Notes on Coastal and Estuarine Studies, Bowman, M.J. et al (eds.), Springer-Verlag, 464 pages Dade, W.B., A.J. Hogg, and B.P. Boudreau, 2001, Physics o f Flow Above the Sediment-Water Interface, in: The Benthic Boundary Layer: Transport Processes and Biogeoehemistry, Boudreau, B.P. and B.B. Jorgensen (eds.), Oxford University Press, 4-43 DePinto, J.V., T.L. Theis, T.C. Young, and S.T. Leach, 1994, Predicting Metals Partitioning During Resuspension Events, in: Transport and Transformation of Contaminants N ear the Sediment-Water Interface, DePinto, J.Y et al.(eds.), CRC Press, 177-204 Dortch, M., C. Ruiz, T. Gerald, R. Hall, 1998, Three-dimensional contaminant transport/fate model, 5th International Conference on Estuarine and Coastal Modeling, Published by ASCE, 75-89 Dronkers, J.J., 1969 Tidal computations for rivers, coastal areas and seas. Journal o f Hydraulics Division, ASCE, Vol. 95, N. H Y l, pp 44-77 Dyke, P.P.O., 2001, Coastal and Shelf Sea Modelling, in: Topics in Environmental Fluid Mechanics, Chatwin, P. et al (eds.), Kluwer Academic Publishers, 249 pages Feng, S., R.T. Cheng, P. Xi, 1986a, On tide-induced residual current and residual transport, 1. Lagrangian residual current. Water Resource Research 22, 1623-1634 Feng, S., R.T.Cheng, P.Xi, 1986b, On tide-induced residual current and residual transport, 2. Residual transport with application in South San Francisco Bay, California, Water Resource Research 22, 1635-1646 Erickson, M.D., 1997, Analytical Chemistry o f PCBs, 2nd Edition, CRC Press, pp 18- 96 212 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fairey, R., J. Downing, C. Roberts, E. Landrau, J. Hunt, B. Anderson, C. J. Wilson, G. Kamahi, F. LaCaro, P. Michael, M. Stephenson and M. Puckett. 1998. Chemistry, Toxicity and Benthic Community Conditions in Sediments o f the San Diego Region, Final Addendum Report. San Diego, CA. Fisher, N. S., J. K. Cochran, S. Krishnaswami and H. D. Livingston, 1988, Predicting the oceanic flux o f radionuclides on sinking biogenic debris. Nature 335: 622-625. Fowler, S.W., 1991. Biologically mediated removal, transformation, and regeneration o f dissolved elements and compounds. In: Mantoura, R.F.C., Martin, J.M., Wollast, R. (eds.). Ocean Margin Processes in Global Change. Wiley, New York, 127-143 Fuller, C. C. 1982, The use o f ^*‘ ^Pb, ^^"^Th, and '^’Cs as tracers o f sedimentary processes in San Francisco Bay, California. Unpublished Masters Thesis, University o f Southern California, Los Angeles, CA. Fairey, R., Downing, J., Roberts, C., Landrau, Eli, Hunt, J., Anderson, B., Wilson, C.J., Kapahi, G., LaCaro, F., Michael, P., Stephenson, Mark, and Puckett, M., 1998, Chemistry, Toxicity and Benthic Community Conditions in Sediments o f the San Diego Region, Final Addendum Report, September 30, 1998 Fairey, R., C. Roberts, M., Jacobi, S. Lamerdin, R. Clark, J. Downing, E. Long, J. Hunt, Anderson, B., J. Newman, R. Tjeerdema, M. Stephenson and C. Wilson, 1998. Assessment o f Sediment toxicity and chemical concentrations in the San Diego Bay region, California, USA. Environmental Toxicology and Chemistry, 17:1570-1581 Fuller, C.C., A. van Geen, M. Baskaran, R. Anima, 1999, Sediment Chronology in San Francisco Bay, California, defined by ^’* ’Pb, ^^"^Th, ’^^Cs and Marine Chemistry 64(1999), 7-27 Gerald, C.F. and Wheatley, P.O., 1994, Applied Numerical Analysis, fifth edition, Addition-Wesley Publishing Company, 616-662 Hammer, M.J., and M.J. Hammer, Jr., 2001, Water and Wastewater Technology, Fourth Edition, Prentice Hall, 2001, pp 121-122 213 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Harada, K. and S. Tsimogai, 1982, and in settling particles in Funka Bay: Does lead dissolve at the bottom? Proceedings o f Annual Meeting o f Geochemistry Society o f Japan, Ryukyu University, 1982. Helz, G.R., Setlock, G.H., Cantillo, A. Yl, and Moore, W.S., 1985, Processes controlling the regional distribution o f ^'*^Pb, ^^^Ra and anthropogenic zinc in estuarine sediments. Earth and Planetary Science Letters, 76, 23-24 Hill, P.S. and l.N. McCave, 2001, Suspended Particle Transport in Benthic Boundary Layers, in: The Benthic Boundary Layer: Transport Processes and Biogeochemistry, Boudreau, B.P. and B.B. Jorgensen (eds.), Oxford University Press, 78-103 Hummel, W., 1997, Binding models for humie substances, in: Modeling in Aquatic Chemistry Chapter V, Grenthe, I. And Puigdomenech, 1. (eds.), OLCD-OCDL Paris, pp 153-206 Hyman, M., J. Rohr, J. Schoonmaker, T. Ratcliffe, B. Chadwick, K.Richter, S.Jenkins, and J.Waysl, 1995, Mixing in the wake of an aircraft carrier, in: Proceedings o f Oceans ’95, Scripps Institute o f Oceanography, University of California, La Jolla, California Jackson, G.A., 1986, Physical Oceanography o f the Southern