University of Southern California Dissertations and Theses

Silicic and germanic acids: Laboratory determination of their molecular diffusivities and field study of their benthic fluxes along the California margin

(USC Thesis Other)

Silicic and germanic acids: Laboratory determination of their molecular diffusivities and field study of their benthic fluxes along the California margin

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content SILICIC AND G ERM ANIC ACIDS: LABORATORY DETER M INA TIO N OF TH EIR M OLECULAR D IFFU SIVITIES AND FIELD STUDY OF TH E IR BENTHIC FLUXES ALONG THE CALIFORNIA MARGIN by Kathleen Marie Cummins A Thesis Presented to the FACULTY OF TH E GRADUATE SCHO O L U N IVER SITY OF SO UTHER N CALIFORNIA In Partial Fulfillment of the Requirements of the Degree M ASTER OF SCIENCE (G EOLO GICAL SC IENC ES) August 2002 Copyright 2002 Kathleen Marie Cummins Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. UMI Number: 1414874 UMI UMI Microform 1414874 Copyright 2003 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. U N IV E R S IT Y O F S O U T H E R N C A L IF O R N IA T H E G R A D U A TE S C H O O L U N IV E R S IT Y PARK LOS A N G E L E S . C A L IF O R N IA 8 0 0 0 7 This thesis, written by Kathleen Marie Cummins under the direction of h&V. Thesis Committee, and approved by all its members, has been pre sented to and accepted by the Dean of The Graduate School, in partial fulfillm ent of the requirements fo r the degree of Master of Science D ta n f)afr August 6, 2002 THESIS COMMITTEE C hairm an Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Acknowledgements The work presented in this thesis was made possible by the contributions from the following institutions and individuals. Funding from the National Science Foundation, grant # O C E 9 911608 to Dr. D. E. Hammond. The Earth Sciences Department at the University of Southern California, for travel funds. The personnel of the OSU ICP Facility, for processing Ge samples from core incubations. The University of Southern California Machine Shop, for constructing the diffusion cells and the core incubation equipment. The captain and crew of the R/V Point Sur, for a safe and productive research cruise. Gerry Smith, Shelley Howard, and Federico Spagnoli, for their laboratory and technical assistance. David Okaya, for the time and effort he put into the fortran code for the numerical simulation. Drs. Will Berelson and Teh-Lung Ku, for their comments and advice regarding this thesis. Dr. Jim McManus for his expert advice and assistance with incubation samples. Steve Colbert, for his willingness to answer my questions, or find the answers, if necessary. Doug Hammond, for his patience, enthusiasm for science, and encouraging me to do math in my head. Jim Bischoff, for having faith in me as a scientist. My friends here at USC, for always being supportive and willing to commiserate. My family, for being proud of me and cheering me on, although they have no idea what I’ve done. John Ferrari, for helping me keep my sanity, and my academic goals in perspective. ii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table of Contents page Acknowledgements ii List of Tables v List of Figures vi Abstract viii Chapter 1: Overview of Research 1 Chapter 2: Molecular Diffusivities of Silicic and Germanic Acids Abstract 4 Introduction 5 Methodology Diffusion Cell 5 Experimental Procedure 7 Sample Analysis 10 Calculation of Diffusivity 14 Results Experimental Results 21 Selection Criteria for Good Data 32 Molecular Diffusivity Results 34 Discussion 38 Conclusion 57 Chapter 3: Nutrient Fluxes from Cores of the California Margin. Abstract 59 Introduction 60 Methodology Incubation Set-Up (Equipment Description) 60 Study Area 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experimental Procedure Sample Analysis Flux Calculation for Incubated Cores Results Macrofauna Observed Core Incubation Benthic Lander (in situ) and Incubation Comparison Discussion Possible Artifacts Pressure Effects Germanium Sequestration Conclusion References Appendix A Numerical Simulation of Diffusivity Cell Description of Model Equations Used in the Numerical Simulation Fortran Code Used in the Numerical Simulation Appendix B Diffusion Experiments: Normalized Concentration (C/C0) Versus Time Appendix C Diffusion Experiments: Observed C /C 0 Versus Numerical Simulation C /C 0 Appendix D Diffusion Experiments: Relative Diffusivities of Solutes Appendix E Determination of Boundary Layer for Core Incubations Appendix F Incubation Nutrient Flux Plots by Station page 62 65 67 68 69 79 83 84 86 87 88 92 93 94 105 110 119 125 131 iv Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Tables Table Page 2-1. Results for Adsorption of Solute to Frit Material 8 2-2. Experimental Conditions for Each Experiment 11 2-3. Diffusivities and Reduced % 2 Generated by the Numerical Simulation 19 2-4. Published Dm Values for KCI 20 2-5. Summary of Observed and Numerical Simulation C /C 0 22 2-6. Molecular Diffusivity Results for Good Experiments 35 2-7. Relative Diffusivities Using Method 1 (Simulated) and Method 2 (Slope2 of Regression) 37 2-8. Ratios of Diffusivities 45 2-9. Pore W ater and Lander Results Comparison from McManus et al (1995) and Sayles et al. (1996) 52 2-10. Hydration Sites for Ions in Sea W ater 54 3-1. Summary of Conditions for Cores 63 3-2. Summary of Core Incubation Data and Fluxes 70 3-3. Incubation Flux Data 77 3-4. Comparison of Benthic Lander and Core Incubation Fluxes 80 E -1. Results of Variable W ater Column Height Experiment 128 E-2. Results of Variable Stir Rate Experiment 129 v Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. List of Figures Figure Page 2-1. Diffusion Cell Schematic 6 2-2. Box Model schematic of Numerical Simulation 16 2-3. Profile in Porous Frit 17 2-4. Normalized Concentration Over Time 28 2-5. Sum of Squares v. D* 30 2-6. Sensitivity of Simulated Results to D* for KCI 31 2-7. Observed v. Model Generated (Simulated) C/C0 Results 39 2-8. Normalized Concentration Plots 41 2-9. Simulation of C/C0 for Ge and KCI as a Test of Relative Diffusivity Approximation (Method 2) 43 2-10. Molecular Diffusivity v. Ionic Strength of Solution 48 2-11. Molecular Diffusivity Ratio (Ge/Si) v. Ionic Strength 56 3-1. Core Incubation Schematic 61 3-2. Map of Study Area 64 3-3. Flux Plot for Core Incubation 75 3-4. Incubation v. Lander (In Situ) Fluxes 81 B-1. Normalized Concentration (C /C 0) Versus Time 106 C -1. Observed C /C 0 Versus Numerical Simulation C /C 0 111 D-1. Relative Diffusivities of Solutes 120 E-1. Boundary Layer v. W ater Column Height 130 vi Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure Page E-2. Boundary Layer v. Stir Rate 130 F-1. Plots Showing Nutrient Fluxes and Ge/Si for Station 1 132 F-2. Plots Showing Nutrient Fluxes and Ge/Si for Station 2 133 F-3. Plots Showing Nutrient Fluxes and Ge/Si for Station 3 134 F-4. Plots Showing Nutrient Fluxes and Ge/Si for Station 4 135 F-5. Plots Showing Nutrient Fluxes and Ge/Si for Station 6 136 F-6. Plots Showing Nutrient Fluxes and Ge/Si for Station 7 137 F-7. Plots Showing Nutrient Fluxes and Ge/Si for Station 8 138 F-8. Plots Showing Nutrient Fluxes and Ge/Si for Station 9 139 F-9. Plots Showing Nutrient Fluxes and Ge/Si for Station 10 140 vii Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Abstract Diffusivities for silicic and germanic acids in dilute KCI and seawater solutions were determined experimentally at 25°C using a diffusion cell. Results (in 10'6 cm2/s), for silicic acid, were 9.03 ±0.29, in low salinity solution (l<0.1), and 7.74 ±0.19, in seawater solutions (l>0.6). Germanic acid results were 7.63 ±0.36 and 7.24 ±0.25, in these solutions. Cores from the California margin were incubated on deck and measured fluxes of nitrate, phosphate, silicon, and germanium were compared to in situ benthic fluxes at the same sites. After corrections for differences up to 5°C between incubation and in situ temperature, fluxes were still 25-30% lower than in situ. However, the Ge/Si flux ratio from incubations was similar to in situ, both exhibited strong fractionation of Ge from Si (~50% the ratio in diatoms). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 1 Overview of Research Silicic acid is an important nutrient used by diatoms to form their opaline tests. The role of diatoms in the Si cycle has recently been summarized by Treguer et al. (1995). Diatoms are photosynthesizers and need to remain in the surface layer of the ocean during their lifetime, thus depleting silicic acid from surface waters as they grow and multiply. When they die, they settle to the sea floor, with some of their tests dissolving (remineralizing) along the way and some reaching the seafloor. As the biogenic opal skeletons that reach the seafloor are buried, some dissolve and enrich the pore waters in silicic acid, sometimes up to 1000 uM (Treguer, 1995; Hurd, 1972). The concentration gradient at the sediment-water interface causes dissolved silica to diffuse out of the sediment into the overlying water column. This flux is an important contribution of silicon to the ocean Si cycle (e.g. Treguer, 1995). Attempts to quantify or constrain the flux of nutrients such as dissolved silica from the sediments have been made by Fanning and Pilson, (1974), Berelson et al., (1987, 1996), McManus, (1995), and many others. Germanium is a Group IVA element just below silicon in the periodic table, indicating that its behavior is similar to that of silicon. This makes Ge a useful tracer of Si in reactions involving silicate precipitation and dissolution 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (Froelich and Andreae, 1981; Froelich et al., 1985). Germanium can substitute for silicon in diatom tests. The amount of Ge that is taken up in the tests directly reflects the ratio of Ge to Si in ocean waters in which the diatoms formed. However, the record in buried opal indicates considerable variability in this ratio and on interpretation is that the Ge/Si ratio in ocean waters varies in response to climate (Mortlock et al., 1991; Shemesh et al., 1989). This makes the Ge/Si ratio a useful proxy for paleoclimate studies, although the mechanism causing this variation is not well understood. The formation and dissolution of biogenic opal is probably the primary control on the germanium distribution in ocean water (Froelich and Andreae, 1981; Murnane et al., 1989). Therefore, it is important to understand the biogeochemical interactions that influence the cycling of germanium in the ocean. King et al. (2000) present evidence from pore water studies that germanium released during opal dissolution is selectively removed by an unidentified precipitation mechanism in suboxic sediments. Murnane et al. (1989) determined that significant fractionation of Ge from Si occurred in anoxic, iron-rich sediments from San Pedro and San Nicolas Basins off the coast of southern California. Hammond et al. (2000) has shown that sequestration of germanium in iron-rich, reducing sediments removes about half the Ge released by opal dissolution. Quantifying the role of diagenesis in fractionating Ge and Si is the driving force behind the research presented in this thesis. This can be done 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. using one of two methods to evaluate the relative fluxes of Ge and Si. The first is to use pore water profiles to calculate a flux. In order to calculate the flux, it is necessary to know the diffusion coefficient of the species of interest. The second is to measure the flux directly using incubation techniques. The first objective of this research was to measure the molecular diffusivities for silicic and germanic acids in aqueous solutions, using a diffusion cell. These molecular diffusivities can then be applied to any pore water profile to calculate fluxes. The second objective was to obtain silicic and germanic acid fluxes using shipboard incubation techniques, from cores collected along the California margin. The California margin is an ideal site for this study because of its high rate of opal cycling (Hammond et al., 2000). The region has high rates of upwelling, creating high productivity, that produces a high rain rate of opal to the sea floor. The burial and subsequent dissolution of this opal allows pore waters to become enriched in silicic acid, as well as germanic acid, thus making the California margin a good site for making flux measurements of silicic and germanic acids from the sediments. The California margin also has a variety of sites, from oxic to anoxic, and this is also a good reason to examine Si and Ge fluxes, which may provide insight to diagenetic processes at these sites. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 2 Molecular Diffusivities of Silicic and Germanic Acids Abstract Germanium and silicon have many similarities in their water column behaviors, but these elements may be fractionated by diagenetic reactions in iron-rich, reducing sediments. Pore water profiles can constrain the magnitude of this fractionation, but diffusion coefficients for both silicic and germanic acid are required to quantitatively determine their transport. The objective of this study was to experimentally determine these diffusivities, using a diffusion cell in the laboratory. A porous frit was filled with solutions containing KCI or seawater, silicic acid, and germanic acid (as a 68Ge tracer). These solutes were allowed to diffuse into a well-mixed reservoir solution of either deionized water or seawater at 25°C. Samples were drawn from the reservoir over time, and a numerical simulation was used to estimate the diffusivity required to best fit the resulting concentration change over about a 4 day period. Relative diffusivities were also determined from regressions of one solute versus another. Results (in 10'6 cm2/s) for silicic acid averaged 9.03 ±0.29 in low salinity solution (I < 0.1) and 7.74 ± 0.19 in sea water solutions (I > 0.6). The results in seawater are about 23% lower than results of Wollast and Garrels (1971), and results for fresh water are about 50% smaller than Applin (1987). Results for naturally occurring germanic acid 4 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (assumed to be 0.99 that of 68Ge) averaged 7.63 ± 0.36 and 7.24 ±0.25 for low and high salinity solutions respectively. Results did not exhibit a significant dependence on the silicic acid concentration. The relative diffusivity of Ge/Si is 0.82 ±0.04 (SDOM ) and 0.93 ±0.04 (SDO M ) in low and high salinity solutions respectively. Introduction The objective of the work described in this chapter was to experimentally determine the diffusion coefficients for silicic and germanic acids in aqueous solutions of varying ionic strengths. These solutions included deionized water, dilute KCI solution and surface sea water. Silicic acid and germanic acid (as a 68Ge tracer) were added to these solutions to carry out the experiments. Methodology Diffusion Cell: This study used what this author has termed a “diffusion cell” that worked similar to the diaphragm cell described by Mills and Lobo (1989). There are, however, some differences. The diffusion cell (Figure 2-1) consists of 3 columns containing a porous frit (HDPE, 1.27 cm diameter x 4 cm length) making up the compartment that the solute diffuses out of. Each column is independent of the others. The frit material was tested for adsorption using 5 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 4 cm Side View Piston prevents evaporation and condensation from forming on cell wall. It is seated against cell wall with o-rings and has a central sample port. Samples are drawn from Sample Reservoir, without removing piston. A small stir bar is rotated at 12 rpm to keep the reservoir well mixed. The volume of fluid at the start of the experiment is 15-20 cm3 . Porous frit material is HDPE with a porosity of about 42% Start solution is cycled through the frit during filling of cell to prevent air pockets within frit. Each column is 4 cm by 1.3 cm diameter. Start Solution Injection Port i Sample Port 'i Piston Sample Reservoir Stir Bar A Porous Frit Column v Top View Stir Bar Porous Frit Figure 2-1. Diffusion Cell Schematic. This figure shows the general features of the cell used in the diffusion experiments. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. all solutes by storing starting solution in contact with the frit material in the columns of the cell. Results (Table 2-1) indicate no adsorption of solute to the frit. At the base of the columns, beneath the porous frit material, is a basal reservoir space (about 1 cm length). A three-way valve, at the base of each column, is used to inject the 68Ge-Si tracer solution that will diffuse into the overlying sample reservoir. The tops of the columns are in contact with a larger cylinder (sample reservoir) with the porous frit flush to the base of the sample reservoir. It is important to note that because the tracer solution diffuses out of the frit, there is a change in concentration at the boundary between the frit and sample reservoir. This change in concentration gives the diffusion cell a slightly different geometry than other methods. The overlying sample reservoir is filled with a solution free of tracer at the beginning of each experiment. A piston fitted with o-rings is pushed down into the sample reservoir to the surface of the solution to prevent evaporation and condensation of sample reservoir solution. Samples are drawn through a central hole in the piston, after each withdrawal, the piston is adjusted to compensate for the volume removed. The hole in the piston is kept sealed when not sampling. Experimental Procedure: Tracer solutions were prepared from D D IW and reagent grade KCI, or from filtered (0.4 pm) surface sea water from the California Borderland (s = 7 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-1. Results for Adsorption of Solute to Frit Material. Length of contact with frit material Si fuM) Ge (dpm/g) KCI (mmho) None 131.8 9209.6 ± 41.6 49.4 ± 0.21 Immediate 128.7 9224.0 ± 42.4 50.7 ± 0.05 5 days 131.6 ± 5.6 j._ _ 6 8 /-^ _ 9293.0 ± 40.9 52.5 ± 0.19 Note: Ge results are for the isotope 6 8 Ge. 34 p.s.u.). To these solutions, small amounts of silicic acid and germanic acid (as 68Ge tracer) were added. The solution pH was measured and found to be 6.5 to 7.5 for various runs. Because pK for silicic and germanic acids are 9.5 and 8.6 respectively, the tracers should have been in their undissociated form. The tracer solution was injected into the basal reservoir and porous frit. It was cycled through the frit several times, until trapped air was removed from the pore space. This procedure was repeated for each column and served to homogenize the solution in the frits because they are independent of each other. Once the cell was filled with the tracer solution, it was immersed in a water bath at constant temperature (25°C). The sample reservoir solution, which was 68Ge and Si free, was stored separately and also placed in the water bath. When the cell had been in the bath for several hours, preferably overnight, the excess 68Ge-Si tracer solution was removed from the sample reservoir and the sample reservoir solution was added. A small stir bar (2 mm diameter x 7 mm length) was used to mix the sample reservoir solution so that it would be homogeneous. The stir bar was situated so that it was not spinning directly on any of the porous frit areas. The bar was rotated slowly (12 rpm) by a small motor with a magnet mounted to it. The slow stir rate was used to minimize the creation of eddy motions within the frit. The stir motor was situated beneath the water bath, so it would not interfere with sampling. After adding the tracer-free solution 9 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. to the sample reservoir, the piston was inserted, the reservoir was given about 5-10 minutes to mix, and the first sample was taken. Experimental conditions for each experiment are listed in Table 2-2. Samples were taken one or two times per day over about a four day period. Although smaller sample aliquots were taken in some of the earlier experiments (expts. 1-6), the optimal sample size was 800 pi (used for all runs after expt.6). The movable piston design was implemented after experiment 10 to avoid evaporation from the sample reservoir and condensation on the cell walls. The 800 pi samples were then analyzed for 68Ge by gamma spectroscopy. After they were counted, two aliquots of 150 pi were taken to analyze for KCI (when KCI was the matrix), by specific conductance. Subsequently, two aliquots of 200 pi were used for Si analysis by colorimetric methods. The composition of the solution filling the frit for each experimental run was also measured and used to normalize each sample run. Sample Analysis: 68Ge decays (287 day half-life) by positron emission with a gamma emitted at 1078 keV. The 68Ge solution used to make the tracer solution was purchased from Brookhaven National Laboratories, Long Island, New York, with an activity of 1.15 pCi on 6/27/00. It was diluted to a convenient working range, about 300 nCi/ml. The activity for the samples was determined by 10 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-2. Experimental Conditions for Each Experiment. Experiment1 ,2 ,3 Cell C0 Si (HM) C0 Ge* (dpm/g) C0 KCI (pmho) Matrix5 Duration (days) Notes 1 A 400 - 1375 100mM KCI/ddiw 2.3 a,b,c 2 A 441 - 1345 100mM KCI/ddiw 3.1 b,c,e 3 A 415 - 1415 100mM KCI/ddiw 3.0 a,b 4 B 1071 - 1377 100mM KCI/ddiw 3.0 a,b,c 5 A 387 - 1305 100mM KCI/ddiw 4.0 a,c 6 B 1006 - 1260 100mM KCI/ddiw 4.0 a.c.f 7 A 378 724 1221 100mM KCI/ddiw 3.1 a,c,f 8 B 914 723 1185 100mM KCI/ddiw 3.1 a,f 9 A 423 381 1103 100mM KCI/ddiw 4.1 c 10 B 1025 393 1039 100mM KCI/ddiw 4.1 c 11 A 345 106 1169 100mM KCI/ddiw 4.0 b 18 A 148 9387 - 100% SW/ddiw 3.9 f 19 B 352 9303 - 100% SW/ddiw 3.9 f 20 A 112 7060 - 100% SW/SW 3.9 d 21 B 311 7971 - 100%SW/SW 3.9 f.d 22 A 119 7767 - 100% SW/SW 4.1 b,d 23 B 328 8525 - 100% SW/SW 4.0 b,d 24 A 138 8792 - 100% SW/SW 3.3 d 25 B 328 8573 - 100% SW/SW 3.3 26 A 136 8673 - 100% SW/80%SW 4.2 f 27 B 336 8879 - 100% SW/80%SW 4.2 f 28 A 142 9498 - ddiw/ddiw 4.1 29 B 499 9852 - ddiw/ddiw 4.1 30 A 154 9081 633 50 mM KCI/ddiw 4.0 31 B 512 9159 671 50 mM KCI/ddiw 4.0 a. Temperature could have been too high for part of run by 1-2° C. b. Stir rate inconsistent c. Condensation on cell wall <1 O O ul d. Convective mixing between sample reservoir and solution in frit. e. Reservoir solution final volume very low less than 10% start volume. f. Stir bar on frit. 1. Experiments 12-17 discarded because a precipitate formed. 2. Experiments 18 on, geometry of cell changed by addition of a piston and new frit material. 3. All experiments run at 25°C, except for experiments 22 and 23, run at 4°C. 4. Ge results are for the isotope 6 8 Ge. 5. Matrix indicates (frit solution)/(sample reservior solution). 11 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. gamma spectroscopy (with an intrinsic Ge well-type detector with well dimensions of 4.5 cm by 1cm diameter) using the 511 keV annihilation peak attributed to the positron emission. This peak was used because it is detected more frequently than the gamma and therefore, has better counting statistics. The raw counts were corrected for background and radioactive decay to obtain the 68Ge concentration on the date that the start solution was made. This is necessary to obtain diffusive fluxes and also is useful for comparison of absolute concentration among starting solutions. A test of the KCI and sea water matrices showed negligible activity at 511keV relative to the samples. Samples and start solutions were counted in the same geometry. The precision of this analysis was determined to be about 1 % based on the uncertainty in counting, which is the square root of the number of counts (Taylor, 1982). Analysis for KCI was by specific conductance, using a V W R Scientific Digital Conductivity Meter, Model 1054, equipped with a flow-through micro conductivity cell. The samples were diluted two fold and conductance was measured. Corrections for background were made when necessary. Concentration is assumed proportional to conductance and absolute concentrations of samples were determined by standardization with 0.1 M (1200 pmho) KCI prepared as a start solution. Analysis for Si was by colorimetric methods. Samples from experiments 1-10 were analyzed using a Hitachi U-1100 Spectrophotometer 12 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. equipped with a flow-through cell. The method used was after the Determination of Reactive Silicate in A Practical Handbook of Seawater Analysis (Strickland and Parsons, 1968). The spectrophotometer data were corrected for drift of the zero and for cross-contamination from sequential samples. There was no additional color development with time from beginning to end of sample analysis. Samples from experiments 11-31 were analyzed using a Lachat Instruments QuikChem 4200 for flow injection analysis. The method used for analysis on the Lachat was QuikChem method 11-114-27-1-A for Silica (Si02) in seawater, but raw data were recalculated to correct for drift and any non-linearity in standard curves. The Lachat data were corrected for drifting base line (blank) over time. Most results were based on peak area, but if bubbles were observed, height was used. Standards for both methods were made from a substock solution of 3560.0 p,M Si in double deionized water that was made from a silicon stock standard 1000 |ig/ml (99.99% spectral purity, analysis lot N43593) from VW R Scientific. Working standards in artificial seawater ranged between 0 and 100 p,M Si for the spectrophotometer; for the Lachat they were and 0 and 10 |o,M Si. All measurements were normalized by dividing observed concentrations by the concentrations observed in the solution used to fill the frit. 13 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Calculation of Diffusivity: A numerical simulation was used to determine diffusivity for each experimental run, based on a one dimensional diffusion box model stepped through time. The model was based Fick’s laws of diffusion, which relate flux to the concentration change over time. The simulation assumes that the concentration within the frit and basal reservoir were uniform at the start, transport within the frit was by molecular diffusion, and the concentration in the sample reservoir was homogeneous and equal to the concentration at the top of the frit. The model was adapted into a FORTRAN code using the following equations which were adapted from Ranalli’s Rheology of the Earth, (1987), but are fundamentally the same as the diffusion equations in Crank (1975). The complete code for sub-routines can be found in Appendix A. First Box (sample Reservoir) at t > 0: Boxes within the frit at t > 0: V < S r ■ *" C /ix + 1 2 C /ir} Initial condition at start of experiment (t = 0): C r = some value > 0 C n x — 1 14 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C ’r = concentration in the sample reservoir after a time step (dt) has elapsed. C r = concentration in the sample reservoir at the previous time step. Ci = concentration of the diffusing species in the uppermost frit (box 1). A = Area of the porous frit material (total area of the 3 frits). ([) = porosity of the frit material (0.42). D* = Diffusion coefficient of the species diffusing out of the frits, dt = incremental time step (30s). V r = Volume of the sample reservoir. dx = box thickness over which species diffuses (0.08 cm). nx = box number. The model has 51 boxes. The sample reservoir is box 0, and boxes 1-50 represent the frit material, with the length of each box corresponding to dx. A schematic diagram of the box model is shown in Figure 2-2. The concentration in the bottom box did not change significantly during the experiment because the diffusion process did not extend that far into the frit. Figure 2-3 shows the extent of the diffusion process in the frit at the start, one day and five days. Even at five days the gradient for the most rapidly diffusing solute only extended about 3 cm into the frit material. This numerical simulation did not correct for the tortuosity of the molecular diffusion path and therefore, a value (D*) was determined (Table 15 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Box 0 Sample Reservoir (well mixed) Box 1 Box 2 ’nx-1 Porous Frit Material (diffusive) nx nx+1 Box 49 Box 50 Figure 2-2. Box Model Schematic of Numerical Simulation. The sample reservoir is box 0 and the porous frit is divided into 50 boxes with length dx. 16 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. depth (cm) Normalized Concentration of KCI 0.2 0.4 0.6 - i r«-i---|-=-i— * -r- 1 -- 1 ---1 ---1 ---1 ---p 0.8 -r— I ---1 — t ■ ■ i Sample 1 Reservoir • Start ■ 1 Day a 5 Days \ Profile in Porous Frit Porous Frit A Frit Solution Reservoir Figure 2-3. Profile in Porous Frit, showing how the concentration changes with depth in the frit material over time. Experiments typically ran about 3-4 days. At 5 days the sample reservoir is about 30% of the starting solution concentration for KCI (diffuses fastest) and the diffusive exchange in the frit has progressed to about 3 cm. This demonstrated that the boundary conditions set in the numerical simulation are preserved over the course of the modeled experiment. 17 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2-3), not the actual molecular diffusivity (Dm ). The molecular diffusivity was determined from D* by applying a cell constant determined from the numerical simulation for KCI, which has a known diffusivity. The cell constant corrects D* to Dm , accounting for effects of tortuosity. Tortuosity refers to the path that a molecule takes when it diffuses. As the name suggests the path is not a straight line. The tortuosity for the diffusion cell was the only parameter that could not be determined before the start of these experiments. All other parameters that affect the diffusivity, such as frit porosity and area, were built into the numerical simulation, including effects of withdrawing samples (Appendix A). The tortuosity was determined by using KCI to calibrate the cell. A known concentration of KCI (50 or 100 mM) was used as the matrix for the diffusing solutions for several experiments. This allowed the calculation of the cell constant that was specific to the diffusion cell. The cell constant (02) was calculated as 02 = Dp u b /D* using the published molecular diffusivity (Dp U b) (Table 2-4) and observed diffusivities (D*) of KCI for each cell. A cell constant was calculated for each run using KCI, but only runs that were good were used to compute the cell constant for either cell (cell A or cell B). Cell constants of 14.1 ± 0.2 and 12.8 for cells A and B respectively were calculated from runs 5, 7, and 8 and used for experiments 1 through 17. The frits were replaced with new material after 18 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-3. Diffusivities and Reduced X2 Generated by the Numerical Simulation. Expt. D* (1<r7 cm2 /s) Reduced X2 Cell Si1 Ge2 KCI3 N4 Si Ge KCI 1 A 7.24 - 16.18 8 1.6 - 6.5 2 A 8.04 - 19.04 14 2.6 - , 4.4 3 A 8.51 - 15.12 8 4.7 - 9.9 4 B 5.38 - 13.24 8 3.0 - 7.9 5 A 6.30 - 13.35 6 0.1 - 1.1 6 B 3.69 - 9.62 6 3.9 - 11.9 7 A 6.26 5.76 13.12 7 0.2 0.7 0.6 8 B 7.59 6.39 16.65 7 0.1 5.0 0.4 9 A 10.53 12.18 21.04 9 1.0 13.3 2.1 10 B 11.65 13.71 25.69 9 2.8 3.6 2.2 11 A 2.51 2.57 5.76 8 10.2 5.9 4.0 18 A 7.46 6.88 - 8 1.5 4.0 - 19 B 7.44 7.07 - 8 3.1 3.4 - 20 A 181.22 172.86 - 8 333.6 1051.3 - 21 B 122.86 125.59 - 8 587.4 266.6 - 22 A 28.88 27.42 - 7 23.8 484.9 - 23 B 7.30 6.21 - 6 14.8 3.9 - 24 A 15.66 16.94 - 7 3.2 9.9 - 25 B 9.49 8.59 - 7 4.4 1.6 - 26 A 7.36 8.03 - 8 1.4 0.3 - 27 B 9.56 9.33 - 8 4.7 4.6 - 28 A 10.85 9.12 - 8 0.7 0.9 - 29 B 27.65 13.85 - 8 10.2 1.2 - 30 A 9.12 7.52 20.32 8 0.5 0.2 0.1 31 B 10.88 5.63 22.94 8 1.3 0.9 0.2 1. Precision of Si is approximately 0.5 uM for Lachat and 0.65 uM for Spectrophotometer. 2. Precision of Ge is based on counting statistics (about 1%). Results are for the isotope 6 8 Ge. 3. Precision of KCI is 0.5mM 4. N is the number of samples drawn during an experiment. 19 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2-4. Published Dm values for KCI. Source Date 0.1M Dm (10-6 ) 0.05M Dm (10-8 ) Mills and Lobo 1983 19.311 Stokes 1959 18.732 Ghosting 1950 18.473 18.643 CRC Handbook of Phys. And Chem. 1964 18.44 Amer. Inst. Of Phys. 1972 18.47 18.64 Li and Gregory 1971 18.42 Average 18.44 18.64 1. Not used in average because it is not similar to other values. 2. Not used because it is incuded in CRC Handbook. 3. Not used because it is included in Amer. Inst, of Physics. Note: Diffusivity for KCI is an average of K* and Cl" diffusivities. I\3 O experiment 17. For experiments after 17, KCI results for experiments 30 and 31 were used to compute cell constants of 9.2 and 8.1 for cells A and B respectively. This large difference in cell constants seems to reflect a difference in the nature of the porous material. Because the new frit material was purchased a year after the first batch, it was suspected that it may not be identical to the original material. The dry density of the new material and the old material was measured for comparison. This was done by weighing to find mass and measuring to find volume of the material. The results were 0.53 and 0.52 g/cm3 for the original and replaced frit material respectively, showing that there was a slight density difference of 0.01 g/cm3. This suggests that there is a slightly different configuration of the pore spaces in the second lot of porous material that has caused the slightly lower cell constants for experiments after expt. 17. Results Experimental Results: Concentrations measured in each experimental run were normalized to the value in the starting tracer solution (C0). A summary of experiments and their normalized KCI, Si, and Ge concentrations (C /C 0) is given in Table 2-5, along with the simulated values of C /C 0. A typical plot of normalized concentrations for KCI, Si, and Ge shows that KCI has the greatest concentration change with time and Si the next, then Ge (Figure 2-4). A 21 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C /C 0. Time js) Observed (C/C0) Si Modeled (C/C0) Si Observed (C/C0) Ge1 Modeled (C/C0) Ge1 Observed (C/C0 ) KCI 0 0.000 0.000 - - 0.007 3660 0.000 0.009 - - 0.022 10860 0.008 0.020 - - 0.033 18000 0.032 0.029 - - 0.044 25140 - 0.037 - - 0.051 175320 0.120 0.118 - - 0.218 184380 0.125 0.121 - - 0.218 194820 0.123 0.125 - - 0.226 0 0.005 0.005 _ 0.007 6780 0.030 0.021 - - 0.041 14040 0.043 0.032 - - 0.063 21360 0.050 0.042 - - 0.078 61260 0.096 0.080 - - 0.149 89160 0.105 0.101 - - 0.156 100740 0.112 0.109 - - 0.167 109440 0.114 0.116 - - 0.175 164400 0.150 0.158 - - 0.219 179160 0.157 0.169 - - 0.227 193620 0.159 - - - 0.238 251400 0.196 - - - 0.279 253140 0.176 - - - 0.286 267720 0.180 - - - 0.297 420 0.019 0.019 _ _ 0.016 9900 0.034 0.039 - - 0.037 73560 0.082 0.094 - - 0.100 93000 0.096 0.105 - - 0.116 145380 0.123 0.131 - - 0.154 173580 0.136 0.143 - - 0.174 231000 0.175 0.167 - - 0.220 259380 0.195 0.178 - - 0.247 300 0.011 0.011 _ 0.011 9780 0.022 0.026 - - 0.028 73440 0.060 0.070 - - 0.090 92880 0.070 0.079 - - 0.105 145260 0.094 0.101 - - 0.143 173580 0.110 0.111 - - 0.163 231120 0.137 0.131 - - 0.203 259140 0.153 0.140 - - 0.227 (C/c0 ) KCI 0.007 0.024 0.044 0.059 0.072 0.208 0.214 0.221 0.007 0.036 0.053 0.068 0.124 0.153 0.164 0.174 0.229 0.244 0.016 0.045 0.117 0.131 0.163 0.179 0.208 0.221 0.011 0.038 0.106 0.119 0.150 0.165 0.193 0.206 22 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C /C 0. (cont.) rime (s) Observed <C/C„) Si Modeled (C/C0 ) Si Observed (C/C0) Ge1 Modeled (C/C0) Ge1 Observed (C/C0) KCI Modeled (C/C0) KCI 1200 0.016 0.016 - - 0.017 0.017 57000 0.071 0.070 - - 0.091 0.097 82800 0.083 0.084 - - 0.111 0.116 161580 0.116 0.118 - - 0.159 0.162 318540 0.165 0.166 - - 0.228 0.225 343260 0.174 0.172 - - 0.239 0.233 1020 0.010 0.010 _ 0.010 0.010 56820 0.039 0.050 - - 0.058 0.078 82560 0.050 0.061 - - 0.076 0.095 161340 0.081 0.087 - - 0.125 0.135 318300 0.132 0.126 - - 0.202 0.192 343020 0.141 0.132 - - 0.214 0.199 2040 0.033 0.033 0.021 0.021 0.035 0.035 25320 0.063 0.064 0.049 0.049 0.076 0.081 82260 0.106 0.103 0.088 0.086 0.134 0.135 111480 0.119 0.118 - 0.100 0.157 0.156 