California Bight, in: Lecture Notes on Coastal and Estuarine Studies Vol. 15, Plankton Dynamics o f the Southern California Bight, Lppley, R.W. (ed.). Springer-Verlag, ppl3-52 James, I.D., 2002, Modelling pollution dispersion, the ecosystem and water quality in coastal waters: a review. Environmental Modeling and Software 17(2002) 363-385 Jeremiason, J.D., S.J. Lisenreich, J.L. Baker, and B.J. Ladie, 1998, PCB decline in settling particles and benthic recycling o f PCBs and PAHs in Lake Superior, Environmental Science and Technology, 32, 3249-3256 Kadko, D., 1993, Excess Po and nutrient recycling within the California Coastal Transition Zone, Journal o f Geophysical Research, Vol. 98, No. C l, pp 857-864 214 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Kaliakin, V.N., 2002, Introduction to Approximate Solution Techniques, Numerical Modeling, and Finite Element Methods, Marcel Dekker, Inc., 39-76 Kaufman, A., Y.-H. Li and K. K. Turekian. 1981. The removal rates o f ^^"^Th and 228Th fj-om waters o f the New York Bight. Earth and Planetary Science Letters 54: 385-392. Kim, G., 2001, Large deficiency o f polonium in the oligotrophic ocean’s interior. Earth and Planetary Science Letters, 192(2001) 15-21 Kharkar, D.P., J. Thomson, K.K. Turekian, W.O. Foster, 1976. Uranium and thorium decays series nuclides in plankton from the Caribbean, Limnology and Oceanography 21,294-299 Kreyszig, L., 1988, Advanced Engineering Mathematics, Sixth Edition, John Wiley & Sons, 1294 pages Krishnaswami, S., Somayajulu, B.L.K., and Chung, Y. , 1975, Pb-210/Ra-226 disequilibrium in the Santa Barbara Basin. Earth and Planetary Science Letters, 27, 388-392 Krone, R.B., 1962, Flume studies o f the transport o f sediment in estuarine shoaling processes, Final Report to San Francisco District US Army Corps o f Engineers, Washington DC. University o f California, Berkeley Ku, T.-L., K. G. Knauss and G. G. Matheiu. 1977. Uranium in open oeean: concentration and isotopic composition. Deep Sea Research 24: 1005-1017. Langmuir, D., 1978, Uranium solution-mineral equilibria at low temperatures with applications to sedimentary ore deposits. Geochimica et Cosmochimica Acta, 42, 547-569 215 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lang, G., R. Schubert, M. Markofsky, H.-U. Fanger, I. Grabemann, H.L. Drasemann, L J.R . Neumann, and R. Riethmuller, 1989, Data interpretation and numerical modeling o f the M ud and Suspended Sediment Experiment 1985, Journal o f Geophysical Research, Vol. 94, No.CIO, pp 14,381-14,395 Largier, J.L., 1995. San Diego Bay Circulation - A study of the circulation of water in San Diego Bay for the purpose o f assessing, monitoring and managing the transport and potential accumulation o f pollutants and sediment in San Diego Bay. Final report to; The California State W ater Resources Control Board and the California Regional Water Quality Control Board, San Diego Region, 1995. Largier, J.L., C.J. Hearn, and D.B. Chadwick, 1996, Density structures in “low inflow estuaries”, in: Buoyaney effects on coastal estuarine dynamics. Coastal Estuarine Studies, Volume 53, edited by D.C. Aubrey and C.T. Friedrichs, pp227-242, American Geophysical Union, Washington, D.C. Largier, J.L., J.T. Hollibaugh, and S.V. Smith, Seasonally hypersaline estuaries in Mediterranean-climate region, Estuarine, Coastal and Shelf Science (1997) 45, 789- 797 Le Norman, C., E. Pettier, and C. Teisson, 1998, Three dimensional modeling of cohesive sediment transport in estuaries, in: Physics o f Estuaries and Coastal Seas, Dronkers and Scheffers (eds.), 1998 Balkema, Rotterdam, the Netherlands Lick, W., 1994, The Flocculation, Deposition, and Resuspension o f Fine-Grained Sediments, in: Transport and Transformation o f Contaminants Near the Sediment- Water Interface, DePinto, J.V et al. (eds.), CRC Press, 35-58 Liu, Y.P., G.E. Millward, and J.R.W. Harris, 1998, Modeling the distributions of dissolved Zn and Ni in the Tamar Estuary using hydrodynamics coupled with chemical kinetics. Estuaries, Coastal and Shelf Sciences (1998) 47, 535-546 Lueck, R.G. and Y. Lu, 1997, The logarithmic layer in a tidal channel. Continental Shelf Research, Vol. 17, N o.14, ppl785-1801 216 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Lumborg, U., and Windelin, A., 2003, Hydrography and cohesive sediment modeling: applieation to the Romo Dyb tidal area, Journal o f Marine Systems 38(2003) 287-303 Mahn, C., J. Gieskes, Rivera-Duarte, L, and B. Chadwick, 2002, Trace Element Distributions in San Diego Bay: Copper, Zinc, Manganese, and the Rare Earth Elements, EOS Transactions (abstract) for the AGU 2002 Fall Meeting, Vol. 83, No. 47, November 19, 2002 Martin, J.L., and S.C. McCutcheon, 1999, Hydrodynamics and Transport for Water Quality Modeling, Lewis Publishers, 794pp Martin, J.M., Meybeck, M., and Fusset, M., 1978 a. Uranium and Thorium in the Zaire Estuary. Netherlands Journal o f Sea Research, 12(3-4), 338-344 Martin, J.M., Jijampurkar, V.N., and Salvadori, P., 1978b, Uranium and thorium isotopes behaviour in estuarine systems, in: Biogeoehemistry o f estuarine sediments, Goldberg E.D. (ed), pp 111-127. UNESCO, Paris. Masque, P., J.A. Sanchez-Cabeza, J.M. Bruach, E. Palacios, and M.Canals, 2002, Balance and residenee times o f Pb-210 and Po-210 in surface waters o f the northwestern Mediterranean Sea, Continental Shelf Research 22(2002) 2127-2146 Massel, S.R., 1989, Hydrodynamics o f Coastal Zones, Elsevier Oceanography Series, 48, Elsevier, 1989, 279-284 Massel, S.R., 1999, Fluid Mechanics for Marine Ecologists, Springer, 253-304 McCain, B. B., S. L. Chan, M. M. Krahn, D. W. Brown, M. S. Myers, J. T. Landahl, S. Pierce, R. C. Clark, Jr. and U. Varanasi. 1992. Chemical contamination and associated fish diseases in San Diego Bay. Environmental Science and Technology 26: 725-733. McDonald, E.T. and R.T.Cheng, 1994, Issues related to modeling the transport o f suspended sediments in northern San Francisco Bay, California, in: Estuarine and Coastal Modeling III, Proceedings o f the 3rd International Conference, M.L. Spaulding et al (eds.), 1994, 551-565 217 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Moijeld, H.O., 1988, Depth dependence o f bottom stress and quadratic drag coefficient for barotropic pressure-driven currents, Journal o f Physical Oceanography 18, 1658-1669 Moore, H.E., S.E.Poet, E.A.Martell, L.H.W ilkening, 1974, Origin o f Rn-222 and its long-lived daughters in air over Hawaii, Journal o f Geophysical Research 79, 5019- 5024 Moore, W.S. and J. Dymond, 1988, Correlation o f ^’* ’Pb removal with organic carbon fluxes in the Pacific Ocean, Nature 331, 339-341 Needham, M.V., 1983, San Diego Bay area fisheries management, in: Proceedings o f the 3rd Symposium on Coastal and Ocean Management, pp.496-515, June 1-4, San Diego, California Nozaki, Y., Dobashi, P., Kato, Y., and Yamamoto, Y., 1998, Distribution o f Ra isotopes and the 210Pb and 210Po in surface seawaters o f the mid Northern Hemisphere, Deep-Sea Research I, 45 (1998), 1263-1284 Nozaki, Y., Zhang, J., and Takeda, A., 1997, '°Pb and ^'**Po in the equatorial Pacific and the Bering Sea: the effects o f biological productivity and boundary scavenging, Deep-Sea Research II, Vol. 44, No. 9-10, pp 2203-2220, 1997 Nozaki, Y., I. Naodo, M.Yahima, 1990, Unusually large ^’'’Po deficiencies relative to 210Pb in the Kuroshio Current o f the East China and Philippines Seas, Journal o f Geophysical Research, 95(1990) 5321-5329 Part-Enander and Sjoberg, 1999, The MATLAB 5 Handbook, Eddison Wesley, Chinese Translation by Wang, Y. et al, China Machine Press, 384 pages Partheniades, E. 1962, A study o f erosion and deposition o f cohesive soils in salt water, PhD. Thesis, University o f California, Berkeley 218 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Partheniades, E., 1965, Erosion and deposition o f cohesive soils, Journal o f the Hydraulics Division, ASCE, Proceedings, Vol. 91, No. H Y l: 105-139 Partheniades, E., 1993, Turbulence, Flocculation and Cohesive Sediment Dynamics, in: Nearshore and Estuarine Cohesive Sediment Transport, Coastal and Estuarine Studies Vol.42, A.J. M ehta (ed.), American Geophysical Union, Washington, D.C., 40-59 Peng, J., Zeng, E., Ku, T.-L. and Luo, S., 2002, Sediment resuspension in San Diego Bay and its significance to redistribution and dilution o f particle-affiliated contaminants. Southern California Coastal W ater Research Project (SCCWRP) Annual Report 2002 Piedra-Cueva I., and M.Mory, 2001, Erosion o f a deposited layer o f cohesive sediment, in: Coastal and Estuarine Fine Sediment Processes, W.H. McAnally and A.J. M ehta (eds.), Elsevier Science, 2001, 41-51 Potter, M.C. and Wiggert, D.C., 1997, M echanics o f Fluids, second edition, Prentice Hall, 1997, 286-292 Quigley, M.S., P.H. Santschi, L. Guo and B.D. Honeyman, 2001, Sorption irreversibility and coagulation behavior o f 234Th with marine organic matter. Marine Chemistry 76 (2001) 27-45 Rawling, M. C., Turner, A., and Tyler, A. O., 1998, Partiele-water interactions o f 2, 2’, 5, 5’-tetrahydrochlorobiphenyl under simulated estuarine conditions. Marine Chemistry 61,115-126 Robbins, J.A., 1978, Geochemical and geophysical applications o f radioactive lead, in: The biogeoehemistry o f lead in the environment, Elsevier/North Holland Biomedical Press, 1978, 285-393 Sabersky, R.H., A.J. Acosta, E.G. Hauptmann, and E.M. Gates, 1999, Fluid Flow: a first course in fluid mechanics, fourth edition, Prentice Hall, pp448-449 219 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. San Diego Bay Interagency Water Quality Panel, 1998. World Wide Web Site: http ://sdbav. sdsc.edu/html/modeling2. html. Santschi, P.H., D.Adler, M.Amdurer, Y-H. Li and J.J. Bell, 1980, Thorium isotopes as analogues for “particle-reactive” pollutants in coastal marine environments, Earth and Planetary Science Letters, 47, 327-335 Santschi, P.H., L. Guo, M. Baskaran, S. Trumbore, J. Southon, T.S. Bianchi, B. Honeyman, and L. Cifuentes, 1995, Isotopic evidence for the contemporary origin o f high-molecular weight organic matter in oceanic environments, Geochimica et Cosmochimica Acta, Vol. 59, No. 3, 625-631 Santschi, P.H., L. Guo, S. Asbill, M. Allison, A. B. Kepple, L.