171300 0.146 0.146 0.128 0.127 0.197 0.193 199800 0.160 0.158 0.139 0.139 0.213 0.209 267000 0.184 0.188 0.164 0.167 0.242 0.246 1800 0.028 0.028 0.022 0.022 0.029 0.029 25080 0.061 0.061 0.051 0.051 0.078 0.078 81960 0.098 0.101 0.086 0.087 0.135 0.133 111240 0.115 0.116 0.100 0.101 0.156 0.154 171000 0.142 0.143 0.125 0.125 0.197 0.192 199500 0.155 0.154 0.137 0.136 - 0.209 266760 0.182 0.180 0.160 0.159 0.241 0.246 840 0.029 0.030 0.011 0.011 0.020 0.020 10140 0.052 0.051 0.035 0.034 0.049 0.052 62040 0.101 0.102 0.085 0.090 0.115 0.123 98040 0.118 0.124 0.109 0.115 0.148 0.154 162000 0.149 0.157 0.144 0.150 0.188 0.197 236520 0.188 0.190 0.178 0.185 0.235 0.238 269040 0.203 0.203 0.194 0.199 0.257 0.255 322920 0.230 0.225 0.231 0.223 0.285 0.282 355140 0.244 0.239 0.249 0.237 0.311 0.298 840 0.021 0.021 0.013 0.013 0.020 0.020 23 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C /C 0. (cont.) rime (s) Observed (C/C„) Si Modeled (C/C0 ) Si Observed (C/C0) Ge1 Modeled (C/C0) Ge1 Observed (C/C0 ) KCI Modeled (C/C0) KCI 10080 0.046 0.045 0.042 0.040 0.056 0.058 62040 0.101 0.102 0.102 0.102 0.134 0.140 98040 0.125 0.128 0.126 0.130 0.170 0.175 162000 0.160 0.165 0.168 0.170 0.218 0.224 236520 0.184 0.201 0.205 0.208 0.265 0.271 269040 0.218 0.216 0.220 0.224 0.283 0.290 322860 0.249 0.241 0.253 0.251 0.329 0.320 355080 0.268 0.257 0.274 0.267 0.351 0.339 780 0.030 0.030 0.025 0.025 0.031 0.031 7380 0.040 0.036 0.035 0.030 0.039 0.042 58920 0.062 0.064 0.056 0.058 0.077 0.086 90900 0.071 0.075 0.066 0.070 0.089 0.103 160560 0.092 0.097 0.086 0.093 0.124 0.134 250380 0.120 0.121 0.116 0.116 0.166 0.167 318720 0.136 0.137 0.136 0.133 0.199 0.190 348900 0.152 0.145 0.148 0.141 0.212 0.201 840 0.002 0.002 0.005 0.005 _ _ 1200 0.007 0.003 0.009 0.006 - - 58800 0.028 0.034 0.029 0.035 - - 76680 0.046 0.039 0.047 0.041 - - 157920 0.058 0.060 0.061 0.060 - - 242340 0.075 0.076 0.074 0.076 - - 313080 0.087 0.088 0.087 0.087 - - 337140 0.093 0.091 0.091 0.091 - - 540 0.003 0.003 0.004 0.004 _ 900 0.010 0.004 0.011 0.005 - - 59700 0.034 0.038 0.034 0.038 - - 77520 0.048 0.044 0.048 0.044 - - 157560 0.066 0.065 0.065 0.065 - - 241980 0.082 0.083 0.080 0.082 - - 312720 0.096 0.096 0.094 0.094 - - 336780 0.100 0.100 0.098 0.099 - - 960 0.024 0.027 0.017 0.026 _ _ 12000 0.138 0.114 0.135 0.111 - - 24660 0.238 0.163 0.234 0.160 - - 85740 0.353 0.290 0.354 0.285 - - 106500 0.360 0.318 0.363 0.313 - - 24 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C /C0. (cont.) Modeled (C/C0) Expt. Time (s) Si Si Ge1 Ge' KCI KCI 21 22 23 24 25 rime (s) Observed (C/C0) Si Modeled (C/C0 ) Si Observed (C/C0 ) Ge1 Modeled (C/C0) Ge1 Observed (C/C0) KCI 179160 0.378 0.394 0.377 0.389 - 275460 0.391 0.460 0.389 0.455 - 337860 0.400 0.489 0.392 0.485 - 420 0.018 0.018 0.012 0.012 _ 11400 0.094 0.092 0.088 0.087 - 24060 0.149 0.132 0.149 0.128 - 85140 0.282 0.238 0.274 0.235 - 105900 0.298 0.262 0.293 0.260 - 178560 0.338 0.330 0.339 0.328 - 274920 0.366 0.393 0.359 0.392 - 337260 0.373 0.425 0.377 0.424 - 300 0.021 0.021 0.004 0.004 _ 6960 0.021 0.040 0.007 0.023 - 20220 0.030 0.060 0.019 0.042 - 86460 0.102 0.109 0.086 0.092 - 173700 0.156 0.149 0.138 0.131 - 266940 0.190 0.180 0.171 0.162 - 357840 0.206 0.205 0.187 0.187 - 420 0.006 0.006 0.001 0.001 _ 6600 0.009 0.013 0.004 0.007 - 80160 0.037 0.046 0.032 0.038 - 173580 0.069 0.069 0.057 0.059 - 264300 0.088 0.085 0.075 0.074 - 342660 0.100 0.098 0.090 0.086 - 900 0.020 0.020 0.010 0.010 _ 7380 0.025 0.032 0.016 0.023 - 31560 0.043 0.054 0.034 0.046 - 91800 0.088 0.085 0.077 0.078 - 180900 0.117 0.115 0.112 0.109 - 264420 0.140 0.137 0.134 0.131 - 282420 0.141 0.141 0.137 0.136 - 600 0.011 0.011 0.008 0.008 _ 7020 0.018 0.020 0.014 0.017 - 31200 0.033 0.038 0.029 0.034 - 91440 0.062 0.064 0.057 0.058 - 180600 0.089 0.089 0.083 0.082 - 25 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C/C0. (cont.) Expt. Time (s) Observed (C/C0) Si Modeled (C/C0) Si Observed (C/C0) Ge1 Modeled (C/C0 ) Ge1 Observed (C/C0) KCI Modeled (C/C0) KCI 264060 0.110 0.107 0.101 0.100 - - 282060 0.111 0.111 0.104 0.103 - - 26 300 0.026 0.026 0.008 0.008 _ _ 21900 0.031 0.042 0.023 0.025 - - 90180 0.058 0.064 0.048 0.048 - - 174480 0.080 0.081 0.067 0.067 - - 263760 0.101 0.095 0.083 0.082 - - 340080 0.106 0.106 0.094 0.093 - - 365100 0.111 0.109 0.097 0.097 - - 27 300 0.008 0.008 0.006 0.006 _ _ 21840 0.037 0.028 0.034 0.026 - - 90240 0.052 0.055 0.051 0.053 - - 174420 0.077 0.076 0.071 0.074 - - 263700 0.092 0.094 0.091 0.091 - - 340020 0.107 0.107 0.104 0.104 - - 365040 0.111 0.111 0.107 0.108 - - 28 600 -0.001 0.001 0.006 0.006 _ . 7200 0.006 0.010 0.012 0.014 - - 15900 0.014 0.017 0.018 0.020 - - 90180 0.047 0.049 0.048 0.049 - - 172800 0.069 0.070 0.067 0.069 - - 263100 0.088 0.088 0.085 0.086 - - 353400 0.107 0.103 0.102 0.100 - - 29 540 0.003 0.003 0.006 0.006 _ _ 7200 0.020 0.019 0.014 0.016 - - 15840 0.038 0.031 0.023 0.025 - - 90120 0.086 0.080 0.057 0.060 - - 172800 0.115 0.112 0.084 0.084 - - 263040 0.136 0.138 0.103 0.104 - - 353340 0.156 0.161 0.124 0.121 - - 30 600 0.002 0.002 0.007 0.007 0.007 0.007 10800 0.010 0.014 0.017 0.017 0.024 0.026 77700 0.041 0.044 0.046 0.045 0.068 0.071 106200 0.052 0.053 0.053 0.052 0.081 0.083 175800 0.076 0.070 0.067 0.068 0.106 0.107 264300 0.083 0.087 0.084 0.084 0.133 0.132 26 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-5. Summary of Observed and Numerical Simulation C /C 0. (cont.) Expt. Time (s) Observed (C/C0) Si Modeled (C/C0) Si Observed (C/C0) Ge1 Modeled (C/C0) Ge1 Observed (C/C0) KCI Modeled (C/C0 ) KCI 349680 0.104 0.102 0.098 0.097 0.154 0.152 31 600 0.007 0.007 0.008 0.008 0.009 0.009 10800 0.017 0.020 0.015 0.017 0.024 0.030 77700 0.053 0.055 0.042 0.041 0.076 0.079 106200 0.062 0.064 0.048 0.048 0.092 0.092 175800 0.082 0.083 0.063 0.062 0.115 0.118 264300 0.105 0.103 0.078 0.077 0.146 0.145 349680 0.121 0.119 0.088 0.090 0.171 0.167 1. Ge results are for isotope 6 8 Ge. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Normalized Concentration NORMALIZED CONCENTRATION OVER TIME 0.25 0.2 • KCI ♦ Si a Ge ♦ ♦ ▲ 0.15 ♦ ▲ ▲ 0.1 ♦ 0.05 ♦ ▲ P Experiment 7 k ■ ■ ■ ■ I ■ ■ ■ • I_I I ■ » I — I I_I ---I--1 I I l, | I I , , 1 ,. I I 1 -.1- I 1 . . J . Figure 2-4. Normalized Concentration Over Time. Shows the concentration change of each species over time for a typical experiment. Each measurement was normalized relative to composition of the solution used to initially fill the frit. The first sample is taken about 15 minutes after the start of the experiment and is not zero because a small amount of the tracer solution may remain in the sample reservoir when the tracer-free solution is added. 0 0.5 1 1.5 2 2.5 3 3.5 Time (days) 28 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. complete set of plots of C /C 0 for each solute of each experiment versus time is in Appendix B. This change in concentration over time for each solute can be related to their respective diffusivities. Because diffusivity is inversely proportional to mass (Jahne, 1987), it is expected that KCI will have a greater diffusivity than silicic acid (H4S i0 4) and silicic acid will have a greater diffusivity than germanic acid (H468G e 0 4). The best estimate of D* was found by using the numerical simulation to calculate concentration at each sampling time observed in individual experiments. This was done fifty times over a predicted range of D*. These results generated by the numerical simulation, were compared to the normalized concentrations observed for each sample in each experimental run. The differences between observed and calculated results were squared and summed with the smallest sum of squares (%2) considered the best-fit diffusivity (D*) for the experiment (Figure 2-5). A best fit diffusivity (D*) was determined for each solute in each experiment (Table 2-3). The cell constant found for each cell using KCI was applied to the D* value of each solute to convert to the molecular diffusivity. Estimating the uncertainty in D* for a single run is not easy. A test of the sensitivity of the results to D* was done by performing simulations in which D* differed by 10% from the best fit value. The corresponding C /C 0 values for the D* values were then plotted against the observed C /C 0 values for KCI (Figure 2-6) for the five good experiments (5,7,8,30,31) with KCI as 29 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S u m o f Squares 1 10-3 8 10-4 6 10-4 4 10-4 2 10-4 0 100 1 Figure 2-5. Sum of Squares v. D*. The observed and model generated concentrations were compared over a range of diffusivities to determine the best fit diffusivity. A numerical simulation was used to calculate diffusivity (D*) at 50 points over a given range of diffusivity for each species. The difference for each concentration pair was squared and summed to find the sum of squares. The lowest sum of squares is considered the best fit diffusivity. I I I I ^ I I ‘" T I | I I I I | I I I I | I i i i | I 1 I I | i i i i • Sum of Squares v. D*.* Experiment 5 ' '_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I_I--1 --1 _I_I_I_I_I--1 --1 --1 --L . 10-6 1.2 10-6 1.4 10-6 1.6 10-6 Diffusivity (cm2/s) 30 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Observed KCI C/CQ v. Simulated C/Co Observed KCI C/Cn v. Simulated C/Co D* varied bv ± 10% i i i I i i i * i I i i 0.25 E r r o r 0.20 Q « 0.15 0* + 10% E r r o r e n | 0.10 5 5 0.05 0.00 0.15 0.20 0.25 0.00 0.05 0.10 P* varied bv ± 10% 0.30 0.25 C h l s q O 0.20 o ■ o 0.15 I 3 E CO 0.10 y ■ M 2 ‘M 0 T r r o T 0.05 C h i w 0.00 ^ 0.00 0.15 0.20 0.25 0.05 0.10 EXPERIMENTS Observed KCI C/Cc EXPERIMENT 7 Observed KCI C/C0 Observed KCI C/Cn v. Simulated C/Co D* varied by ± 10% 0.30 0.25 o° < 0.20 O | 0.15 a E 5 5 0.10 E r r o r 0.05 0.00 0.25 0.15 0.20 0.00 0.05 0.10 Observed KCI C/C„ v. Simulated C/Co D* varied bv ± 10% y » M 2 * M 0 B F D * + 1 0 % 0 . 9 9 8 7 5 0.10 y » M 2 M O D +10% BFD* KCI 10% 0 . 9 B 9 9 3 0 . 0 0 7 2 5 3 9 C h U a l 2 . 0 3 4 o » O 5 I 0.06 M • M O O a l u S B F D M O * E r r o r 0 . 9 9 8 8 4 EXPERIMENTS Observed KCI C/CQ 0.00 0.02 EXPERIMENT30 0.04 0.06 0.08 0.10 0.12 0.14 0.16 Observed KCI C/C0 Observed KCI C/CQ v. Simulated C/Co D* varied by ± 10% 0.20 0.15 □ E 5 5 D* +10% 0.10 0.05 E r r o r 0.00 0.10 0.15 0.20 0.00 0.05 EXPERIMENT 31 Observed KCI C/C0 Figure 2-6. Sensitivity of Simulated Results to D* for KCI. These plots show how the observed versus the numerical simulation C /C0 would change if the D* varied by 10%. 31 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the matrix. Results of these simulations indicated that a change of 10% in D* yielded a 4% difference in slope, with about 1% uncertainty. Thus, the fractional change in D* is 2.5 times greater than the fractional change in simulated v .observed slope. The 1% uncertainty is attributed to random errors in C /C 0. It indicates that slopes of 0.99 to 1.01 would be adequate fits. This is equivalent to an uncertainty of 2.5% in D* for each experiment. Precision for Ge and Si should be comparable. Once this step was completed the data were examined to identify any procedural problems. Selection Criteria for Good Data First, a reasonable D* value is one that cannot exceed Dm . Second, a reasonable agreement between the observed C /C 0 and the numerical simulation C /C0 can be used as an indicator of how well the model fits the observed data (i.e. how realistic is the D* value determined by the numerical simulation). This can be done using a reduced %2 test, incorporating the analytical precision of the analysis for that solute. This test allows evaluation of whether the diffusion model is a good description of the solute transport that occurred. The reduced %2 was determined using the equation below. / 2 \ ( X ' G 2 X I 32 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where: %2 = sum of difference squared determined by the numerical simulation. N-2 = the number of observations minus two degrees of freedom o2x i = the uncertainty in the analytical precision of the analysis squared. If the model is a good description of the data then the reduced % 2 should be about or less than 1 (Taylor, 1982). It is expected that sometimes the reduced % 2 should exceed 1. The probability of this happening depends on the degrees of freedom. For example, if the degrees of freedom are 6 then the probability of exceeding 1 is 42% (Taylor, 1982). The reduced % 2 calculation also depends on how well the precision of the analysis is known. For KCI the precision of the conductivity meter is 0.5 mM for experiments 1 through 11 and 0.8 for experiments 30 and 31. The Si precision depends on the instrument of analysis. Typically, the spectrophotometer has a precision of 0.65 pM and the Lachat has a precision of 0.5 pM. W hen normalized to the starting solution (tracer solution), this provides an estimate for C /C0 precision. However, in some cases, Si precision was better or worse than this typical value and the estimate of a for Si may be off by as much as a factor of 2 for any individual run. The Ge precision is based on counting statistics, where the precision is proportional to the square root of total number of counts (Taylor, 1982). Typically samples were counted until the 33 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. precision was at 1%. In some cases, the analytical precision for Si and Ge was known to be questionable, and it seemed in the case of some experiments the precision may not have been as good as estimated. In these cases it seems reasonable to allow a reduced %2 that is higher than 1 when all other criteria indicate that the experiment is good. Consequently a cut off of 5 was chosen as an acceptable value for reduced %2. Values of reduced %2 < 5 were considered acceptable fits. Third, relative diffusivities of Si to KCI, Ge to KCI and Ge to Si, were examined. This should minimize effects of some factors that may affect diffusivity, such as minor temperature and stirring inconsistencies, because they should affect each solute equally. Molecular Diffusivity Results Results for eleven experiments meeting the criteria above, are found in Table 2-6, which gives the molecular diffusivity (Dm ) determined from the numerical simulation and lists the results in order of increasing ionic strength (I). The average Dm for Si, Ge, and KCI are also listed, grouped as high ionic strength and low ionic strength solutions. Uncertainties are calculated from the variability of replicate experiments, and the standard deviation of each group is comparable to the 2.5% precision predicted from the sensitivity test. Molecular diffusivities (D m ) for dissolved monomeric silica (Si), determined using the numerical simulations and conversion of D* to Dm using 34 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2-6. Molecular Diffusivity Results for Good Experiments. Dm is in 10'6 cm2 /s and Ge results are for isotope 68Ge. Experiment Cell Molecular Diffusivity (Dm ) Si Ge Cell KCI1 Constant Matrix Ionic Strength (I) Tracer Solution pM Si 28 A 9.95 8.36 - 9.2 ddiw 0 142 30 A 8.36 6.90 18.63 9.2 25 mM KCI 0.025 154 31 B 8.83 (4.57)® 18.63 8.1 25 mM KCI 0.025 512 5 A 8.76 - 18.44 14.1 50 mM KCI 0.05 387 7 A 8.70 8.01 18.44 14.1 50 mM KCI 0.05 378 8 B 9.59 7.57 18.44 12.8 50 mM KCI 0.05 914 18° A 6.84 6.31 - 9.2 50% sw 0.35 148 19° B 6.04 5.74 - 8.1 50% sw 0.35 352 26 A (6.75)b 7.36 - 9.2 90% sw 0.63 136 27 B 7.76 7.58 - 8.1 90% sw 0.63 336 25 B 7.71 6.98 - 8.1 100% sw 0.70 328 D J < 0 .1 Si Ge Average 9.03 7.71 1. Assumed from literature. Standard Dev. of group 0.61 0.63 a. Excluded, Chauvenet Criteria. SDOMd 0.18 0.31 b. Excluded, poor precision in Si analysis. Precision0 0.29 0.36 c. The large density gradient (DDIW above sea water) may have caused these experiments to stratify, D m I > 0.6 Si Ge relusting in diffusivities that are too low. Average 7.74 7.31 d. Standard deviation of mean calculated from Standard Dev. of group 0.04 0.30 replicate measurements, or assuming replicate SDOM 0.13 0.18 precision is 2.5%, whichever is larger. Precision® 0.23 0.25 e. Including uncertainty in cell constant of ± 2.5%. C O cn a cell constant range from 9.95 to 6.04 x 10"6 cm2/s. Germanic acid (Ge) molecular diffusivity values ranged from 4.57 to 8.36 x 10"6 cm2/s. It is worth noting that Ge Dm for experiment 31 is significantly smaller than the other good runs. It was excluded based on Chauvenet’s criterion. It is also worth noting that Si Dm for experiment 26 is 6.75 x 10-6 cm2/s. This seems a bit low, based on similar experiments, and is also smaller than the Dm for Ge (7.36 x 10"6 cm2 /s) for that experiment. The plot for Si v. time is quite noisy suggesting poor precision for Si analysis. These Si data for experiment 26 were excluded based on the noisy data set and on the large uncertainty in the Ge/Si ratio (Table 2-7). Average Dm values for each solute are also calculated for solutions with ionic strength greater than 0.6 and less than 0.1. Uncertainties were computed based on replicates and assuming no experiment has a precision better than 2.5%. The uncertainty of 2.5% in the cell constant was also incorporated in the final estimate of uncertainty. The ionic strength reflects an average of the concentrations for the solutions in the sample reservoir and the frit. In all good experiments, except for experiment 25, the sample reservoir had a lower ionic strength than the solution in the frit. This was done to prevent convective mixing. Diffusivities in low salinity solutions (ionic strength less than 0.1) for dissolved monomeric silica are higher than diffusivities in high salinity solutions (e.g. seawater, 35%o) by about 15%. The average Si Dm for low salinity solutions is 9.03 ± 0.29 x 10-6 cm2/s, while 36 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2-7. Relative Diffusivities Using Method 1 (Simulated) and Method 2 (Slope2 of Regression). Experiment Cell Method 1 Si/KCI Method 2 Si/KCI Method 1 Ge/KCI Method 2 Ge/KCI Method 1 Ge/Si Method 2 Ge/Si 1 * A 0.45 0.34 ± 0.05 - - - - 2 * A 0.42 0.41 ± 0.04 - - - - 3 * A 0.56 0.58 ± 0.01 - - - - 4 * B 0.41 0.43 ± 0.02 - - - - 5 A 0.47 0.50 ± 0.01 - - - - 6 * B 0.38 0.42 ± 0.02 - - - - 7 A 0.48 0.51 + 0.03 0.44 0.46 ± 0.02 0.92 0.90 ± 0.02 8 B 0.52 0.51 ± 0.04 