-S. Wen, 2001, Accumulation rates and sources o f sediments and organic carbon on the Palos Verdes shelf based on radioisotopic tracers ’^^Cs, ^'V b, ^^"*Th, and '"^C, Marine Chemistry 73, 125-152 Sanudo-Wilhelmy, S. A., Rivera-Duarte, 1. And Flegal, A.R., 1996, Distribution o f colloidal trace metals in the San Francisco Bay Estuary, Geochimica et Cosmochimica Acta 60, 1259-1282 • 1 9 T S Sarin, M.M., S. Krishnaswami, R.Ramesh, B.L.K. Somayajulu, U decay series nuclides in the northeastern Arabian Sea: scavenging rates and cycling processes. Continental Shelf Research, 14 (1994), 251-265 Schroeder, R.A., 1994, Point-source identification from PCB analyses o f surficial sediment in San Diego Bay, California, Presentation at the 1994 Ocean Science Meeting, San Diego, California Shannon, L.V., Cherry, R.D., and Orren, M.M., 1970. Polonium-210 and lead-210 in the marine environment. Geochimica et Cosmochimica Acta, 34, 701-711 Senate Committee on Toxics and Public Safety Management, 1988, Hearings on San Diego Bay Toxic Contamination, Senate Committee on Toxics and Public Safety Management. 220 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Shannon , L.V., R.D.Cherry, M.M. Orren, 1970, Polonium-210 and Lead-210 in the marine environment, Geochimica et Cosmochimica Acta 34, 701-711 Shaw, C.T., 1992, Using Computational Fluid Dynamics, Prentice Hall, 251 pages Sheng, Y.P., Eliason, D.E., and Chen, X.-J., 1992, Modeling three-dimensional circulation and sediment transport in lakes and estuaries, in: Estuarine and Coastal Modeling, Proceedings o f the 2nd International Conference, Spaulding, M.L. et al. (eds.), 1992, Published by ASCE Sheng, Y.P., 1996, Pollutant Load Reduction Models for Estuaries, in: Estuarine and Coastal Modeling, Proceedings o f the 4th International Conference, Spaulding, M.L. and R.T. Cheng (eds.), 1996, Published by ASCE Sternberg, R.W., I.Berhane, A.S.Ogston, 1999, Measurement o f size and settling velocity of suspended aggregates on the northern California continental shelf. Marine Geology 154(1999), 43-53 Taki, K., 2001. Critical shear stress for cohesive sediment transport, in: Coastal and Estuarine Fine Sediment Processes, W.H.McAnally and A.J. M ehta (eds.), ppl73- 187, Proceedings in Marine Sciences Series 3, Elsevier Science, 2001 Tanaka N., Takeda Y., and Tsunogai S., (1983) Biological effect on removal o f Th- 234, Po-210 and Pb-210 from surface water in Funka Bay, Japan, Geochimica et Cosmochimica Acta 47, 1783-1790 Tang, L., and E.E. Adams, 1998, Interfacing hydrodynamics and water quality models with the Eulerian-Lagrangian Method, 5th International Conference on Estuarine and Coastal Modeling, Spaulding, M.L., and Cheng, R.T. (eds.). Published by ASCE, 153-165 Tappin, A.D., J.D Burton, G.E. Millward, P.J. Statham, 1997, A numerical transport model for predicting the distributions o f Cd, Cu, Ni, Pb and Zn in the southern North Sea: the sensitivity o f model results to the uncertainties in the magnitudes o f metal inputs, Journal o f Marine Systems 13 (1997) 173-204 221 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Torfs, H., J. Jiang, and A.J. Mehta, 2001, Assessment o f the erodibility o f fine/coarse sediment mixtures, in: Coastal and Estuarine Fine Sediment Processes, W.H.McAnally and A.J. Mehta (eds.), Elsevier Sciences B.V., 109-123 Turekian, K.K., Y.Nozaki, E.K.1977, Geochemistry o f atmospheric radon and radon products. Annual Review o f Earth and Planetary Sciences, 227-255 Turekian, K.K., D.P.Kharkar, and J.Thomson, 1974. The fates o f Pb-210 and Po-210 in the ocean surface. Journal Des Researches Atmospheriques 8, 639-646 Turekian K.K., 1977, The fate o f trace metals in the ocean, Geochimica et Cosmochimica Acta 42, 1139-1144 Turner, A. and G.E. Millward, 2002, Suspended particles: their role in estuarine biogeochemical cycles, Estuarine, Coastal and Shelf Science (2002), 55, 857-883 Van de Kreeke, J., 1988, Hydrodynamics o f tidal inlets, in: Lecture Notes on Coastal and Estuarine Studies, Vol.29: Hydrodynamics and Sediment Dynamics o f Tidal Inlets, D.G. Aubrey and L.W eishar (eds.). Springer-Verlag New York, Inc., 1988, 1- 23 van Meter, P.C. and J.T.Wilson, 1997, Similar rates o f decrease o f persistent, hydrophobic and particle-reactive contaminants in riverine systems. Environmental Science and Technology, 32, No. 21, 3312-3317 van Rijn, L.C., 1989, Handbook-Sediment transport by currents and waves. Report H461, Delft Hydraulics, Delft, the Netherlands Wang, X.H., 2002, Tidal-induced sediment resuspension and the bottom boundary layer in an idealized estuary with a muddy bed. Journal o f Physical Oceanography 32(11), 3113-3131 222 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Wang, J.; N. Foster, S. Armalis, D. Larson, A. Zirino, and K. Olsen, 1995, Remote stripping electrode for in situ monitoring o f labile copper in the marine environment, Analytica Chimica Acta (1995) 223-231. Wang, P.P., R.T.Cheng, K.Richter, E.S.Gross, D.Sutton and J.W .Gartner, 1998, Modeling tidal hydrodynamics o f San Diego Bay, California, Journal o f the American Water Resources Association, Vol.34, N o.5, 1123-1140 Williams, P.M., 1986, Chemistry o f the Dissolved and Particulate Phases in the Water Column, in: Lecture Notes on Coastal and Estuarine Studies Vol. 15, Plankton Dynamics o f the Southern California Bight, Eppley, R.W. (ed.). Springer-Verlag, pp 53-83 Xu, M., 2000, Fortran PowerStation Version 4.0, Tsinghua Press (in Chinese), 341 pages Yen, T.-F., 1999, Environmental Chemistry: Essentials o f Chemistry for Engineering Practice, Prentice Flail, 762 pages Young, D.R. and T.C.Heesen, 1974, Inputs and distributions o f chlorinated hydrocarbons in three Southern California harbors. Technical Report 36. Southern California Coastal Water Researeh Project, El Segundo, CA Zeng, E. and A. Khan. 1994. Preliminary study o f seasonal variation o f carbon and nitrogen in sediments off Point Loma. in: J. N. Cross , C. Francisco and D. Hallock (Eds.), Southern California Coastal Water Research Project Annual Report 1992-93. Southern California Coastal Water Research Project, Westminster, CA, USA. pp. 91- 99 Zeng, E.Y., J. Peng, D. Tsukada, and T.-L. Ku, 2002, In situ measurements o f chlorinated biphenyls in the waters o f San Diego Bay, California, Environmental Science and Technology 2002, 36, 4975-4980 223 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. Zeng, E. Y., C.C. Yu, and K. Tran, 1999, In situ measurements o f chlorinated hydrocarbons in the water column off the Palos Verdes Peninsula, California, Environmental Science and Technology, 33, 392-398 Zhu, C., and Anderson, G., 2002, Environmental Applications o f Geochemical Modeling, Cambridge University Press, 284 pages 224 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission. 7) CD ■o o Q . C o CD Q . ■o CD ( / ) o' 3 Appendix I Data Summary for All Water Samples Summer 1999 Water Samples O O ■o c q ' O ’ CD CD ■ D O Q . C 9- o o ■ o o CD Q . T D CD (/) C/) to Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t 1 2 8 6/25/1999 n/a n/a 0.027 + 0.002 n/a n/a 0.096 ± 0.003 n/a n/a 1 2 9 6/25/1999 n/a n/a 0.043 + 0.002 n/a n/a 0.141 i 0.006 n/a n/a 1_30 6/25/1999 n/a n/a 0.035 + 0.002 n/a n/a 0.091 + 0.005 n/a n/a 1 3 1 6/29/1999 n/a n/a 0.035 ± 0.003 n/a n/a 0.079 + 0.003 n/a n/a a 1-26 6/25/1999 n/a n/a 0.034 + 0.003 n/a n/a 0.092 ± 0.004 n/a n/a al-27 6/25/1999 n/a n/a 0.033 + 0.003 n/a n/a 0.099 + 0.004 n/a n/a 2__22 6/25/1999 n/a n/a 0.029 ± 0.002 n/a n/a 0.054 + 0.003 n/a n/a 2_23 6/25/1999 n/a n/a 0.053 ± 0.003 n/a n/a 0.074 + 0.003 n/a n/a a2-24 6/25/1999 n/a n/a 0.031 ± 0.002 n/a n/a 0.065 + 0.004 n/a n/a a2-25 6/25/1999 n/a n/a 0.049 + 0.002 n/a n/a 0.084 + 0.005 n/a n/a 3_32 6/29/1999 n/a n/a 0.024 ± 0.002 n/a n/a 0.084 + 0.006 n/a n/a 3_33 6/29/1999 n/a n/a 0.026 + 0.002 n/a n/a 0.039 + 0.003 n/a n/a a3-34 6/29/1999 n/a n/a 0.021 ± 0.002 n/a n/a 0.030 + 0.002 n/a n/a a3-36 6/29/1999 n/a n/a 0.018 + 0.001 n/a n/a 0.039 + 0.002 n/a n/a 4_17 6/25/1999 n/a n/a 0.020 + 0.002 n/a n/a 0.034 ± 0.001 n/a n/a 4 1 8 6/25/1999 n/a n/a 0.017 + 0.002 n/a n/a 0.047 + 0.003 n/a n/a a4-19 6/25/1999 n/a n/a 0.020 + 0.002 n/a n/a 0.031 + 0.001 n/a n/a a4-20 6/25/1999 n/a n/a 0.006 ± 0.001 n/a n/a 0.030 ± 0.002 n/a n/a a4-21 6/25/1999 n/a n/a 0.016 + 0.001 n/a n/a 0.027 + 0.001 n/a n/a 5 1 5 6/25/1999 n/a n/a n/a n/a n/a 0.029 + 0.004 n/a n/a 5_16 6/25/1999 n/a n/a 0.019 ± 0.002 n/a n/a 0.031 + 0.003 n/a n/a a5-13 6/18/1999 n/a n/a 0.017 + 0.001 n/a n/a n/a n/a n/a a5-14 6/18/1999 n/a n/a 0.010 + 0.002 n/a n/a 0.042 ± 0.010 n/a n/a 6_10 6/18/1999 n/a n/a 0.017 + 0.001 n/a n/a 0.025 + 0.002 n/a n/a 6 1 1 6/18/1999 n/a n/a 0.022 ± 0.002 n/a n/a n/a n/a n/a a6-12 6/18/1999 n/a n/a 0.018 + 0.001 n/a n/a 0.025 + 0.001 n/a n/a 7_8 6/18/1999 n/a n/a 0.007 + 0.001 n/a n/a 0.034 ± 0.007 n/a n/a 7 3 CD ■ D O