0.41 0.41 ± 0.03 0.79 0.81 ± 0.01 9 * A 0.50 0.55 + 0.02 0.58 0.66 ± 0.03 1.16 1.19 ± 0.04 10 * B 0.45 0.55 ± 0.04 0.53 0.61 ± 0.01 1.18 1.12 ± 0.06 11 * A 0.44 0.41 ± 0.03 0.45 0.43 ± 0.02 1.02 1.05 ± 0.03 18 A - - - - 0.92 0.92 ± 0.03 19 B - - - - 0.95 0.93 + 0.01 20 * A - - - - 0.95 1.05 + 0.03 21 * B - - - - 1.02 1.06 ± 0.01 22 * A - - - - 0.95 0.96 + 0.02 23 * B - - - - 0.85 0.84 + 0.05 24 * A - - - - 1.08 1.09 + 0.03 25 B - - - - 0.91 0.92 + 0.03 26 * A - - - - 1.09 0.96 ± 0.09 27 B - - - - 0.98 0.97 ± 0.04 28 A - - - - 0.84 0.79 ± 0.01 29 * B - - - - 0.50 0.60 ± 0.07 30 A 0.45 0.49 ± 0.06 0.37 0.38 ± 0.01 0.82 0.76 ± 0.08 31 * B 0.47 0.50 ± 0.02 0.25 0.25 ± 0.02 0.52 0.50 ± 0.02 indicates bad runs, bold rows indicate good runs. Note: All Ge results are for the isotope 6 8 Ge. Method 1 is from numerical simulation and method 2 is from slope of the regression. the Si Dm for higher salinity solutions is 7.74 ± 0.23 x 10-6 cm2/s. There is no obvious difference between the two low salinity solutions (25 and 50 mM KCI) in Si Dm results. Ge Dm results tend to be higher in the low salinity solutions (7.71 ± 0.36) than in the high salinity solutions (7.31 ±0.25), with the exception of experiment 31, which appears anomalously low. Discussion The observed concentrations were plotted against the simulated concentrations. They should yield a slope of one if the fit is good. If the fit is not good, it may indicate non-diffusive behavior in the cell. A complete set of plots is in Appendix C. An example of a good fit and a bad fit are shown in Figures 2-7 (a) and (b). Figure 2-7 (a) shows the fit of the observed Ge C /C 0 versus the numerical simulation C /C 0 of Ge for experiment 26. The slope of the line is one and points are distributed randomly about the regression. Figure 2-7 (b) is the same, using data for experiment 20. The fit is clearly not good. The points are too low for the earlier samples, too high for the middle samples and too low for the later samples. The curve is clearly concave down. This pattern and the high values for D* (Table 2-3), suggest convective mixing occurred. Experiments 20 and 21 also exhibited a pronounced concave down curve and a very high D* value and they were almost certainly influenced by convection. They were run immediately after 18 and 19, and insufficient time was allowed for the cell to dry between runs. 38 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Experim ent 26 Observed v. Modeled Ge C/C, y = M1+M2*M0 Value 0.0093804 0.04 0.06 M odel G enerated G e C/C, 0.50 0.40 o O O o O ■ o a t a i n ■ Q O ~ i—1 —1 —1 —■ —r Experim ent 20 Observed v. Modeled Ge C/Cc (b) 0.30 - 0.20 - 0.10 - 0.00 _1 _ • • y = M1+M2*M0 Value Error ml 0.062713 0.039576 m2 0.79126 0.12433 Chisq 0.017865 NA R 2 0.87097 NA 0.00 0.10 0.20 0.30 M odel G enerated G e C/C0 0.40 0.50 Figure 2-7. Observed v. Model Generated (Simulated) C /C 0 Results. This figure shows the results for germanic acid in two experiments. (a) This figure shows the best fit of the numerical simulation to the observed concentrations. The simulation is a good fit because points are randomly distributed around the regression line and it has a reduced % 2 of 0.3. (b) This figure shows the best fit for another experiment. It exhibits non-diffusive behavior, indicated by curvature as points are below at beginning and end and high in the middle of the experiment. It has a reduced % 2 of 1051.3. 39 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As a result the tracer solution was diluted by about 10% with DD IW from rinsing the cell to clean it. Undiluted matrix was placed in the overlying reservoir and its higher density apparently was sufficient to cause convection. The relative diffusivity was determined using two methods (Table 2-7). The first method used the ratio of the molecular diffusivity (D m ) values generated by the numerical simulation, for Si:KCI, Ge:KCI and Ge:Si. The draw back of this method is that there is no direct measure of uncertainty. Method two used the slope generated by plotting the normalized concentrations of Si and Ge against KCI and Ge against Si. Figure 2-8 shows an example of Si v. KCI and Ge v. Si. A complete set of relative diffusivity plots for each experiment is in Appendix D. The drawback to method two is that it does not indicate if convection occurred. The regression slope of one solute versus another (method 2) can be used as an estimate of the square root of the diffusivity ratio. By squaring the slope of the regression an estimate of relative diffusivity can be determined. Crank (1975) has shown that if the surface concentration is nearly constant during the experiment, the total flux of material M across the frit should be given by: where: C0 = concentration in the frit initially (pmol/cm3). D = diffusivity (cm2/s) 40 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 0.20 E X P E R IM E N T 7 Si v. K C I C /C o 0.15 o O O 0.10 < 7 5 Value Error ml 0.008543 0.002095 0.05 m2 0.71224 0.012638 Chisq NA 0.99843 NA 0.00 0.00 0.05 0.10 0.15 0.20 0.25 KCI C/C0 E X P E R IM E N T 7 Ge v. Si C/Cn 0.16 0.12 o O O £ 0.08 Value Error m l -0.011116 0.0013086 0.04 m2 0.94822 0.010304 Chlsq NA NA 0.99953 0.00 0.00 0.05 0.10 0.15 0.20 Si C/C0 Figure 2-8. Normalized Concentration Plots. This was the approach for method 2. Plots show the normalized concentration of Si v. KCI and Ge v. Si. The square of the slope should be equal to the ratio of diffusivities for each solute pair. These plots were used to calculate relative diffusivities and uncertainties for each experiment. 41 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. M = Mass of the solute that diffused into the sample reservoir (pmol/cm2). t = time over which the solute diffuses (s). From this relation, it is clear that the concentration ratio of the two solutes will be: A C0\ and that: / D \ v 2 v Ay / \ 1/2 A vAy ( c /c 0\ (C/C0)b Consequently, if normalized concentrations for one solute (a) is plotted versus those for a second solute (b), the square of the slope can be used to obtain the ratio of diffusivities. This method is particularly useful, because the slope uncertainty can be used to determine the uncertainty of the estimate of relative diffusivity. To test whether the approximation for method 2 is reasonable, a numerical simulation was run using the best-fit diffusivity (D*) for Ge and KCI from experiment 30. These solutes have the greatest difference in diffusivity, so they provide the best test of this approximation. This simulation produced 120 C /C 0 values for each solute calculated over about 5 days. This yielded a concentration curve for Ge and KCI as shown in Figure 2-9 (a). From these C/C0 values, 7 points were chosen that represent typical sampling times for experiments. Results for Ge were plotted against results for KCI to produce 42 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. o O O 0.20 0.15 0.10 0.05 0.00 ♦ KCI C/C, Ge C/C, Simulated C/Cn v. Time Experiment 30 D * = 20.3 x 10-r D* = 7.5 x 10-7 0.79 1.63 2.46 3.29 4.13 4.96 Time (days) 0.10 0.08 0° 0.06 O Q > O 0.04 0.02 0.00 i i—i—|—i—> — i i | ■ i—i—i—| —i i t “ |'" Simulated Ge v. KCI C/C( Experiment 30 (b) — r—r —i— [ i i y = M1+M2*M0 Value Error ml 0.0014 0.0008 m2 0.628 0.008 Chisq 6.24e-06 NA R2 0.99914 NA - * ■ — » 1 ■ * 1 I 1 ■ ■ I ■ 1 1 1 • ■ ■ 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 KCI C/C„ Figure 2-9. Simulation of C /C 0 for Ge and KCI as a Test of Relative Diffusivity Approximation (Method 2). (a) The simulation of 120 C/C0 values based on a given D* for Ge and KCI. (b) The regression for 7 data points taken over four days that best represent the average sampling interval for all experiments. 43 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. the regression shown in Figure 2-9 (b). The slope of that regression is 0.628 ±0.008 and the square of that slope is 0.394 ±0.016, which is 7% higher than the ratio of simulated diffusivities of 0.37. So, it appears that this method (regression squared) is biased to slightly higher relative diffusivities than using Dm or D* ratios. If a shorter time interval is used, or diffusivities are more similar (Ge v. Si, for example). The approximation becomes even better. Table 2-8 compares the two methods by which relative diffusivities were determined. The first method used the ratio of molecular diffusivities generated by the numerical simulation. The second method used the slope of the regression squared of the observed C /C 0. The table shows relative diffusivities for good experiments, listed in order of increasing ionic strength, the average relative diffusivities with the standard deviation of the mean (SDO M ) and the standard deviation of the group (st. dev.). The average was also calculated without experiment 31, because the Ge results for that experiment are about half that of Si, which is cause for re-examination of the data. The D* value for Ge was low for this experiment, but the fit of the numerically simulated and the observed C /C 0 is good. The relative diffusivities of G e to KCI and Ge to Si are also low due to the low Ge result. Stirring and temperature problems are not likely to be the cause of the low Ge result for that experiment, because they would affect all solutes equally, not just one. The recalculation of the average without experiment 31 was 44 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2-8. Ratios of Diffusivities. Results for Ge are for the isotope 68Ge. Experiment Method 1 Simulated Si/KCI Method 2 Slope Si/KCI Method 1 Simulated Ge/KCI Method 2 Slope Ge/KCI Method 1 Simulated Ge/Si Method 2 Slope Ge/Si Ionic Strength (I) p.M Si Start 28 - - - - - - 0.84 0.79 0.01 0 142 30 0.45 0.49 ± 0.06 0.37 0.38 ± 0.01 0.82 0.76 ± 0.08 0.025 154 31 0.47 0.50 ± 0.02 0.25 0.25 ± 0.02 (0.52) (0.50 ± 0.02) 0.025 512 5 0.47 0.50 ± 0.01 - - - - - - 0.05 387 7 0.48 0.51 + 0.03 0.44 0.46 ± 0.02 0.92 0.90 ± 0.02 0.05 378 8 0.52 0.51 ± 0.04 0.41 0.41 ± 0.03 0.79 0.81 ± 0.01 0.05 914 18 - - - - - - 0.92 0.92 ± 0.03 0.35 148 19 - - - - - - 0.95 0.93 ± 0.01 0.35 352 26 - - - - - - (1.09) (0.96 ± 0.09) 0.63 136 27 - - - - - - 0.98 0.97 ± 0.04 0.63 336 25 - - - - - - 0.91 0.92 ± 0.03 0.7 328 Average (Ge/Si) KCI (I < .1)_____________________________________________________________________________________ methods 1 & 2 average 0.48 0.50 0.41 0.42 0.84 0.82 0.83 ± St Dev. 0.03 0.01 0.01 0.01 0.06 0.06 0.05 + SDOM 0.01 0.00 0.03 0.04 0.03 0.03 0.04 SW (1 > .6) average - - - - - 0.94 0.95 0.94 ± St Dev. - - - - - 0.05 0.04 0.05 ± SDOM - - - - - 0.04 0.03 0.04 Note: results in parentheses are not used in calculation of averages. oi done to see if Chauvenet’s Principle (Taylor, 1982) could be applied. The relative diffusivity for Ge to KCI and Ge to Si were beyond 3 sigma for that average, so that experiment was excluded from the calculation of the average, although the cause of the discrepancy is unknown. Averages for relative diffusivity were calculated for high (I > 0.6) and low (I < 0.1) salinity solutions based on methods 1 and 2 (Table 2-8). Based on the test using the numerical simulation, it would seem reasonable to choose the ratio of molecular diffusivities found by method 1 rather than method 2 (the regression squared method), as being a more reliable estimate, as the latter may be a few per cent high if diffusivity differs by a factor of 2. This is apparent for Ge/KCI and Si/KCI. However, for Ge/Si the estimates agree within experimental uncertainty and an average is used. The average of the two methods used to determine Ge/Si in low salinity and high salinity solutions are 0.83 ±0.04 (SDO M ) and 0.94 ±0.04 (SDO M ) respectively. Hammond et al. (2000) predict a lower limit for relative diffusivity for Ge/Si of 0.83, based on the square root of the solute mass ratio, assuming no hydration. This value is in line with the low salinity relative diffusivities of this study, suggesting that there are little or no waters of hydration for solutes in the low salinity solutions. However, the average of high salinity diffusivities (0.94) is consistent with a solute mass ratio of 0.88. This ratio would be achieved if both solutes each had 12 waters of hydration. It is not known why the lower salinity solution appears to have fewer waters 46 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. of hydration. It is likely that the true answer is somewhere in between these two estimates. Polymerization is another factor that should be considered. It should decrease Dm by making the molecule larger and slowing random molecular motions due to the additional mass. Ignoring the matrix of the tracer solution, there is no trend or pattern of higher concentration Si solutions having a lower Dm . Experiment 26 has a Dm for Si that is lower than the Dm for Ge, this might suggest polymerization, however, the tracer solution concentration of silicic acid was 142 jiM, which is the lowest concentration used for this suite of experiments. The Si results for this run (experiment 26) were quite noisy and consequently discarded. If polymerization were a factor it should be evident in the tracer solutions with higher silicic acid content, which it is not. Another line of evidence is that Applin (1987), presents a calculation of polymerization as a function of pH v. % silicic acid. All the tracer solutions for this study are all well within the monomeric silicic acid field. There are only two other studies published that attempt to experimentally determine the diffusion coefficient for silicic acid, and no measurements of germanic acid diffusivity have been discovered in the literature. Applin (1987), experimentally determined a diffusion coefficient for silicic acid in fresh water of 2.2 x10'5 cm2/s (25.5 ±0.5°C) and also found that the diffusivity decreased slightly with increasing silicic acid concentration. 47 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Wollast and Garrels (1971) experimentally determined a diffusion coefficient for silicic acid in seawater of 1.0 ±0.05 x10'5 cm2/s at 25° C. The molecular diffusivity (Dm ) for Si and Ge from this study was plotted against ionic strength (Figure 2-10 (a) and (b)) to see if the matrix played any role in the diffusion process. Solutions of low ionic strength (l<0.1) have higher Dm values than solutions of higher ionic strength. As ionic strength increases, Dm for silicic acid appears to pass through a minima (Figure 2-10 (a)). Molecular diffusivities for germanic acid versus ionic strength are shown in Figure 2-10 (b). This Figure shows the same pattern where Ge Dm passes through a minima. Some solutes appear to have a similar behavior (Li and Gregory, 1974), and this may be related to decreasing hydration numbers at higher ionic strength, thus overcoming the effect of increased viscosity at the higher ionic strengths. However, little difference in the availability of free H2O molecules should occur at these ionic strengths, as will be shown later. Alternatively, this may be an experimental artifact due to the development of a stagnant boundary layer in the reservoir during the experiment. Runs at intermediate ionic strength might have been strongly stratified because sea water in frits was overlain by D D IW in the sample reservoir. Perhaps stirring was insufficient to overcome this stratification, influencing results for both solutes. Consequently, the intermediate salinity results are considered suspect. 48 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 10 9 C M E o < o 8 o 5 7 E ° 6 I 1 1 1 1 I “ J — I — P — T — I — | — I — I — I — I — | I I I I | I I I — I — pTTTT Dm Si v. Ionic Strength (a) Low Salinity matrix • Si 142 ♦ Dm Si 356 ▲ Dm Si 533 T Dm Si 914 Dm Si W ollast Viscosity Seawater matrix I I T 10t 'X (±) 50% i » » ■ 1- ■ « » « I « 1 ' ■ 90% 1 ■ ■ * ■ 1 * ■ ■ ■ 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Ionic Strength C M E 0) C D E 12 Dm Ge v. Ionic Strength 10 - - Low Salinity matrix 8 6 90% 100% Seawater matrix 50% 4 • Ge Dm Viscosity 2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Ionic Strength Figure 2-10. Molecular Diffusivity v. Ionic Strength of Solution. Molecular diffusivity of Si (a) and Ge (b) versus ionic strength of Solution. Data points in parenthesis were not used in the calculation of the average diffusivity for either Si or Ge. The line indicates the expected relation due to the effect of salinity on viscosity. The results for 50% sea water are considered questionable due to possible stratification of the sample reservoir. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Molecular diffusivity results in low salinity (KCI matrix) and fresh water (DDIW ) solutions for this study were compared to those of Applin (1987). Applin’s result of 2.2 x 10"5 cm2/s is more than twice that of this study and the results of Wollast and Garrels (1971). Applin attributed his high diffusivity to a combination of effects related to sea water. The first effect was a difference in viscosity. Li and Gregory (1974) predict sea water diffusivity should be 8% lower than fresh water. Second, Applin suggests an overall decrease in the mobility of silicic acid in sea water, due to the dissociated silicic acid species Si(OH)3' which is more susceptible to hydration, although only 5% of the total is dissociated at pH = 8.2. He also adds that complexes of dissolved silica would tend to slow the diffusion of dissolved silica. According to Applin the above factors might partially explain the difference between his results and those observed by Wollast and Garrels (1971) in sea water. However, they do not explain the difference from measurements of this study in low ionic strength water. The technique employed by Applin for his diffusion experiments was an agar gel method. Although his technique gave results for KCI in agreement with Robinson and Stokes (1959), it is suggested by Mills and Lobo (1989) that the agar method is subject to large systematic errors and extrapolation to zero agar concentration may not yield results consistent with other methods. Consequently, the difference between the results presented here and those of Applin (1987) is attributed to an unidentified artifact of the agar method. 50 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results presented here for dissolved monomeric Si in seawater from this study (Table 2-7) are about 23% lower than the Dm value of 10.0 ±0.5 x 10'5 cm2/s of Wollast and Garrels (1971). Molecular diffusivity (Dm ) of silicic acid in sea water for this study is 7.74 ± 0.23 x 10'5 cm2 /s at 25°C. The technique employed by Wollast and Garrels was to use a cell containing an inner sample reservoir and outer tracer reservoir separated by a Millipore filter. Sampling required the withdrawal of 10 mis from the inner reservoir, temporarily creating a difference in pressure between their sample reservoir and their spiked tracer solution. Perhaps this pressure difference was sufficient to drive advective flow from the outer to the inner reservoir, effectively increasing the transport of Si and thus increasing the Dm observed. A constraint on Dm is provided by field studies. McManus et al. (1995) and Sayles et al. (1996) make comparisons between benthic lander fluxes measured in situ and pore water derived fluxes of silicic acid (Table 2-9). The ratio of fluxes from the two approaches is (weighted mean) 1.12 ±0.15. The pore water fluxes are based on the Dm reported by Wollast and Garrels and assumes that flux is proportional to the diffusivity. If the diffusivity obtained by the present study had been used, the pore water/lander flux ratio would have been 0.87 ±0.11. Consequently, the field observations suggest the true value for Si Dm lies between the present results and those of Wollast and Garrels (1971). An intermediate value is also close to that predicted by 51 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 2-9. Pore W ater and Lander Results Comparison from McManus et al., (1995), and Sayles et al., (1996). McManus etal., (1995). Latitude Longitude Lander fluxes Pore Water Fluxes #of Cores fs j O (/> 140° w 0.46 ± 0.02 0.38 ± 0.04 2 0° 140° W 0.46 ± 0.02 0.51 ± 0.06 2 2° N 140° W 0.43 ± 0.05 0.55 ± 0.06 2 5° N 140° W 0.22 ± 0.02 0.25 ± 0.03 2 Sum 1.57 0.06 1.69 0.10 8 PW flux/Lander flux 1.08 ± 0.07 Sayles etal., (1996). #of Latitude Longitude Lander fluxes Pore Water Fluxes Cores 31.8° N 64.3° W 0.047 ± 0.004 0.057 ± 0.005 3 PW flux/Lander flux 1.21 ± 0.10 Note: Flux units are mmol/m2 day and SDOM was calculated for Sayles et al. based on their quoted standard deviation. PW flux/Lander flux (weighted by number of cores) 1.12 ± 0.15 taking the value from the low salinity solution (9.03 ± 0.29 x 10'6 cm2/s) and reducing it by 8 % (8.31 ± 0.27 x 10'6 cm2/s) to account for viscosity effect. The difference between the predicted Dm (8.31 ± 0.27 x 10'6 cm2/s) and the results of this study (7.74 ± 0.23 x 10'6 cm2/s) is 0.57 ± 0.35 x 10'6 cm2/s. This difference is within 2o, suggesting that the predicted and observed results are not that different. Of course additional systematic errors may exist in Table 2-9 due to uncertainties in tortuosity, porosity, and temperature correction of diffusivity, and consequently this test is not definitive. The difference between results for low ionic strength solutions and sea water brings into question the extent of solute hydration in these two matrices. Solutes develop a hydration layer, due to the charge of the ion or its polarity. It should depend on the ionic strength of the solution. Calculations based on Robinson and Stokes, (1959), suggest that hydration of ions or molecules by H20 in the low salinity solutions is very small or negligible. Table 2-10 is a summary including effective diameter from Garrels and Christ (1965) and the number of hydration sites for several ions in sea water. Based on estimates of effective diameter for hydrated ions (Garrels and Christ, 1965), magnesium ion (Mg+2) should have 70 w ater molecules attached, but Robinson and Stokes (1959) report only 12 sites for water molecules. Although the latter estimate is just for the first layer, it is evident that a volume calculation may not be the best determination of how many 53 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 2-10. Hydration Sites for Ions in Sea Water. Sea W ater Effective Diam eter (A) Hydration Sites Notes Na+ 4 5 a,b cr 3 4 a,c SO42 4 5 a,d Mg+2 8 12 a,b a. Effective diameter values from Garrels and Christ(1965). b. Hydration sites from Robinson and Stokes (1959). c. Hydration sites estimated from effective diameter d. Same effective diameter as N a \ so assume same number of sites. Ol - C * . molecules make up the hydration layer. Because the second and third layer are not strongly bound to the first layer and the water molecules are constantly interacting with the surrounding solution it is questionable whether the second and third layers could be considered added mass that would influence the molecular diffusivity. To estimate the number of sites for Cl" and SO 4" 2, the effective diameter was used. It was assumed that since the sodium ion and the sulfate ion had the same effective diameter that it would have the same surface area and thus the same number of hydration sites. The chloride ion has a slightly smaller effective diameter than Na+, about 25% smaller, so a hydration number of 4 was assigned. Based on the 4 major constituents of sea water (Na+, Mg2+, Cl", S 0 42') and their respective concentrations (0.5M , 0.055M , 0.6M, 0.03M ) in sea water, the total number of moles hydrating the ions is 5.7 out of 55.6 moles of water molecules per liter of sea water. It appears that there are plenty of water molecules available to hydrate the silicic and germanic acids in sea water and differences in hydration between solutions used here should be minimum. Figure 2-11 shows the ratio of molecular diffusivity for germanic and silicic acids (Ge/Si) versus ionic strength. Silicic acid concentration in tracer solutions is indicated in the legend. There does not appear to be a significant change in the ratio of diffusivities with increasing ionic strength. Because the isotope 68Ge was used in all the experiments an adjustment must be made to determine Dm of natural inorganic germanium 55 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ge/Si 1 . 2 0 n — 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — 1 — i— 1 — 1 — 1 — r Ratio of Ge/Si v. Ionic Strength 1.00 0.80 0.60 0.40 0.00 i ■ » |_|_|_L . I I • 142 [iM Si ■ 356 |.iM Si a 912 |aM Si i i » i I » i i » 0.20 0.40 Ionic strength 0.60 0.80 Figure 2-11. Molecular Diffusivity Ratio (Ge/Si) v. Ionic Strength. Shows the ratio of molecular diffusivity (determined by method 2 ) of silicic and germanic acids (Ge/Si) versus ionic strength of the solution. 56 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (atomic weight = 72.6). Using the relationship reported by Jahne (1987), of the square root of molecular weight being inversely proportional to the diffusivity and an intermediate value for hydration waters of 6 H2O, the difference in molecular diffusivity between 68Ge and 73G e should be 1%. Consequently, measurements of 68Ge presented here should be reduced by 1 % to estimate diffusivity of naturally occurring Ge, resulting in values of 7.63 ± 0.36 and 7.24 ±0.25 (10‘6 cm2/s) for low and high salinity solutions, respectively. Conclusion The molecular diffusivities of silicic and germanic acids appear to be dependent on salinity of the solution, but are independent of silicic acid concentration from 140 to 900 |iM Si.. The average Dm for silicic and germanic acids in low salinity (I < 0.1) solutions are 9.03 ±0.29 and 7.63 ±0.36 x10"6 cm2 /s respectively. The average Dm for silicic and germanic acids in high salinity solutions (I > 0.6) are 7.74 ±0.19 and 7.24 ±0.25 x10‘6 cm2/s respectively. The relative diffusivity of Ge to Si has a smaller variability and is a more useful number than the individual molecular diffusivities, when working with pore water profiles to calculate fluxes. The relative diffusivities of Ge/Si also appear to be dependent on salinity. Low salinity solutions have a relative diffusivity average of 0.82 ± 0.04 (SDO M ), which corresponds to the square 57 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. root of the ratio of solute masses for unhydrated molecules. The high salinity solutions (seawater) have a relative diffusivity average of 0.93 ± 0.04 (SDOM ) which suggests some water of hydration is present. However, it is unknown why this ratio seems to increase in going from fresh water to sea water. 58 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Chapter 3 Field Study of Benthic Fluxes Along the California Margin. Abstract Cores were collected from 9 sites along the California margin in July 2001, to measure nutrient fluxes and Ge/Si flux ratios for comparison with data obtained from in situ benthic flux chambers. The cores were incubated on board the ship in a cold room and samples were drawn over a two day period. Nitrate, phosphate, silicon, and germanium were analyzed and plotted against time/height of water column in the core. The slope of this plot defines the flux. In some cases, curvature was observed and non-linear fits were necessary. These fluxes were compared to fluxes determined by in situ benthic chambers. In some cases, core incubation flux temperature varied by up to 5° C from in situ temperature, and estimates of the importance of this difference were made. Even after a temperature correction, fluxes were about 25-30% lower than in situ fluxes for all nutrients. The Ge/Si ratio from core incubations exhibited strong fractionation of Ge from Si at stations 1, 2, 3, 6 , 7, and 10; flux ratios were about 50% of the ratio in diatoms. These results were comparable to in situ Ge/Si fluxes, with the exception of station 4 (Tanner Basin) where in situ data show a preferential release of Ge. Incubations exhibited small fractionation at stations 8 and 9. Overall, this pattern is similar to expected results with sequestration at sites with near- 59 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. surface pore waters rich in iron (stations 1, 2, 3, 6 , and 7). The sequestration at station 10 (3400m ) was unexpected. Introduction The purpose of this experiment is to evaluate whether shipboard core incubation can provide measurements equal to in-situ benthic flux chambers. Potential advantages of this technique are the ease of core collection, shorter duration of stations, and less investment in capital equipment. However, this technique may introduce artifacts. Methodology Incubation Set-up (Equipment Description) The core liners used to collect the samples for incubation were acrylic (extruded), fitted with a PVC plug at the bottom and a movable PVC piston at the top (Figure 3.1), both made of polyvinyl chloride (PVC). The piston and plug were sealed with o-rings. The piston had a 12 volt DC fractional horsepower motor (Hankscraft) mounted to the top that turned a magnet at 32 rpm. This magnet coupled with a 1 inch stir bar located in a basket assembly just below the piston (inside the core liner) to keep the water column well mixed. The piston also had a sample port that consisted of a 3 foot length of 1 /8 inch nylaflow tubing with a 2 way valve attached at the end. The valve was coupled with a syringe for drawing the sample. 60 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Core Incubation Schematic Motor magnet Piston C stir bar ) Sample Port The piston is driven down into the water column as the sample is drawn. The stir bar spins at about 30 rpm and is contained in a sm all basket attached to the piston. Water Column The ideal water column height is about 10 cm. A t a stir rate o f 32 rpm and water column height of 10 cm a boundary layer o f about 120 pm is produced. The ideal sedim ent-water interface is horizontal. PVC Plug ^ Figure 3-1. Core Incubation Schematic. This figure shows the general features of the core set up used for the shipboard incubations. 61 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Study Area Cores were collected from 9 study sites near the coast of Southern California and Central California (Table 3-1, Figure 3-2). Benthic flux chambers (Berelson and Hammond, 1986) were deployed at these sites for 1 to 2 days to measure in situ nutrient fluxes. Protocols for deployment and data reduction are given in Hammond et al. (1996). The spatial pattern of fluxes and time variability will be examined in future work. Experimental Procedure Cores were retrieved using a multi-corer. After recovery, they were promptly placed in a cold room. Spot checks (1-2 times per day) indicated a relatively constant temperature of 5.5 ±1.1 °C. However, temperature may have varied somewhat more than this depending on traffic in and out of the cold room and how much the cooling rate was adjusted to counter this effect. Some temporal and spatial inhomogeneity in the cold room temperature must have existed as cores for station 1 (next to cold room wall) exhibited ice formation in the overlying water at the end of their incubation. Desirable cores for incubation had a horizontal or nearly horizontal sediment-water interface. The spring-loaded arm on the multi-core rack was released and the bottom plug was then inserted in the bottom of the core. Then the multi core rack was removed and the piston was inserted into the top of the core, taking care to avoid trapping air bubbles in the core. The piston was then 62 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-1. Summary of Conditions for Cores. Station Core Location Lat. Lon. Date Time In Depth Notes Core °C In situ °C 1-7 A San Pedro 33°35.72 118°31.98 4-Jul 1333 897 a,f,g 6.0 5.2 B San Pedro 33°35.72 118°31.98 4-Jul 1333 897 a.f.g 2-4 A Santa Monica 33°44.32 118°47.06 3-Jul 1905 904 b,d,e 5.5 ±0.7 5.1 B Santa Monica 33°44.32 118°47.06 3-Jul 1905 904 b,d 3-23 A Catalina 33°16.37 118°38.96 7-Jul 840 1328 b,h,i 6.3 ±1.0 4.1 B Catalina 33°16.37 118°38.96 7-Jul 840 1328 b,j 4-19 A Tanner 32°59.12 119°45.46 6-Jul 1536 1539 b,c 6.6 ±0.9 3.9 B Tanner 32°59.12 119°45.46 6-Jul 1536 1539 b,c,e,i,j 6-48 A Morro Bay 6 35°17.92 120°58.35 9-Jul 2312 100 a,k,l,n 4.4 ±0.5 9.7 B Morro Bay 6 35°17.92 120°58.35 9-Jul 2312 100 b,i,k,m,n 7-46 A Morro Bay 7 35°17.86 121°01.57 9-Jul 2028 198 b,k 4.6 ±0.4 8.6 B Morro Bay 7 35°17.86 121°01.57 9-Jul 2028 198 b,k 8-51 A Sta 8/02 min 35°26.14 121°24.35 10-Jul 1951 704 a,o 4.6 ±0.5 4.9 B Sta 8/02 min 35°26.14 121°24.35 10-Jul 1951 704 0 9-57 A Sta 9/1500m 35°33.80 122°03.04 11-Jul 2200 1500 i,m,p 5.2 ±0.6 2 .7 B Sta 9/1500m 35°33.80 122°03.04 11-Jul 2200 1500 i.k.q 10-69 A Sta 10 36°07.35' 122°35.58' 13-Jul 727 3444 r,s,t 6.3 ±0.5 1.6 B Sta 10 36°07.35' 122°35.58' 13-Jul 727 3444 t a. Piston leakage greater than 100 mis. b. Piston leakage less than 100 mis. c. Possible oxygenation of water column. d. Oxidation visible in sediments. e. Irregular stirring. f. Cores partially frozen at time of last withdrawl. g. Worms, tube-like, amphopods, forams, algal mat. h. Brine shrimp, tube worms. i. Brittle star. j. Stalk like tubes, brown fluffy seds. k. Polychaetes I. Jelly fish, m. Shrimp n. Fluffy brown/green seds. o. Burrows present, but no animals, p. Sandy core top with mud. q. Tree-like, stick-like tube worms, r. epibenthic foram. s. Burrows, t. Brown/green mud. 38° .O ' detail j San Francisco Monterey v Bay 180 1 2 0 °W 36° S 100 km Los Angeles 34° % C B T B PE 32° 118°W Figure 3-2. Map of Study Area. California borderland, stations are numbered. 1. San Pedro (SP). 2. Santa Monica (SM). 3. Catalina Basin (CB). 4. Tanner Basin (TB). 6 . Morro Bay (100 M). 7. Morro Bay (200 m). 8 . Oxygen Minimum Zone. 9. Slope (1500 m). 10. Slope (3400 m). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. driven down until a water column height of about 10 cm remained. Then, stirring started at a rate of 32 rpm. These parameters were selected based on lab studies of the dissolution of Alabaster to estimate boundary layer thickness, and should create a boundary layer of about 120pm. Details are found in appendix D, as well as a complete description of the procedure used to determine stir rate. Five samples were drawn over a two day period. Samples were drawn with Teflon syringes and filtered through a 0.45 pm filter. The first 4 mis of sample were discarded (volume of the nylaflow tubing). The second 4 mis were used to rinse the syringe prior to drawing the sample. A draw of 30 ml was then taken, filtered, and split approximately 2 0 mis for nutrients (PO4, NO3, and Si) and 10 mis for Ge and metals. The split for nutrients was stored at 4°C for approximately 2-3 weeks in nalgene HD PE 30 ml bottles and the split for Ge was stored at 4°C for approximately 3 months in acid washed nalgene HDPE 15 ml bottles. Sample Analysis Nutrients, (phosphate (P 0 4), nitrate + nitrite (N 0 3) and silicic acid (Si)) were analyzed colorimetrically using a Lachat Instruments QuikChem 4200, by flow injection analysis. The method used for phosphate was QuikChem method 31-115-01-3- A, for Orthophosphate in Seawater, with a precision of ±0.1 pM. Standards 65 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ranged from 0 to 4 pM and exhibited a linear response over this range. They were made from a 100 pM working stock standard solution. Samples exceeding this range were diluted with artificial sea water. The working stock standard solution was made from a 4m M PO43' stock standard which was made by dissolving 0.5444g of primary standard grade anhydrous potassium dihydrogen phosphate (KH2PO4, FW 136.09). The method for nitrate was QuikChem Method 31-107-04-1-A, Nitrate in Brackish or Seawater 0.5 to 50.0 pM (7 to 700pg N/L as NO3" or NO2') with a precision of ±0.1 pM. Standards ranged from 0 to 50.0 pM and were made from a working stock standard solution of 1250 pM. The working stock standard solution was made from a stock nitrate standard, made by dissolving 0.5055 g of potassium nitrate (KNO3, FW 101.11). Previous experience has shown nitrite to be an insignificant contribution to the nitrate (Hammond, pers. comm.). The method for silica was QuikChem Method 11-114-27-1-A, Silica (Si0 2 ) in Seawater 1.66 to 333 pmol (0.1 to 20.0 mg Si0 2 /L) with a precision of 0.5pM. Standards ranged from 0 to 100 pM Si and were made from a working stock standard of 5 mM Si, which was made from a 1000 ppm stock standard from VW R. As a check, previously made standards with in the 0- 100 pM range were also incorporated into the standardization. Samples exceeding this range were diluted with artificial sea water. 