Q . C o CD Q . ■ D CD ( f i W o' o o o o ■ D < q' Summer 1999 Water Samples (continued) Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t 7_9 6/18/1999 n/a n/a 0.014 + 0.002 n/a n/a 0.017 ± 0.001 n/a n/a a7-5 6/15/1999 n/a n/a 0.017 + 0.001 n/a n/a 0.054 + 0.008 n/a n/a a7-6 6/15/1999 n/a n/a 0.016 + 0.002 n/a n/a 0.027 + 0.002 n/a n/a a7-7 6/15/1999 n/a n/a 0.015 ± 0.001 n/a n/a 0.015 + 0.002 n/a n/a 8 2 6/15/1999 n/a n/a 0.030 + 0.002 n/a n/a 0.226 ± 0.023 n/a n/a 8 4 6/15/1999 n/a n/a 0.018 + 0.001 n/a n/a n/a n/a n/a 9_1 6/15/1999 n/a n/a 0.040 + 0.001 n/a n/a 0.099 + 0.006 n/a n/a CD Note: for the above samples, each sample is identified by station number and the water sampling depth in feet, e.g. a7_5 means the samples was collected at station a7 and the sampling depth was 5 feet (1.5m). CD ■ D o Q . C a o o ■ D o CD Q . ■ D CD C/) C/) Winter 2000 Water Samples Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t 1 2/22/2000 0.012 + 0.002 n/a 0.035 + 0.010 0.043 ±0.011 n/a 0.534 + 0.011 n/a n/a al 2/22/2000 0.012 + 0.001 n/a 0.027 + 0.011 0.033 + 0.001 n/a 0.273 ± 0.008 n/a n/a 2 2/22/2000 0.013 + 0.001 n/a 0.017 ± 0.005 0.021 ± 0.002 n/a 0.248 + 0.055 n/a n/a 3 2/17/2000 0.009 + 0.002 n/a 0.015 + 0.003 0.019 ± 0.001 n/a 0.048 + 0.003 n/a n/a 4 2/17/2000 0.012 + 0.002 n/a 0.011 + 0.004 0.013 + 0.001 n/a 0.068 + 0.007 n/a n/a 5 2/17/2000 0.014 + 0.002 n/a 0.017 ± 0.006 0.021 + 0.001 n/a 0.076 + 0.012 n/a n/a 6 2/25/2000 0.005 + 0.001 n/a 0.005 + 0.003 0.007 ± 0.001 n/a 0.032 + 0.001 n/a n/a 8 2/29/2000 0.004 ± 0.001 n/a 0.003 + 0.003 0.004 + 0.000 n/a 0.030 ± 0.001 n/a n/a 9 2/29/2000 0.005 + 0.001 n/a 0.005 ± 0.003 0.007 ± 0.001 n/a 0.048 + 0.007 n/a n/a N i K ) 0\ CD ■ D — i O o . c o CD Q . ■ D CD C/) W o' o o M ay-December 2002 Water Samples (excluding samples collected on July 26-27, 2002; September 6, 2002, and on February 16-17, 2003, which are listed below) o o ■ D c q ’ c 3" CD — i CD " O — i O o . c 9 - o o 3 CD Q . O C -o CD C/> o' =5 Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t SMBI 6/12/2002 0.023 + 0.001 n/a 0.055 ± 0.003 0.036 + 0.001 n/a 0.067 + 0.003 n/a n/a SMB2 6/23/2002 n/a n/a n/a n/a n/a 0.076 ± 0.003 n/a n/a OBP 7/6/2002 0.047 ± 0.002 0.012 ± 0.001 0.069 + 0.002 0.028 ± 0.001 0.023 + 0.001 0.066 + 0.003 0.46 + 0.07 1.04 ± 0.16 IP 7/6/2002 0.031 ± 0.002 n/a 0.046 ± 0.002 0.015 + 0.001 n/a 0.065 + 0.003 0.14 + 0.02 n/a SDR(I) 7/6/2002 n/a n/a 0.393 + 0.009 n/a n/a 0.611 + 0.024 n/a 0.94 + 0.14 SDR(II) 11/9/2002 n/a n/a n/a 0.068 + 0.007 0.162 + 0.007 n/a 0.09 + 0.01 0.09 + 0.10 OTR(I) 7/6/2002 n/a n/a 0.267 + 0.009 n/a n/a 0.412 ± 0.016 n/a n/a OTR(II) 7/27/2002 n/a n/a 0.042 + 0.002 n/a n/a 0.108 ± 0.004 n/a 0.39 + 0.06 OTR(III) 12/26/2002 n/a n/a n/a n/a n/a 0.068 + 0.006 n/a 0.39 + 0.06 SWR(I) 7/6/2002 n/a n/a 0.024 n/a n/a 0.027 + 0.006 n/a n/a SWR(II) 12/26/2002 n/a n/a n/a n/a n/a 0.050 ± 0.005 n/a n/a CW 7/11/2002 0.014 + 0.001 0.019 ± 0.001 0.024 + 0.001 0.007 ± 0.001 0.022 + 0.001 0.028 + 0.003 0.15 ± 0.10 0.01 + 0.10 SI(I) 7/11/2002 0.012 + 0.001 0.045 ± 0.002 0.041 ± 0.002 0.018 + 0.001 0.043 + 0.001 0.077 ± 0.005 0.28 ± 0.10 0.29 ± 0.10 SI(II) 11/9/2002 n/a n/a n/a 0.020 ± 0.001 0.060 + 0.001 n/a 0.19 + 0.29 0.31 + 0.05 SI(III) 12/26/2002 n/a n/a n/a 0.014 ± 0.002 0.048 + 0.003 0.063 + 0.005 0.00 ± 0.15 0.74 ±0.11 HIW(I) 7/11/2002 n/a n/a 0.046 + 0.002 n/a n/a 0.055 + 0.002 n/a n/a HIW(II) 12/26/2002 n/a n/a n/a 0.018 + 0.003 0.039 + 0.003 0.058 ± 0.005 n/a n/a HIE(I) 7/11/2002 n/a n/a 0.028 + 0.002 n/a n/a 0.035 + 0.001 n/a 0.39 ± 0.06 HIE(II) 12/26/2002 n/a n/a n/a 0.010 ± 0.001 0.020 ± 0.002 0.030 + 0.003 0.07 ± 0.10 0.55 ± 0.08 MM 11/8/2002 n/a n/a n/a n/a n/a n/a n/a n/a GP(I) 11/9/2002 n/a n/a n/a 0.015 ± 0.001 0.039 + 0.002 n/a 0.38 + 0.06 0.55 + 0.08 GP(II) 12/26/2002 n/a n/a n/a 0.008 ± 0.001 0.027 ± 0.002 0.036 + 0.003 0.00 ± 0.20 0.48 ± 0.07 TJR 12/26/2002 n/a n/a n/a n/a n/a 0.029 ± 0.005 n/a n/a CC 12/26/2002 n/a n/a n/a n/a n/a 0.058 + 0.007 n/a n/a WSB 12/26/2002 n/a n/a n/a 0.028 + 0.004 0.025 + 0.004 0.054 ± 0.008 1.28 + 0.19 1.28 + 0.19 ESB 12/26/2002 n/a n/a n/a n/a n/a 0.070 + 0.012 n/a n/a Rain 11/9/2002 n/a n/a 0.557 + 0.022 n/a n/a 0.239 ± 0.014 n/a 0.00 + 0.09 to to 7J CD ■ D — i O o . c o CD Q . ■ D CD (/) o ’ 3 O 3 * CD O O ■ D July 26-27 Tidal Cycle Sampling Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t 23:03 7/26/2002 n/a n/a 0.047 + 0.003 n/a n/a 0.068 + 0.002 n/a 1.16 ± 0.2 1:20 7/27/2002 n/a n/a 0.049 ± 0.003 n/a n/a 0.065 ± 0.003 n/a 0.61 + 0.1 3:48 7/27/2002 n/a n/a 0.034 + 0.002 n/a n/a 0.049 + 0.002 n/a 0.52 + 0.1 6:50 7/27/2002 n/a n/a 0.043 + 0.003 n/a n/a 0.049 + 0.003 n/a 1.14 + 0.2 10:55 7/27/2002 n/a n/a 0.055 + 0.002 n/a n/a 0.085 ± 0.003 n/a 1.02 ± 0.2 12:35 7/27/2002 n/a n/a 0.056 + 0.003 n/a n/a 0.059 + 0.002 n/a 0.80 ± 0.1 CD 3" CD September 6, 2002 Sampling (SDB and Open Ocean) CD ■o — i O a . c g- o 3 ■ o — 5 o CD D. O C ■o CD C/) o' 3 Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t 10 9/6/2002 0.018 ± 0.001 0.044 + 0.003 n/a 0.009 + 0.000 0.055 + 0.002 n/a 0.50 ± 0.08 0.71 ± 0.1 20 9/6/2002 0.022 + 0.001 0.036 ± 0.002 n/a 0.010 + 0.000 0.033 ± 0.001 n/a 0.36 ± 0.05 1.19 ± 0.2 30 9/6/2002 0.034 + 0.002 0.033 ± 0.002 n/a 0.018 + 0.001 0.040 + 0.001 n/a 0.79 ± 0.12 0.87 ± 0.1 40 9/6/2002 0.027 + 0.002 0.043 + 0.003 n/a 0.027 + 0.001 0.061 + 0.002 n/a 0.65 ± 0.10 1.35 ± 0.20 50 9/6/2002 0.042 + 0.003 0.032 + 0.002 n/a 0.026 + 0.001 0.050 + 0.001 n/a 0.62 ± 0.09 0.93 ± 0.1 60 9/6/2002 0.033 + 0.002 0.061 + 0.004 n/a 0.043 + 0.001 0.105 ± 0.003 n/a 0.27 + 0.04 1.11 + 0.2 70 9/6/2002 0.047 + 0.003 0.042 ± 0.003 n/a 0.030 + 0.001 0.046 ± 0.001 n/a 0.74 ±0.11 0.96 + 0.1 80 9/6/2002 0.051 + 0.003 0.023 + 0.001 n/a 0.046 + 0.001 0.041 + 0.001 n/a 1.54 ± 0.23 2.06 ± 0.3 90 9/6/2002 0.119 + 0.007 0.092 + 0.006 n/a 0.057 + 0.002 0.014 + 0.000 n/a 2.27 ± 0.34 2.29 ± 0.3 100 9/6/2002 0.040 + 0.002 0.062 ± 0.004 n/a 0.032 + 0.001 0.033 ± 0.001 n/a 0.76 ±0.11 1.24 ± 0.2 50b 9/6/2002 0.030 + 0.002 0.061 + 0.004 n/a 0.045 ± 0.001 0.069 + 0.002 n/a 0.38 ± 0.06 0.90 ± 0.1 60b 9/6/2002 0.042 + 0.003 0.053 ± 0.003 n/a 0.054 + 0.002 0.077 ± 0.002 n/a 0.96 + 0.14 1.69 ± 0.3 70b 9/6/2002 0.056 ± 0.003 0.224 ± 0.013 n/a 0.104 + 0.003 0.337 + 0.010 n/a 0.75 ± 0.11 3.97 ± 0.6 80b 9/6/2002 0.086 + 0.005 0.041 + 0.002 n/a 0.139 ± 0.004 0.075 + 0.002 n/a 0.80 ± 0.12 1.68 ± 0.3 90b 9/6/2002 0.062 + 0.004 0.039 ± 0.002 n/a 0.138 + 0.004 0.054 ± 0.002 n/a 2.61 ± 0.39 3.16 ± 0.5 100b 9/6/2002 0.140 + 0.008 0.057 + 0.003 n/a 0.124 + 0.004 0.083 + 0.002 n/a 1.59 ± 0.24 2.47 + 0.4 to K ) 00 7J CD ■ D — i O o . c o CD Q . ■ D CD (/) W o' o o — t y CD O O ■ D February 16-17 Tidal Cycle Sampling 3- CD — i CD ■ D — i O o . c a o o ■ D — i o CD Q . ■ D CD C / ) (/) Station date Pb-d Pb-p Pb-t Po-d Po-p Po-t Th-d Th-t SI-21:40 2/16/2003 n/a n/a n/a 0.016 + 0.001 0.016 + 0.001 0.032 ± 0.001 n/a n/a SI-12:03 2/17/2003 n/a n/a n/a 0.014 + 0.001 0.012 + 0.001 0.025 + 0.001 n/a n/a SI-3:10 2/17/2003 n/a n/a n/a 0.007 + 0.000 0.010 + 0.001 0.017 + 0.001 n/a n/a SI-6:20 2/17/2003 n/a n/a n/a 0.015 ± 0.001 0.014 ± 0.001 0.028 ± 0.001 n/a n/a SI-9:07 2/17/2003 n/a n/a n/a 0.019 ± 0.001 0.016 + 0.001 0.035 ± 0.001 n/a n/a SI-11:20 2/17/2003 n/a n/a n/a 0.012 + 0.001 0.018 + 0.001 0.030 + 0.001 n/a n/a SI-13:35 2/17/2003 n/a n/a n/a 0.009 + 0.000 0.015 + 0.001 0.025 + 0.001 n/a n/a SI-16:00 2/17/2003 n/a n/a n/a 0.004 + 0.000 0.015 + 0.001 0.019 + 0.001 n/a n/a SI-I7:45 2/17/2003 n/a n/a n/a 0.012 + 0.001 0.015 ± 0.001 0.028 + 0.001 n/a n/a 81-22:13 2/17/2003 n/a n/a n/a 0.011 ± 0.001 0.019 ± 0.001 0.030 ± 0.001 n/a n/a GP-22:0I 2/16/2003 n/a n/a n/a 0.008 + 0.000 0.009 + 0.000 0.017 + 0.001 n/a n/a GP-I2:50 2/17/2003 n/a n/a n/a 0.000 + 0.000 0.009 + 0.000 0.009 + 0.000 n/a n/a GP-3:27 2/17/2003 n/a n/a n/a 0.006 + 0.000 0.006 ± 0.000 0.012 ± 0.000 n/a n/a GP-7:00 2/17/2003 n/a n/a n/a 0.004 + 0.000 0.008 + 0.000 0.012 + 0.000 n/a n/a GP-9:33 2/17/2003 n/a n/a n/a 0.008 + 0.000 0.011 + 0.001 0.019 + 0.001 n/a n/a GP-II:45 2/17/2003 n/a n/a n/a 0.004 ± 0.000 0.008 ± 0.000 0.012 + 0.000 n/a n/a GP-I4:00 2/17/2003 n/a n/a n/a 0.005 + 0.000 0.011 + 0.001 0.017 ± 0.001 n/a n/a GP-I6:19 2/17/2003 n/a n/a n/a 0.007 + 0.000 0.006 + 0.000 0.013 + 0.000 n/a n/a GP-I8:06 2/17/2003 n/a n/a n/a 0.008 + 0.000 0.011 + 0.001 0.019 + 0.001 n/a n/a GP-22:53 2/17/2003 n/a n/a n/a 0.005 + 0.000 0.008 ± 0.000 0.013 ± 0.000 n/a n/a H1W 1 2/16/2003 n/a n/a n/a 0.011 + 0.001 0.010 + 0.001 0.022 + 0.001 n/a n/a HIW2 2/17/2003 n/a n/a n/a 0.010 + 0.001 0.015 + 0.001 0.026 + 0.001 n/a n/a H1W3-1 2/17/2003 n/a n/a n/a 0.017 ± 0.001 0.020 + 0.001 0.037 + 0.001 n/a n/a HIW3-2 2/17/2003 n/a n/a n/a 0.017 + 0.001 0.017 + 0.001 0.034 + 0.001 n/a n/a HIE1 2/16/2003 n/a n/a n/a 0.011 + 0.001 0.012 + 0.001 0.023 ± 0.001 n/a n/a HIE2 2/17/2003 n/a n/a n/a 0.008 ± 0.000 0.014 + 0.001 0.021 + 0.001 n/a n/a HIE3-1 2/17/2003 n/a n/a n/a 0.011 ± 0.001 0.011 + 0.001 0.022 + 0.001 n/a n/a HIE3-2 2/17/2003 n/a n/a n/a 0.007 + 0.000 0.008 ± 0.000 0.016 + 0.001 n/a n/a K ) K ) 'O Appendix II RS-TRIM Program Flow Chart start o f main program datain bcin y \ . update toxics storm evaooration n=maxStep? physical data boundary conditions intemediate result random number generator shipping events simulator resuspension simulator scavenging simulator settling simulator simulation results 230 Reproduced with permission ofthe copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Geochemical studies of beryllium isotopes in marine and continental natural water systems