66 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Germanium analysis was done at Oregon State University using isotope dilution method of Mortlock and Froelich (1996) as modified by Hammond et al. (2000). Precision of analysis was ±2.5% (Hammond, pers. Comm.). Flux calculation for incubated cores The rate of concentration change in the overlying water depends on the ratio of flux to water height. However, water height changed during the course of each incubation as samples were drawn. To eliminate this effect, the ratio of time/height for each interval was calculated and the concentration in overlying water was plotted against the total elapsed time/height. The slope of this plot defines the flux. Despite the attempt to linearize results, curvature was apparent in most experiments, indicating a decreasing flux with time. Consequently, results were fit with both linear and non-linear functions. For non-linear fits, the slope at the beginning of the experiment was taken as the best estimate of flux. Uncertainties for each incubation were calculated from the uncertainty in slope. To calculate Ge/Si ratios, a regression of Ge v. Si was used. 67 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Results Macrofauna observed Cores from station 1 and 2 (897 and 904m respectively) exhibited a surficial layer of iron-rich (oxidized) sediments (about 1 cm thick) with no evidence of animals. Station 1 appeared to have an algal mat. Station 3 (1328m ) exhibited stalk like tubes (forams?), suggesting an oxic/(sub-oxic?) environment. Station 4 had brown fluffy sediment and core B contained stalk-like tubes (forams?) and a brittle star. Stations 6 and 7 (100 and 198m respectively) exhibited numerous polychaetes, and station 6 also had shrimp and brittle stars, and a jelly fish. Station 6 by far had the greatest number and variety of animals. A polychaete was observed inside the jelly fish at one point and then it crawled out. Station 8 (704m ) is at the depth of the oxygen minimum zone, exhibited no visible animals, but did have some deep burrows (to about 15 cm). Station 9 (1500m ) had a sandy surface with some mud and exhibited brittle stars, shrimp, polychaetes and a tubeworm. Station 10 (3444m ) had an epibenthic foram and a burrow with a worm in one core and no visible animals in the other, suggesting a sub-oxic environment. All animals observed in cores were on top of the sediment unless otherwise noted. Observed activity seemed to be at a minimum, but constant throughout the incubation ( ~ 2 days). 68 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Core Incubation Data for samples drawn during incubations is given in Table 3-2. A typical plot for core incubation flux results for one station is shown in Figure 3-3. The complete data set for stations along the central and southern California margin are shown graphically in appendix F. Each station shows flux plots for nitrate (NO 3), phosphate (PO 4), silicic acid (Si), and the ratio of germanic acid to silicic acid (Ge/Si in pM/jnM). Curvature in most plots was readily apparent, although curvature is rarely seen for in situ flux chamber results. The cause of curvature is probably related to the short water column height (1 0 cm or less) and long duration (about 2 days) that caused relatively large changes in nutrient concentrations. This decreased the concentration gradient and probably suppressed the flux during later stages of each run. Alternatively, it may reflect changes in biological activity with time or effects of decreasing oxygen (not measured) during the incubation. Because of these effects, the best estimate for flux should be the slope at the beginning of the experiment, when these perturbations are smallest. For nitrate an exponential function was chosen as a reasonable approximation of the non-linearity, except when changes of observed concentration were less than 10%. Then, a linear fit appeared to be a reasonable approximation. For phosphate, data was a bit noisier than expected, based on analytical uncertainty, and the curvature was not always 69 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3-2. Summary of Core Incubation Data and Fluxes. w c Station_____ Sample Id 1 01-07-04-1 a 01-07-04-2a 01-07-04-3a 01-07-04-4a 01-07-04-5a 01-07-7B-1b 01-07-7B-2b 01-07-7B-3b 01-07-7B-4b 01-07-7B-5b 2 02-04-4B-1a 02-04-4B-2a 02-04-4B-3a 02-04-4B-4a 02-04-4B-5a 02-04-7B-1b 02-04-7B-2b 02-04-7B-3b 02-04-7B-4b 02-04-7B-5b 3 03-23-24-1 a 03-23-24-2a ~ v j o Time (min) (days) Height (cm) (m) 36 0.03 10.61 0.11 388 0.27 10.11 0.10 1401 0.97 9.44 0.09 1986 1.38 7.36 0.07 2761 1.92 6.52 0.07 48 0.03 9.61 0.10 395 0.27 9.08 0.09 1399 0.97 7.61 0.08 1985 1.38 5.77 0.06 2760 1.92 5.23 0.05 52 0.04 8.51 0.09 635 0.44 8.01 0.08 1329 0.92 7.47 0.07 2344 1.63 6.99 0.07 2939 2.04 6.48 0.06 42 0.03 8.61 0.09 632 0.44 8.14 0.08 1327 0.92 7.70 0.08 2337 1.62 7.20 0.07 2927 2.03 5.75 0.06 65 0.05 10.61 0.11 610 0.42 10.04 0.10 Observed Concentration PQ4 (uM) NQ3 (uM) Si (uM) Ge (uM) 3.411 33.690 116.1 79.0 3.539 29.923 124.2 87.1 4.669 25.887 136.3 82.7 4.608 23.127 141.2 88.8 5.370 20.241 151.2 92.0 3.425 33.260 117.4 - 3.609 30.575 124.6 - 4.019 25.487 137.5 - 4.184 22.543 143.0 - 4.500 23.620 170.0 - 3.479 31.040 120.1 84.6 3.741 27.106 126.5 83.6 3.877 23.041 134.0 90.1 4.171 18.171 143.9 88.5 4.333 16.427 147.4 87.9 3.468 30.151 120.3 - 3.655 24.283 126.9 - 4.184 21.350 137.4 - 4.629 17.151 146.1 - 4.854 15.029 152.2 - 3.241 40.560 127.8 88.1 3.121 38.921 130.3 98.4 Table 3-2. Summary o f Core Incubation Data an d Fluxes, (cont.) c o 5 re L _ ■ * - » c a > o c o O ■ a Q ) £ a > to n O C M 1 1 1 1 1 00 1 C M 1 1 C M C O CD N - to to T - IO N-* to 0 0 CD 0 0 0 00 C D C D C D CD CD C D 0 T — 0 0 C M T — T” CD 00 00 00 T — Is - I s- O C O C M Is - to 0 M ; to to to O ID C D cm" C D T “ C O CD 0 0 c \i C M 00 T * 0 00 00 C M 0 0 C O 00 00 1 0 h - C O C O C O Is - 00 C O T— T— T~ V C O C O C D to C D CD v - CD IO C D CD V O 00 T— to C D o> h - CO CD O T — to lO C M T f CD 00 C O 00 N - 00 CD C M LO C M 0 00 Is - 00 05 CD to C D CD C M C D C D CD T — C D C D C D 1 0 0 CD Is-* to T j- C M O CD O 00 n ! C D 00 00 00 T f CO C O 00 00 ^ r T f 00 C O 00 0 0 C O C M O C D O 00 0 T— 0 T - C D O C D 0 0 0 0 T * “ 0 d d d d d d d d d d O ’ i- £ ' _ E gi ,0 '3 X to >\ re •u II a > a E re CO c o 5 re 4 -i (0 0 0 T - to to to C O to to C M C M CD C O C O C M C M 0 0 r - T“ C O 00 C D I s- CD 00 CD C M T— O X — to to 00 00 T f O 0 C M C M T — O v T “ C M p T — C M CO O CO 00 00 00 00 C O 00 00 00 00 00 C O C O 00 0 0 00 0 0 C O C M CD 00 0 0 C M O O O t- O 0 5 T- O O O O O O O O O C M CD T — 0 0 C D Is- Tf* T— 0 0 CD C M Is - T- C D to T— 0 0 T — to C D C D C O C M I s- C M 00 to 0 Is - C O O to C O O C O CD 00 00 d 0 ■ T — CD 0 0 C O d CD CD 0 0 Is - ’ V “ d CD CD d 0 0 p to p 00 0 O C M M- C M O p C M O 00 0 Is- 0 Is - 00 p C M O (D T " 0 I s- CD p Is- 0 to O V-* V“* C M O d T “ ” c \i 0 d d T ”* C M O d 0 r - C M O CN LO C O T f 0 0 C M C M in TJ- CN C D r N C l O O C D T — C M Is - C M 00 00 C O to to to O h - t— X— C M to C D C M C M ^— 0 0 I s - CD t } - C M CD C M O CD C M O M - CD C M C M T— C M C M T— C M C M I- h - C O C O (0 JO -O _Q JO _Q C D C D C D C D C D jQ _Q JO JO C D 00 to ■ * “ C M 00 to C M C O to T “ C M 00 in ■4 C D CD C O C O C O 00 00 C O C O C O C M C M C M C M C M T— C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M C O 00 00 00 C O 00 00 00 C D C D CD C D CD CD C D CD CD C D 0 0 C M C M C M C M C M C M C M C M t— t— T — * * — T— T “ T — T — ’ T — T— M- C O 00 00 00 0 0 00 00 0 0 T f Tf- T}- XT ■4 CD 0 0 O O O O O O O O O O O O O O O 0 O (O 71 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 06-48-21-2 a 5 8 2 0.40 9.89 0.10 2.352 22.648 61.8 36.8 06-48-21-3 a 1362 0.95 7.36 0 .07 3.428 17.645 101.6 06-48-21-4 a 1842 1 .2 8 6.77 0 .07 3.590 13.942 122.8 55.6 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3-2. Summary of Core Incubation Data and Fluxes, (cont.) w c Time Height1 Station Sample Id (min) (days) (cm) 06-48-21-5a 2107 1.46 6.24 06-48-23-1 b 28 0.02 11.51 06-48-23-2b 534 0.37 10.82 06-48-23-3b 1314 0.91 10.18 06-48-23-4b 1793 1.25 9.66 06-48-23-5b 2058 1.43 9.05 07-46-27-1 a 85 0.06 11.61 07-46-27-2a 745 0.52 11.08 07-46-27-3a 1529 1.06 10.34 07-46-27-4a 2005 1.39 9.81 07-46-27-5a 2270 1.58 9.29 07-46-25-1 b 64 0.04 11.61 07-46-25-2b 724 0.50 11.06 07-46-25-3b 1507 1.05 10.53 07-46-25-4b 1984 1.38 10.01 07-46-25-5b 2249 1.56 9.48 08-51-24-1 a 77 0.05 10.11 08-51-24-2a 582 0.40 8.77 08-51-24-3a 1647 1.14 6.60 08-51-24-4a 2397 1.66 4.93 08-51-24-5a 2975 2.07 3.92 08-51-X-1b 41 0.03 9.31 Observed Concentration (m) P04 (pM) N03 (pM) Si (pM) Ge (pM) 0.06 3.720 12.522 138.3 59.3 0.12 2.086 24.591 41.7 - 0.11 2.214 22.313 67.0 - 0.10 2.667 19.669 98.8 - 0.10 2.895 17.803 118.1 - 0.09 2.839 16.870 128.8 - 0.12 2.685 29.259 52.0 34.6 0.11 3.578 27.191 75.1 44.7 0.10 3.979 23.515 98.8 51.5 0.10 4.512 21.578 115.3 53.6 0.09 4.069 20.154 125.3 55.4 0.12 2.623 29.639 51.6 - 0.11 3.307 25.610 80.3 - 0.11 4.608 18.277 110.9 - 0.10 3.959 20.846 129.9 - 0.09 4.841 16.380 139.9 - 0.10 3.131 38.138 109.1 80.7 0.09 3.323 36.366 126.4 83.2 0.07 3.343 26.946 175.6 105.4 0.05 3.269 18.606 216.0 115.0 0.04 3.498 12.964 252.1 123.9 0.09 2.954 35.650 107.5 73.9 'si M Table 3-2. Summary o f Core Incubation Data an d Fluxes. c o T O L- C 0) o c o O ■ o 0) © i n s i O W in o C O 0 0 C D C M T - C O o C D 0 0 T - 0 0 C M v C M N - d in C O C O in o > d C M C M o C D X — G > T f - in C D C O 00 T — t j - C O h- in C O h- C D O in h * . r ^ - C D C D C O o T“ C M C M X~ T * “ X — v T~ x- X “ C M T * “ T _ T — C M C M C M C M _ E O ) o in 5 H T O ■ o II o a E C O co c o + 5 C O *-* <0 O x - x - 00 a > c d c \i c o n- a > o o in O C O O) C O G O O 00 x - in h - S r (O t- C O C O C M C M C M h - « ^ M l- C O C M x r - N . C O C M x r - G O C O 00 C O C M O) C O 00 N o o o o d o d o S U) C M O N C M N C M C O C O N N 00 C M ^ ^ C O r (D O O x “ x-* C M C O 0 5 T “ C O C O in C O C O C D T “ C M C M J O J O C M C O J O m X X X X in in in lo i i i i 0 0 0 0 00 c o o o o o r O r C O C O o c o co t j* in -r- T - C M C M C O O C O 05 C M C M X — C O C O C M h-. in N- C M 00 T~ in in C D C M C D h- h- T — C D C M o 00 in C O q 00 C O C M C M C O C D T— 00 in C O T f C M d d T — C D c s i 00 00 C O 00 cd 00 Tj- 'M ’ '^r - M - tT C O ■ x t C O C O C O C O C O C O C O co N- C D tj- C O C O C M C M in 00 ^r h- N* C O O h - C M in in O O in 00 05 r — C O C M C D C M T~ C O C M q O C M o x — h - C O C O G) C M C O C O C O C O C O C O C O C O 00 C M cd cd cd C M O C D C D 00 x — O C D C D h - C O C O m in T — O q o T — v - T - o O q o o o o d d d d o d d d d o d d d d d v 00 o 00 in co C O 00 C O 00 C O co in q ■ M * 00 C O q in o Tj- C D 00 C M h- C M h^ d C D C D 00 00 x - ^ d o C D 00 cd cd in in X — v x — C M - M - x f r C D co 00 O h - N- 00 C D q m C D 'Xt a > O io C D Tj; q o in q o C M d o d X “ “ x - d d d X “ o d d x — x - C M Tj- C D X - r r v m o O m in o in O C M C O h - in tj- C M co in C O C O C D C M X — h- h - C O T — h - h - C O x — h - 00 ■ x t m r ^ . C M C M T — C M C M T — x — T — C O C O C O C O C O _D J D .Q J O JO C O c o C O C O C O X- C M C O Tj- in x — C M C O in x — C M C O in T — Tl C O C O C O C O C O in in id in in C M C M C M C M C M C M C M C M C M C M C M C M C M C M C M N- r i r i hi. hi. hi- h^ r i r i C D C D C D C D C D in in in in in in m in in m C O q C O q q C D C D C D C D C D C D C D C D C D C D o o d d o o o o o o o o o o o x — T - X— X“ C D 73 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10-69-WCS-1b 7 0 0 .05 8.31 0.08 2.732 38.459 181.7 121.4 10-69-WCS-2b 7 9 5 0.55 7.78 0.08 2.875 38.761 197.8 135.3 10-69-WCS-3b 1389 0.96 7.26 0.07 2.658 38.331 210.2 132.0 Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3-2. Summary of Core Incubation Data and Fluxes, (cont.) WC Time Height1 Observed Concentration Station Sample Id (min) (days) (cm) (m) P04 (pM) N03 (uM) Si (uM) Ge (uM) 10-69-WCS-4b 1545 1.07 6.73 0.07 2.462 36.496 213.0 133.1 10-69-WCS-5b 1715 1.19 6.21 0.06 2.864 38.519 217.6 135.1 1. Water column height represents height prior to drawing sample. -vj STATION 6 Morro Bay 6 Core B 24 Ertor| 0.2907' Chisq 20 Error Chisq NA 16 0 2 4 6 8 10 12 14 16 days/m 4.0 3.5 | 3.0 O CL 2.5 2.0 0 2 4 6 8 10 12 14 16 Days/m y = m1+m2*M0+m3*MQ*M0 ftQQQR4 y - M1+M2*M0 STATION 6 Morro Bay 6 Core B days/m Figure 3-3. Flux Plot for Core Incubation. Example of a typical flux plot for nutrients (N 0 3, P 0 4, and Si). Best representation for fits were exponential curve for N 0 3, linear for P 0 4, and quadratic for Si. 75 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. obvious. Consequently, a linear fit was used for phosphate. For silicate, some curvature was observed and quadratic fits were used. Fluxes of each nutrient for each core are given in Table 3-3. These have been averaged and the uncertainty taken as the larger of (a) the standard deviation of the mean calculated from replicate results, or (b) the standard deviation of the mean calculated from the uncertainties of the two results for each station. Overall, NO3 fluxes show removal of NO3 (negative flux) from the water column, reflecting denitrification. The exception is station 10 which shows a small positive flux, although the uncertainty exceeds the average flux. A single epibenthic foram and a worm burrow was observed in one of the two cores for station 1 0 , while no evidence of animals was visible in the other core, suggesting that the environment was sub-oxic. Overall, P04 fluxes show a positive flux from the sediments into the water column, with the exception of station 3 (Catalina Basin). This station shows a small negative flux (-0 .0 1 ± 0 .0 0 ) of PO4 from the water column into the sediments. The uncertainty for the analysis of this station is quite small, suggesting that there may be removal of PO4 at this site. At stations 6 and 7 (shallow water, Morro Bay) the highest fluxes are observed. Overall, Si fluxes show a positive flux from the sediments into the water column. The largest fluxes are from stations 6 and 7 (shallow water, Morro Bay). 76 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-3. Incubation Flux Data. Nitrate Fluxes Core Linear Fit Exponential Fit average of Cores sdom sdom* 1A -0.429 ± 0.069 -0.585 + 0.077 -0.57 0.01 0.05 1B -0.442 ± 0.070 -0.558 ± 0.072 2A -0.469 ± 0.052 -0.669 ± 0.098 -0.65 0.02 0.07 2B -0.403 ± 0.080 -0.637 ± 0.096 3A 0.051 ± 0.042 -0.231 + 0.044 -0.24 0.01 0.03 3B -0.231 ± 0.023 -0.248 ± 0.024 4A -0.232 ± 0.020 -0.252 ± 0.023 -0.24 0.01 0.02 4B -0.021 ± 0.032 -0.229 ± 0.033 6A -0.545 ± 0.023 -0.738 ± 0.031 -0.67 0.07 0.02 6B -0.491 ± 0.029 -0.598 ± 0.020 7A -0.564 ± 0.017 -0.669 + 0.023 -0.88 0.21 0.01 7B -0.823 ± 0.035 -1.087 ± 0.011 8A -0.499 ± 0.048 -0.816 ± 0.031 -0.72 0.09 0.07 8B -0.556 + 0.086 -0.633 ± 0.128 9A -0.036 ± 0.040 -0.07 0.04 0.05 9B -0.110 ± 0.091 10A 0.015 ± 0.019 0.01 0.01 0.01 10B -0.004 ± 0.016 Phosphate fluxes Linear uncert average Core fit in fit of Cores sdom sdom* 1A 0.065 ± 0.013 0.047 0.019 0.007 1B 0.028 ± 0.003 2A 0.026 ± 0.002 0.034 0.008 0.003 2B 0.041 ± 0.006 3A -0.007 ± 0.003 -0.006 0.000 0.002 3B -0.006 ± 0.002 4A -0.008 ± 0.005 0.000 0.008 0.003 4B 0.008 ± 0.002 6A 0.078 ± 0.012 0.067 0.011 0.007 6B 0.055 ± 0.008 7A 0.091 ± 0.027 0.115 0.024 0.014 7B 0.139 ± 0.006 8A 0.005 ± 0.002 0.000 0.005 0.004 8B -0.004 ± 0.009 9A 0.012 + 0.005 0.009 0.003 0.004 9B 0.005 ± 0.005 10A 0.039 ± 0.007 0.021 0.018 0.006 10B 0.003 ± 0.009 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Table 3-3. Incubation Flux Data, (continued) Si fluxes Quadratic uncert average Core fit in fit of Cores sdom sdom* 1A 1.883 ± 0.464 2.058 0.175 0.271 1B 2.233 ± 0.281 2A 1.419 ± 0.074 1.552 0.133 0.077 2B 1.685 ± 0.134 3A 0.805 ± 0.072 0.781 0.024 0.057 3B 0.756 ± 0.088 4A 2.515 ± 0.157 2.569 0.054 0.106 4B 2.624 ± 0.142 6A 5.983 ± 0.403 6.913 0.930 0.246 6B 7.844 ± 0.284 7A 5.518 ± 0.309 6.467 0.949 0.173 7B 7.416 ± 0.157 8A 4.279 + 0.238 3.690 0.590 0.119 8B 3.100 ± 0.012 9A 2.608 ± 0.064 2.707 0.100 0.080 9B 2.807 ± 0.146 10A 3.018 ± 0.047 2.939 0.079 0.078 10B 2.860 ± 0.149 Note: SDOM was calculated from agreement of replicate cores. SDOM* was predicted from uncertainties in flux measurement for each core. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Although core pairs (A and B) came from the same location, there was spatial heterogeneity observed in the macrofauna. This suggests that there may be some spatial variability in the fluxes observed in pairs from each site. Nitrate uptake in replicate cores varied by about 10%, except at stations 9 and 10, where variability was much higher (70-100% ), due to low uptake rates. Phosphate fluxes were much smaller than nitrate and exhibited greater variability between replicate cores (30% average), except at station 10 where fluxes were small and variability was about 85% . Silica fluxes were larger than nitrate and phosphate. Variability between replicate cores is less than 10%, except for stations 6 , 7, and 8 , where variability is about 15%. Benthic Lander (In situ) and Incubation Comparison Results from comparison of benthic lander (in situ) and core incubation (shipboard) generated fluxes are given in Table 3-4 and plotted in Figure 3-4. If shipboard incubation is a good representation of in situ methods for measuring fluxes, then the results of each method should be identical. Because in situ results should be free of artifacts due to pressure and temperature changes, these are taken as the standard. Figure 3-4 shows the incubation and the benthic chamber (in situ) fluxes for each station plotted against one another. A 1:1 line has been added, to represent the 79 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Reproduced w ith permission o f th e copyright owner. Further reproduction prohibited without permission. Table 3-4. Comparison of Benthic Lander and Core Incubation Fluxes. Flux results in mmo/m 2/d. ______________________ NQ3 Flux ______________________ P04 Flux _______________________Si FLUX Station_______ Location________ Core Inc._________Lander______ Core Inc. Lander Core Inc. Lander 1-7 San Pedro -0.57 ± 0.05 -1.24 ± 0.02 0.05 ± 0.02 0.09 ± 0.02 2.06 ± 0.27 2.82 ± 0.05 2-4 Santa Monica -0.65 ± 0.07 -1.34 ± 0.11 0.03 ± 0.01 0.12 ± 0.01 1.55 ± 0.13 2.16 + 0.13 3-23 Catalina -0.24 ± 0.03 -0.38 ± 0.04 -0.01 + 0.00 0.01 ± 0.01 0.78 ± 0.06 0.97 ± 0.01 4-19 Tanner -0.24 ± 0.02 -0.47 + 0.10 0.00 + 0.01 0.01 + 0.01 2.57 ± 0.11 2.24 ± 0.27 6-48 Morro Bay 6 -0.67 ± 0.07 -1.23 ± 0.06 0.07 ± 0.01 0.14 ± 0.04 6.91 ± 0.93 13.75 ± 0.59 7-46 Morro Bay 7 -0.88 ± 0.21 -1.34 ± 0.06 0.11 ± 0.02 0.11 ± 0.02 6.47 + 0.95 9.25 ± 0.29 8-51 Sta 8/02 min -0.72 + 0.09 - 0.00 + 0.01 - 3.69 ± 0.59 - 9-57 Sta 9/1500m -0.07 ± 0.05 -0.15 ± 0.02 0.01 + 0.00 0.00 ± 0.00 2.71 ± 0.10 2.04 ± 0.24 10-69 Sta 10 0.01 ± 0.01 -0.21 ± 0.03 0.02 ± 0.02 0.02 ± 0.00 2.94 ± 0.08 3.13 ± 0.00 STATION Ge/sr LOCATION Core Inc. Lander 1-7 San Pedro 0.30 ± 0.12 0.33 ± 0.04 2-4 Santa Monica 0.16 ± 0.10 0.34 ± 0.21 3-23 Catalina 0.29 ± 0.30 - - 4-19 Tanner 0.69 ± 0.07 1.20 ± 0.19 1.03 ± 0.24 6-48 Morro Bay 6 0.31 ± 0.01 0.41 + 0.06 0.39 ± 0.06 7-46 Morro Bay 7 0.28 ± 0.04 0.35 ± 0.06 8-51 Sta 8/02 min 0.48 ± 0.04 - - 9-57 Sta 9/1500m 0.50 ± 0.07 1.19 ± 0.61 10-69 Sta 10 0.32 ± 0.14 - - 1. The expected ratio of Ge/Si in oceans is 0.70. Note: Core incubation average is the mean of 2 replicates ± the SDOM or the uncertainty based on fitting, which ever is larger. Lander average is based on the mean of 2 or 3 chambers. o o o 1.00 NO, Flux 3 Core Incubation v. Lander 0.50 sta. 10 U . 0.00 sta. 3 sta. 9 sta. 4 sta.1 sta. 2 sta. 7.-1-00 sta. 7 -1.50 -2.00 o.oo 0.50 -1.00 -0.50 -1.50 Lander Flux x 3 II. C _ o r e . a s o c o o -5— i— i — I — I — I — . — r PO Flux Core Incubation v. Lander (b) sta. 7 sta. 6 sta. 10 sta. 2 sta. 9 sta. 3 -0.05 0.00 0.05 0.10 0.15 0.20 Lander Flux Si Flux Core Incubation v. Lander sta. 7 sta. 4 sta. 10 sta. 1 £ 2.0 sta. 2 sta. 3 -2.0 0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0 Lander Flux Figure 3-4. Incubation v. Lander (In Situ) Fluxes. The line through the plot represents a one to one slope. This would be the trend of the data if the two methods were identical. 8 1 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. trend of fluxes that should be observed if the two methods are comparable. In general, the core incubation fluxes are smaller than in situ results. In the case of nitrate (NO3), it is clear that the incubation derived fluxes are not as negative as the in situ fluxes. In other words, the removal of NO3 from the water column to the sediments for the incubations is happening at a slower rate than the in situ removal (core incubation flux is lower). Stations 10, 9, 4, and 3 fall much closer to the 1 to 1 line than the rest of the stations. However, the uncertainty in the incubation sample analysis is fairly large and the error bars for 5 of the analyses cross the 1 to 1 line, suggesting that although the results are low for the incubation they may not be greatly different from the in situ fluxes. In the case of phosphate (PO 4), the majority of the incubation derived fluxes are lower than the in situ derived fluxes. There are, however, two stations where in situ and incubation fluxes are the sam e (stations 7 and 10). Station 7 is shallow with many animals, and station 10 is deep with very few animals. The uncertainty for the station 10 incubation is somewhat large compared to the analysis (0.02 ±0.03). Station 9 has a higher incubation derived flux and a very small uncertainty. The station 4 incubation flux (0.00 ±0.01) also falls close to the 1 to 1 line, and its uncertainty is reasonably small suggesting that it is not significantly different than the in situ derived flux. 82 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. In the case of Si, incubation derived fluxes are often smaller than in situ fluxes (Figure 3-4 (c)), although stations 4 and 9 have slightly higher fluxes. Stations 10 and 3 have fluxes equal to the in situ derived fluxes. Stations 6 and 7 have incubation derived fluxes that are lower than the in situ derived fluxes. Discussion Possible artifacts There were some systematic temperature differences between in situ and core incubation conditions. The largest effect is observed for stations 6 and 7, where incubation temperature was about 5°C colder than in situ. A 5°C decrease reduces diffusivity by about 12% (Li and Gregory, 1974). A 5°C decrease also reduces pore water Si by 12% (McManus et al., 1995). Thus, combining these effects should produce a flux at stations 6 and 7 that is about 25% low. Similar effects may occur for NO3 and PO4, although reaction kinetics may have a different temperature dependence than Si. At station 10, incubation temperature was about 5°C warmer than in situ, and thus these fluxes from cores may be 25% too high. Lander fluxes at station 8 are almost certainly too low because a rapid loss of the Cs/Br spike indicated that chambers appear to have leaked (Hammond, pers. comm.). The sediments were relatively coarse grained, and easily subjected to winnowing and strong boundary currents. It is likely 83 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. that chamber penetration was poor. These were deleted from the comparisons. It should be noted that over the course of the incubation, the oxygen supply that animals would use is decreasing with time. Most animals captured in cores appeared to be dead or very sluggish by the end of the incubation. This may be due to the consumption of the finite oxygen supply or possibly due to the change in pressure and or temperature. Oxygenation of cores is another possibility. Although the overlying water in the cores was not exposed to the atmosphere for more than about a minute (60 seconds), there may have been some small bubbles that were trapped in the cores and possibly increased the oxygen content in the water column. This was most likely a concern at stations 1, 2 and 4. An addition of oxygen might cause the nitrate uptake to be reduced. Some bacteria use nitrate as an electron acceptor when oxygen is not available. If oxygen were to become available through the addition of small bubbles trapped in the core, these bacteria would stop using the nitrate and consume the oxygen, thus reducing the nitrate uptake. Alternatively, bacterial activities may have been altered due to temperature and pressure changes. Pressure Effects W hen cores are removed from their in situ environment to the surface of the ocean, the pressure difference has the potential to alter the 84 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. metabolism of any organisms present. A few studies have done shipboard incubations and in situ flux measurements at the same stations. Studies of shallow water sites (most <100 m) in the Adriatic Sea, indicated reasonable agreement (±30% with no systematic bias) between O 2 fluxes measured with incubated cores and in situ landers (Tahey et al., 1996; Giordani et al., 1998). Rowe et al., (1997), observed a shipboard oxygen consumption that was two times greater than in situ consumption on the Greenland shelf, suggesting that shipboard incubations overestimate fluxes. It should be noted that the depths of these cores were relatively shallow (less than 500 m). Wijsman, J. W . M., (2001), compiled a summary of deck incubations that were a factor of 2 higher than in situ incubations at depths greater than 1500m. However, these were not from simultaneous samplings. On the other hand, at greater depths (800-1500 m) it has been observed that reducing pressure can significantly decrease metabolic activity of bacteria, as indicated by 3H-thymidine incorporation, used as an indicator of productivity (Tholosan et al., 1999; Deming and Colwel, 1985). Effects were quite variable on different expeditions, but typically showed a 20-70% decrease. This decrease was not strongly depth dependent, between 800 and 2000 m. In summary, conflicting results exist for the comparison of shipboard incubation and lander O 2 consumption. No literature has been found that actually compares shipboard incubation and in situ fluxes for nutrients with simultaneous sampling, as was done in this study. The incubation 85 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. (shipboard) fluxes, for this study, in general are about 20-50% lower than in situ fluxes. Germanium sequestration: There appears to be a strong fractionation of Ge at stations 1,2,3, (San Pedro, Santa Monica, and Catalina basins). This is not surprising due to the Fe-rich sediments and previous observations by Hammond et al. (2000). The lander data also supports strong fractionation at stations 1 and 2 (no in situ data for station 3). There appears to be no fractionation at station 4 (Tanner basin), although lander data suggests a preferential release of Ge. Published results of previous work suggest this station does not significantly fractionate Ge from Si (Hammond et al., 2000). Stations 6 and 7 both show strong fractionation of Ge in both core incubation and lander data. The sediments at these two stations are believed to be Fe-rich. Measurements of iron in pore waters have not yet been made, but this work will be undertaken by colleagues at the University of Minnesota. Stations 8 and 9 show a small fractionation for the core incubation, (no in situ data for station 8 , large uncertainty for station 9). Station 10 also shows a strong fractionation of Ge in the core incubation data (no in situ data). The station 10 behavior is surprising because the oxygen penetration depth is believed to be > 1 cm at this station, in which case the oxidation front would be quite deep preventing Fe+2 from being very effective in sequestration of Ge. 86 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Conclusion The core incubation method for measuring fluxes appears to provide results that are comparable to measuring nutrient fluxes in situ, although they appear to be systematically low. However, temperature control is essential. The effect of temperature on fluxes was roughly estimated at stations where deck incubation temperature differed from in situ values. After these approximate temperature adjustments are made, the incubations still appear to have a systematic offset of flux, appearing about 25-30% too low. This off set is about the same for all nutrients (NO 3, P 0 4, Si). The fractionation of Ge/Si flux was observed in San Pedro, Santa Monica, and Catalina basins where it was expected due to the presence of near surface Fe-rich anoxic sediments. The difference between the core incubation and in situ Ge/Si flux is not significantly different. The core incubation method does appear to be a reasonable method for measuring ratios of Ge/Si benthic fluxes. 87 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. References Applin, K. R., 1987. The diffusion of dissolved silica in dilute aqueous solution. Geochimica et Cosmochimica Acta, vol. 51, pp. 2147-2151. Berelson, W. M., and Hammond D. E., 1986. The calibration of a new free- vehicle benthic flux chamber for use in the deep sea. Deep-Sea Research, vol. 33, no. 10, pp. 1439-1454. Berelson, W. M., McManus, J., Coale, K. H., Johnson, K. S., Kilgore, T., Burdige, D., and Pilskaln, C., 1996. Biogenic matter diagenesis on the sea floor: A comparison between two continental margin transects. Journal of Marine Research, vol. 54, pp 731-762. Berelson, W. M., Hammond, D. E., Johnson, K. S., 1987. Benthic fluxes and the cycling of biogenic silica and carbon in two southern California borderland basins. Geochimica et Cosmochimica Acta, vol. 51, pp. 1345-1363. Crank, J., 1975. The Mathematics of Diffusion, second edition. Clarendon Press, Oxford. 414 pp. Deming, J. W ., and Colwell, R. R., 1985. Observations of barophilic microbial activity in samples of sediment and intercepted particulates from the Demerara abyssal plain. Applied Environmental Microbiology, vol. 50, pp1002-1006. Fanning, K. A., and Pilson, M. E. Q., 1974. The diffusion of dissolved silica out of deep-sea sediments. Journal of Geophysical Research, vol. 79, no. 9, pp. 1293-1297. Froelich, P. N., and Andreae, M. O., 1981. The marine geochemistry of germanium: Ekasilicon. Science, vol. 213, pp. 205-207. Froelich, P. N., Hambrick, G. A., Andreae, M. O., and Mortlock, R. A., 1985. The geochemistry of inorganic germanium in natural waters. Journal of Geophysical Research, vol. 90, no. C1 pp. 1133-1141. Garrels, R. M., and Christ, C. L., 1965. Solutions. Minerals, and Equilibria. Harper and Row, New York. 88 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Giordani, P., Hammond, D. E., Balboni, V., Miserocchi, S., Malaguti, A., and Poletti, R., 1998. Benthic-Pelagic coupling in the Adriatic Sea: Studies of carbon and nutrient cycling in the EU RO M ARG E-AS Project, in Hopkins, T., Artegiani, A., Cauwet, C., Degobbis, D., and Malej, A. (eds.) Ecosystems Research Report, the Adriatic Sea, EU/Enivironment Series, Brussels, Journal of Marine Systems. Hammond, D. E., McManus, J., Berelson, W . M., Kilgore, T. E., and Pope, R. H., 1996. Early diagenesis of organic material in equatorial Pacific sediments: Stoichiometry and kinetics. Deep-Sea Research II, vol.43, no. 4-6, pp. 1365-1412. Hammond, D. E., McManus, J., Berelson, W . M., Meredith, C., Klinkhammer, G. P., and Coale, K. H., 2000. Diagenetic fractionation of Ge and Si in reducing sediments: The missing Ge sink and a possible mechanism to cause glacial/interglacial variations in oceanic Ge/Si. Geochimica et Cosmochimica Acta, vol. 64, no. 14, pp. 2453-2465. Hurd, D. C., 1972. Interactions of biogenic opal, sediment and seawater in the Central Equatorial Pacific. Geochimica et Cosmochimica Acta, vol. 37, pp. 2257-2282. Jahne, B., Heinz, G., and Deitrich, W ., 1987. Measurement of the diffusion coefficients of sparingly soluble gases in water. Journal of Geophysical Research, vol. 92, pp. 10,767-10,776. King, S. L., Froelich, P. N., and Jahnke, R. A.,. 2000. Early diagenesis of germanium in sediments of the Antarctic South Atlantic: In search of the missing Ge sink. Geochimica et Cosmochimica Acta, vol. 64, no. 8, pp. 1375-1390. Li, Y. H., and Gregory, S., 1974, Diffusion of ions in sea water and in deep- sea sediments. Geochimica et Cosmochimica Acta, Vol. 38, pp. 703- 714. McManus, J., Hammond, D. E., Berelson, W . M., Kilgore, T. E., DeMaster, D. J., Ragueneau, O. G., and Collier R. W ., 1995. Early diagenesis of biogenic opal: Dissolusion rates, kinetics and paleoceanographic implications. Deep-Sea Research II, vol. 42, no. 2-3, pp. 871-903. Mills, R. and Lobo, V. M. M., 1989. Self-diffusion in electrolyte solutions. Physical Sciences Data 3 6 . Elsevier Science Publishing Company, Inc., NY. 89 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Mortlock, R. A and Froelich, P. N., 1996. Determination of germanium by isotope-dilution hydride-generation inductively-coupled mass spectrometry. Analytics Chimica Acta, vol. 332, pp. 645-652. Mortlock, R. A., Charles, C.D., Froelich, P. N., Zibello, A., Saltzman, J., Hays, J. D., and Burckle, L. H., 1991. Evidence for lower productivity in the Antarctic Ocean during the last glaciation. Nature, vol. 351, pp. 220- 223. Murnane, R. J., and Stallard, R. F., 1988. Germanium/silicon fractionation during biogenic opal formation. Paleoceanography, vol. 3, no. 4, pp. 461-469. Murnane, R. J., Leslie, B., Hammond, D. E., and Stallard, R. F., 1989. Germanium geochemistry in the Southern California Borderlands. Geochimica et Cosmochimica Acta, vol. 53, pp. 2873-2882. Murnane, R. J., and Stallard, R. F., 1988. Germanium/silicon fractionation during biogenic opal formation. Paleoceanography, vol. 3, no. 4, pp. 461-469. Renalli, Giorgio, Rheology of the Earth, 1987, p. 149, eqn. 7.11. Allen and Unwin, Boston , Massachusetts. Robinson, R. A., and Stokes, R. H., 1959. Electorlvte Solutions. The Measurement and Interpretation of Conductance. Chemical Potential and Diffusion in Solutions of Simple Electrolytes, second edition (revised). Butterworths, London. 571pp. Rowe, G. T., Boland, G. S., Escobar Briones, E. G., Cruz-Kaegi, M. E., Newton, A., Piepenberg, D., Walsh, I., Deming, J., 1997. Sediment community biomass and respiration in the northeast water Polynya, Greenland: a numerical simulation of benthic lander and spade core data. Journal of Marine Systems, vol. 10, pp. 497-515. Santschi, P. H., Bower, P., Nyffeler, U. P., Azevedo, A., and Broecker, W. S., 1983. Estimates of the resistance to chemical transport posed by the deep-sea boundary layer. Limnology and Oceanography, vol. 28, no. 5, pp. 899-912. Sayles, F. L., Deuser, W . G., Goudreau, J. E., Dickinson, W . H., Jickells, T. D., and King, P., 1996. The benthic cycle of biogenic opal at the Bermuda Atlantic Time Series site. D eep-Sea Research I, Vol. 43., No 4., pp. 383-409. 90 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Shemesh, A., Mortlock, R. A., Froelich, P. N., 1989. Late Cenezoic Ge/Si record of marine biogenic opal: implications for variations of riverine fluxes to the ocean. Paleoceanography, vol. 4, no. 3, pp. 221-234. Strickland, J. D. H. , and Parsons, T. R., 1968. A Handbook of Seawater Analysis. Bulletin 167, Fish. Res. Board Can. (Ottawa). Tahey, T. M., Duineveld, G. C. A., DeWild, P. A. W . J., Berghuis, E. M., and Kok, A., 1996. Sediment 0 2 demand, density and biomass of the benthos and phytopigments along the northwestern Adriatic coast: the extent of the Po enrichment. Oceanologica Acta, vol. 19, pp. 117-130. Taylor, J. R., 1982. An Introduction to Error Analysis, second edition. University Science Books, Sausalito, California. 327 pp. Tholosan, O., Garcin, J., Bianchi, A., 1999. Effects of hydrostatic pressure on microbial activity through a 2000 m deep water column in the NW Mediterranean Sea. Marine Ecology Progress Series, vol 183, pp. 49- 57. Treguer, P., Nelson, D. M., Van Bennekom, A. J., DeMaster, D. J., Leynaert, A., Queguiner, B., 1995. The silica balance in the world ocean: a reestimate. Science, vol. 268, pp. 375-379. Wijsman, J. W . M., 2001. Early diagenetic processes in northwestern Black Sea sediments. Dissertation, Rijksuniversiteit Grohihgen, pp. 7-21. Wollast, R. and Garrels, R. M., 1971. Diffusion coefficient of silica in seawater. Nature Physical Science, Vol. 229, p. 94. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix A Numerical Simulation of Diffusivity Cell Description of Model A computer model using Fick’s laws for diffusion, which relates flux to the concentration change over time, by a constant (D m ) that is specific to the species, is used to find a best fit value of diffusivity based on the observed concentrations of the experiment through time. • The model is a fortran program that uses a one dimensional diffusion equation to calculate a series of concentration values at given time intervals for various diffusivities (D*). • The program is first order over time and second order over space. • The program then performs a sum of squares (% 2) using the calculated and observed values, which is then plotted against the corresponding D* values. • The D* value correlating with the smallest chi2 is determined to be the best fit. • To remain true to the diffusive behavior governed by Fick’s laws for diffusion, the model uses the first observed value (taken about 15 min. after the start) to back calculate a starting concentration in the sample reservoir. 92 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Equations Used in the Numerical Simulation First Box (sample Reservoir) at t > 0: C r ~ C r ■ * " 2 ' A<j)D*dt v V rA * Boxes within the frit at t > 0: v Sx '^{O x-1 ■ * " Qtx+1 Initial condition at start of experiment (t = 0): C r = some value > 0 Cnx = 1 C ’r = the concentration in the sample reservoir at any t greater than 0. C r = the concentration in the sample reservoir at the previous time step. Ci = the normalized starting concentration of the diffusing species in the frit. A = Area of the porous frit material. < | ) = porosity of the frit material. D* = Diffusion Coefficient of the diffusing species within the frit, dt = incremental time step. V R = Volume of the sample reservoir. dx = box thickness over which species diffuses. nx = box number. 93 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Fortran Code Used for the Numerical Simulation Diffusion Sub-routine Code c================================================= c diffuseld: do 1D diffusion for one time step: c diffuse1d(C,diffusivity,nx,dx,dt,Cnew) c NOTE: c concentration is incoming and outgoing values, c diffusivity defines physical model c model parameters are nx,dx, and dt. c c Diffusion is patterned after thermal conduction which is a c diffusive process: c Concentration == temperature c diffusivity == diffusivity = conductivity/(spec c heat*density) c c Make sure units of diffusivity, conductivity, dx and dt are c consistent. c Boundary conditions not dealt with within subroutine - must do c externally c================================================= c solve Heat conduction equation from Ranalli's Rheology c of the Earth, 1987, pg 149, eq. 7.11 c c heat: dT/dt = (Kconduct/cspec_heat*dens) dA 2T/dxA 2 c c diffusion: dC/dt = (diffusivity) dA 2C/dxA 2 c================================================= c 20oct00 modify from fd_latentHeat6.f c 26jan01 modify from conductld.f c================================================= subroutine diffuse1d(C,diffusivity,nx,dx,dt, + porosity,Area,Vo,Cwork) real C(nx),diffusivity(nx),Cwork(nx) real dx,dt real porosity,Area,Vo integer nx 94 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. c nx is the number of boxes. c clear outgoing values of concentration do ix=1,nx Cwork(ix) = 0. enddo c loop over spatial points do ix = 2,nx -1 Fo = diffusivity(ix) * dt/(dx*dx) Cwork(ix) = ( C(ix-1) + C(ix+1) - (2. -1 ./F o )*C (ix )) * Fo enddo c do Boundary Conditions: C(1) and C(nx): F = 2 * (porosity*Area*diffusivity(1)*dt)/(Vo*dx) Cwork(1) = C(1) + (F * (C(2) - C(1))) Fo = diffusivity(nx) * dt/(dx*dx) Cwork(nx) = C(nx) + (Fo * (C(nx-1) - C(nx))) c transfer updated concentration back into original array do ix=1,nx C(ix) = Cwork(ix) enddo call checkexp(C,nx) RETURN end c================================================== c subroutine to stop program if diffusivity is too large. subroutine stability1d(diffusivity,nx,dx,dt) real diffusivity(nx) integer nx, ix xmin = diffusivity(l) * dt/(dx*dx) xmax = diffusivity(l) * dt/(dx*dx) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. do ix=1 ,nx Fo = diffusivity(ix) * dt/(dx*dx) if(Fo .LT. xmin)xmin=Fo if(Fo .GT. xmax)xmax=Fo if(Fo .GT. 0.5)then write(6,100)ix,Fo stop elseif(Fo .EQ. 0.)then write(6,200)ix endif enddo write(6,300)xmin,xmax return 100 format('DIFFUSION STABILITY FAILED> Node',i8; Fo=',f10.4,/, + 'DIFFUSIO N STABILITY > Fo should be < 0.5',/, + 'DIFFUSIO N STABILITY > decrease dt or increase dx') 200 form at('DIFFUSION STABILITY FAILED> Fo = 0. at node ',i8) 300 format('DIFFUSION STABILITY> Fo ranges between',f10.4,' and',f10.4) end c================================================== c subroutine to test for extreme concentrations. subroutine checkexp(array,nx) real array(nx) exponthresh = 32. exponthresnm = -exponthresh do ix=1,nx if(x .NE. 0.)then x = alog10(abs(array(ix))) if(x .GT. exponthresh )write(6,100),ix if(x .LT. exponthreshm)write(6,200),ix endif enddo 100 format('CHECK W ARNING > concentration exponent > 10A32 at ix=',i10) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 200 format('CHECK W ARNING > concentration exponent < 10A -32 at ix=',i10) return end c================================================== Diffusion Cell Simulation Code c================================================ c 1-D test of diffusion (spookychi2dVol.f) c 26jan01 initial installation c 01feb01 put in proper boundary conditions and use real c parameters. c 31may01 print actual concentration diffusion as a function of c time. c 22jun01 put in mid-stream volume change c 14nov01 skip chisum calculation for bad data points (but c reduce volume anyway). c================================================ c Declarations real C(5000),Cwork(5000), diffusivity(5000), withdraw(25) integer getpar, itwithdraw(25) real Cmeasured(25),Ccalculated(25) real finaldiffuse(1000),finalchisum(1000) real Cleftmost(15000) c================================================ c (A) Retrieve input parameters which regulate modeling. c define default values diffuse= 0.00000137 CO = 0. C1 = 1. nt = 0 nx = 51 dt = 30. dx = 0.08 97 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. porosity= .42 Area = 3.81 Vo = 0. i1 = -1 \2 = -1 id = -1 idprint = -1 n withdraw = -1 do i=1,25 withdraw(i)=-1. enddo dvolumeO = 0.8 dvolum el = 0.8 DiffO = 0 .0 0 0 0 0 0 7 Diffl = 0 .0 0 0 0 0 5 0 nDiff = 5 0 do i=1,25 Cmeasured(i)=0. Ccalculated(i)=0. enddo c retrieve command-line values call setpar() ichk = getpar("dt","f',dt) ichk = getpar("dx","f",dx) ichk = getpar("nx","d",nx) ichk = getpar("nt","d",nt) ichk = getpar("diffusivity","fI,diffuse) ichk = getpar("C0","f,,C0) ichk = getpar("C1",T,C 1) ichk = getpar("porosity","f",porosity) ichk = getpar("Area","f,,Area) ichk = getpar( Vo","f',Vo) nwithdraw = getpar("times","vf',withdraw) ichk = getpar("dvolume0","f',dvolumeO) dvolum el = dvolumeO Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ichk = getpar("dvolume1","f',dvolumel) ichk = getpar("DiffO",T,DiffO) ichk = getpar("Diff1","f,,Diff1) ichk = getpar("nDiff","d",nDiff) nCmeasured = getparfCm easured'V'vf'.Cm easured) id = (i2+i1 +1)/10 ichk = getpar("i1","d",i1) ichk = getpar("i2",,,d",i2) ichk = getpar("id","d",id) ichk = getpar("idprint","d",idprint) call endpar() c check validity of values if(nt .EQ. 0 .OR. Vo*diffuse .EQ. 0)then write(6,10) write(6,11) write(6,12) stop endif if(nCmeasured .NE. nwithdraw)then write(6,2) write(6,10) write(6,11) write(6,12) stop endif 2 format('ERROR> Number of C_measure= does not equal tim e s -) 10 format('spookychi2 nx= nt= dx= dt= ' + '[diffusivity= C0= C1 =]',/, +' [porosity= area= Vo=] [withdraw= dvolumeO= dvolum el =] [i1 i2= id=]',/, +' Cmeasured= DiffO= Diff1= n D iff-) 11 format( + 1 nx= number of cells in model [51]',/, + ' dx= spacing of cells in model [0.08]',/, + 1 nt= number of time steps to compute [must give]1 ,/, + ' dt= number of sec between time steps [30]',/, + ' diffusivity= diffusivity value for cells [must give]',/, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. + ' porosity= porosity value for left cell [0.42]',/, + ' area= Area of frit [3.81]',/, + ' Vo= starting Volume of left cell [must give]',/, + ' times= times in sec of volume withdrawal [off]',/, + ' dvolume0= change in volume of withdrawal (first time) [0.8]',/, + ' dvolumel = change in volume of withdrawal (later times) [0.8]',/, + ' co= initial concentration of cell #0 [0]',/, + ' C1 = initial concentration of cells 1-nx [1]') format(/, + ' idprint= print every #idprint time steps [off]',/, + ' i1 = runtime print starting at this cell [off]’,/, + ' i2= runtime print ending at this cell [off]',/, + ' id= increment between cells [nx/10]',/ + ' Cmeasured= series of Measured Concentrations + ' Diff0= diffusivity search start [0.0000007]',/, + ' Diff1 = diffusivity search end [0.0000050]',/, + ' nDiff= number of diffusivity points [50]') c print values which will be used write(6,100)nt,dt write(6,110)nx,dx write(6,120)diffuse w rite(6,122)porosity write(6,124)Area write(6,126)Vo write(6,130)C0,C1 w rite(6,132)diffuse*dt/(dx*dx) 100 format('diffuse1d> nt,dt = ',i10,f15.6) 110 format('diffuse1d> nx,dx = ',i10,f15.6) 120 format('diffuse1d> diffusivity= ’,e15.6) 122 format('diffuse1d> porosity = ',f15.6) 124 format('diffuse1d> area = ',f15.6) 126 format('diffuse1d> Vo = ',f15.6) 130 format('diffuse1d> C0,C1 = ',2f15.6) 132 format('diffuse1d> <fourier> = ',f20.10) if(nwithdraw .EQ. -1)write(6,140) if(nwithdraw .GT. -1)then do i=1,nwithdraw write(6,142)i,withdraw(i) enddo endif 100 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 140 format('diffuse1d> NO VO LU M E W ITHDRAW ALS') 142 format('diffuse1d> Volume withdrawal #, at time:1 ,i5,f15.0) write(6,160)i1,i2,id,idprint 160 format('diffuse1d> Print rows= ',2i10,/, + 'diffuse1d> skip every= ',i10,/, + 'diffuse1d> until = ',i1 0 ) 0================================================== c if volume withdrawals, determine at which time steps do i=1,nwithdraw itwithdraw(i) = ifix(withdraw(i) / dt) enddo 0================================================== delDiff = (Diffl - DiffO) / float(nDiff-l) w rite(6,1100)Diff0, Diffl ,ndiff,delDiff write(6,1200) do i=1,nwithdraw write(6,1210)i,withdraw(i),Cmeasured(i) enddo 1100 format('Diffusivity Search> start at :',e15.6,/, + 'Diffusivity Search> end at :',e15.6,/, + 'Diffusivity Search> #searches:',i15,/, + 'Diffusivity Search> increment:',e15.6) 1200 format('Measured Concentrations to match') 1210 format(' i,time (sec), C_measured:',i5,f10.0,f10.4) c================================================== C================================================== DO 8000 IDIFF = 1.NDIFF diffuse = DiffO + delDiff * (idiff-1) volume = Vo write(6,8100)IDiff,ndiff,diffuse 8100 form at('DIFFUSIVITY LOOP: ’,i8,' OF',i8,'> diffusivity=',e15.6) 8101 c===============================================— = 101 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. c (B) define physical model and initial conditions, check stability. c Set physical model for entire 1D model do ix=1,nx diffusivity(ix) = diffuse enddo write(6,200) call stability1d(diffusivity,nx,dx,dt) write(6,210) 200 format('diffuse1d> Check Stability of Model') 210 format('diffuse1d> Stability OK') call getC0(C,diffusivity,nx,dx,dt,porosity,Area,volume, Cwork,C0, + Cmeasured(1),itwithdraw(1)) c Initial conditions do ix=1,nx C(ix) = C1 enddo c now reset leftmost cell C(1) = CO c================================================ c (C) model ready to go. Loop over all requested time steps. c Loop over time do it=1,nt call diffuse1d(C,diffusivity,nx,dx,dt, + porosity,Area, volume,Cwork) Cleftmost(it) = C(1) do i=1,nwithdraw if(it .EQ. itwithdraw(i))then if(i .EQ. 1)then volume = volume - dvolumeO else volume = volume - dvolum el endif Ccalculated(i) = C(1) write(6,350)it,withdraw(i),volume,Ccalculated(i) goto310 endif 102 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. enddo 310 continue 350 form at(VO LUM E W ITHDRAW AL i,sec,vol,Ccalc>\i8,f12.0,f8.4,f8.4) enddo c now done c w rite(6,400)(C(k),k=1,51) c400 format(10f7.3, 10f7.3, 10f7.3, 10f7.3, 10f7.3) c now calculate chi-squared chisum = 0. do i=1,nwithdraw if(Cmeasured(i) .GE. 0.)then chisum = chisum + (Cmeasured(i)-Ccalculated(i))**2. endif enddo write(6,9010)chisum 9010 format('END OF DIFFU SIVITY LOOP> C H I-SQ UA RED = ',e15.6) finaldiffuse(idiff) = diffuse finalchisum(idiff) = chisum 8000 continue c now summarize write(6,9400) 9400 formatCFINALLY, W E SUM M ARIZE D IFFU SIVITY & CHI- SQUARED:') do i=1,ndiff write(6,9500)finaldiffuse(i),finalchisum(i) 9500 format(2f15.10) enddo c now print concentration diffusion as a function of time write(6,9600) 9600 form at('CONCENTRATION HISTO R Y O VER TIME:') do i=1,nt,120 write(6,9700)dt*(i-1),Cleftmost(i) 9700 format(f10.0, f15.10) enddo end 103 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. C = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = c subroutine to find reservoir concentration at time zero using c first draw, subroutine getC0(C,diffusivity,nx,dx,dt,porosity,Area,volume,Cwork,CO, + COIargest,nt_first) real C(nx),diffusivity(nx),Cwork(nx) real dx,dt,porosity .Area,volume,CO integer nx,nt_first C1 = 1.00 dC = C O Iarg est/100. do 1000 iloop_C = 1,100 CO = dC * float(iloop_C) c Initial conditions do ix=1,nx C(ix) = C1 enddo c now reset leftmost cell C(1) = CO c================================================== c (C) model ready to go. Loop over all requested time steps. c Loop over time do it=1,nt_first call diffuse1d(C,diffusivity,nx,dx,dt, + porosity .Area, volume,Cwork) enddo if(C(1) .GE. C0largest)then write(6,5000)diffusivity( 1), CO return endif 5000 format('C0 SEARCH for TH IS DIFFU SIVITY:',f15.8,f15.5) 1000 continue write(6,2000)diffusivity(1) 2000 format('FATAL> CO SEARCH FAILED FOR D IFFU SIVITY',e15.6) stop end 104 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix B Diffusion Experiments: Normalized Concentration (C/CQ ) Versus Time Plots show C /C 0 v. time for each solute of each experimental run. Plots are organized by experiment. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure B-1. Normalized Concentration (C /C 0) Verus Time. 0.250 0.300 - i----- ------i---- 1 -----1 -----1 -----.---- 1 -----1 -----i— i-----------1 ---- 1 -----i— i---- 1 ---- -— ; 0.200 • Observed Si 1 A A A Observed KCI 1 0.250 • Observed Si I a Observed KCI 1 o O 0.200 A 0.150 • • • ! e 0.150 A * 0.100 O 0.100 0.050 A A EXPERIMENT 1 0.050 EXPERIMENT 2 0.000 i A . • • Observed C/C0 v. Time • ■ ■ ■ . ■ ___i _ _ __ _ __________i _ _ >__j— i — .— 0.000 • Observed C/C0 v. Time 0 10® 5 10* 110* 2 10s 2 10s 010® 5 10 * 1 10 s 1.510s 2 10s Time (a) Time (a) 0.250 .................................................................................... 0.200 -----1 -----1 -----i— i-----1 -----1 -----i— i-----1 ---------------- ------1 — i— i— i— r— r — r— • Observed Si 1 • Observed Si 1 a Observed KCI 1 a Observed KCI 1 A 0.150 A 0.150 A U o • • 1 0.100 « • • 0.100 • o • 0.050 EXPERIMENT 3 0.050 EXPERIMENT 4 0.000 • Observed C/C0 v. Time ■ - ■ ........................ i 0.000 • Observed C/C0 v. Time 010® 5 10 * 110s 1.510s 2 10s 0 10® 5 10 * 1 10 s 1.510s 2 10s Time (a) Time (a) g 0.150 • Observed Si a Observed KCI EXPERIMENT 5 Observed C/C„ v. Time O 0.150 * | | 0 .1 0 0 o • Observed SI I a Observed KCI I EXPERIMENT 6 Observed C/CD v. Time 0 .0 0 0 i i ... I ■ 0 .0 0 0 0 10® 510* 1 10s 1.510s 210s 2.510s 3 10s 3.510s 010® 5 10 * 1 10s 1.510s 2 10 s 2.510s 310s 3.510s Time (a) Time (a) ° 0.150 Observed SI Observed Ge Observed KCI EXPERIMENT 7 Observed C/C0 v. Time 5 10* 1 10 s 1.510s 2 10s Time (s) 310s Observed Si Observed Ge Observed KCI 0 .2 0 0 O 0.150 & 0 .1 0 0 EXPERIMENTS 0.050 Observed C/C0 v. Time 0 1 0 ® 1 10s 1.510s 210s Time (s) 2.510s 310s 106 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure B-1. Normalized Concentration (C /C 0) Verus Time, (cont.) 0.350 0.300 0.250 o ° Q 0 .2 0 0 S S 0.150 w .o ° 0 .1 0 0 0.050 0 .0 0 0 0 • Observed Si A 0.350 O Observed Si ♦ Observed Ge ♦ Observed Ge a Observed KCI 0.300 a Observed KCi EXPERIMENT 9 Observed C/C0 v. Time 100 5 10* 1105 1.510s 2 10s 2.510s 310s 3.510s 4 10 s Time (s) 0.250 0 .2 0 0 0.150 0 .1 0 0 0.050 EXPERIMENT 10 V * Observed C/C„ v. Time 0 .0 0 0 0 100 5104 110s 1.510s 210s 2.5 10s 3 10s 3.510s 4 1 0 s Time {% ) O Observed SI ♦ Observed Ge a Observed KCI a o .i5 o o * 8 t 2 0 .1 0 0 EXPERIMENT 11 Observed C/CD v. Time 0100 5104 1 1 0 s 1.510s 210s 2.5 10s 3 10s 3.510s Time (s) y 0.060 s t S 0.040 J O O O Observed Si ♦ Observed Ge EXPERIMENT 18 Observed C/C0 v. Time 0100 5 1Q4 1 1 0 s 1.510s 2 10 s 2.510s 3 10s 3.510s Time {«) 0 .0 2 0 0 .0 0 0 O Observed Si ♦ Observed Ge EXPERIMENT 19 Observed C/Cc v. Time 0 100 5 1 0 4 1 1 0 s 1.510s 2 10s 2.510s 310s 3.510s Time (s) y 0.300 £ S 0.200 A o 0 .1 0 0 • 9: • • O Observed SI : ® ♦ Observed Ge ■ * EXPERIMENT 20 Observed C/C„ v. Time > ................. . _______________---------------------------------------------------------------- 0100 5 104 1 10s 1.510s 2 10s 2.510s 3 10s 3.510s Time (•) 0.400 0.350 0.300 0.250 0 .2 0 0 0.150 0 .1 0 0 0.050 0 .0 0 0 • 2 6 : 8 O Observed SI ♦ Observed Ge • '* EXPERIMENT 21 Observed C/C0 v. Time : O O 0.150 O Observed Si ♦ Observed Ge 010® 5 104 110s 1.510s 2 10s 2.5 10s 3 10s 3.510s Time (s) EXPERIMENT 22 Observed C/C„ v. Time ■ ■ i i . . . . . . . . 010® 5 104 1 10 s 1.510s 2 10 s 2.510s 3 10s 3.510s 4 1 0 s Time (s) 107 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure B-1. Normalized Concentration (C /C 0) Verus Time, (cont.) 0 .1 2 0 0 .1 0 0 O 0.080 0 1 j> 0.060 O Observed SI ♦ Observed Ge EXPERIMENT 23 Observed C/C0 v. Time ■ i i ............................................. 0 1 0 « 5 io « 1 1 0 * 1 .5 1 0 * 2 1 0 * 2.5 10* 3 10* 3 .5 1 0 * Time (s) t - l 1 I T -l-T -M 7 ' O Observed Si ’■ ' 1 ' ' ' » 1 * ' ' ' 1 ' ' 1 ' 5 8 : ♦ Observed Ge ♦ o ♦ o ♦ EXPERIMENT 24 o * 4 Observed 0/Co v. Time 110* 1.510* 2 10* 2.510* 310* Time (s) ■ ■ ■ ■ '■ i i ■ ■ ■ i ............... | i ■ i | ■ i i i i - i '- t i i M-r 1 >11 !■>' 1 > i ■ i 1 ■ i i ■ 1 ■ i i ■ 1 i i ■ I11 O Observed Si O o O Observed Si O 0.100 4 Observed Gc 4 * . 0.100 4 Observed Go ♦ ♦ - 0.080 O 4 0° 0