PDF

Fractionation of nitrogen isotopes during early diagenesis

PDF

Characterization of geochemical and lithologic variations in Milankovitch cycles: Green River Formation, Wyoming

PDF

A proxy for reconstructing histories of carbon oxidation in the Northeast Pacific using the carbon isotopic composition of benthic foraminifera

PDF

Analysis of the effect of varying port spacing on initial dilution for a multiport outfall diffuser

PDF

Investigation of asphaltic macrostructure and its free radical conversion

PDF

Biopolymer plugging effect in porous media: Gelation modeling and lead stabilization

PDF

A geochemical and hydrological assessment of groundwater in the Portuguese Bend landslide, California

PDF

Evaluating gas-phase biofilter performance in treating chlorinated VOCs from industrial emissions: Principle, modeling and design

PDF

Groundwater modeling and geochemical tracer (CFC-12 and tritium) distribution in the Abalone Cove landslide, Palos Verdes, California

PDF

Holocene sedimentation in the southern Gulf of California and its climatic implications

PDF

Biofiltration of petroleum hydrocarbon vapors and co-treatment of volatile organic compounds in low-pH sulfide biofilters

PDF

Biological nitrification-denitrification and sulfate reduction of reverse osmosis brine rejects

PDF

Evolutionary paleoecology and taphonomy of the earliest animals: Evidence from the Neoproterozoic and Cambrian of southwest China

PDF

Gas migration from oil /gas fields and gas storage fields: Environmental implications

PDF

Early Jurassic reef eclipse: Paleoecology and sclerochronology of the "Lithiotis" facies bivalves

PDF

Geotechnical applications of biopolymer technology

PDF

EFAB: A novel, high-aspect-ratio true three-dimensional microfabrication process for rapid, low-cost desktop micromachining of MEMS

PDF

Cyclostratigraphy and chronology of the Albian stage (Piobbico core, Italy)

PDF

Geology and structural evolution of the southern Shadow Mountains, San Bernardino County, California

Asset Metadata

Creator Peng, Jian (author)

Core Title Integrated geochemical and hydrodynamic modeling of San Diego Bay, California

School Graduate School

Degree Doctor of Philosophy

Degree Program Geological Sciences

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag environmental sciences,geochemistry,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Advisor Hammond, Douglas (committee chair), Ku, Teh-Lung (committee chair), Stott, Lowell (committee member), Yen, Teh-Fu (committee member)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-523668

Unique identifier UC11340844

Identifier 3140535.pdf (filename),usctheses-c16-523668 (legacy record id)

Legacy Identifier 3140535.pdf

Dmrecord 523668

Document Type Dissertation

Rights Peng, Jian

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA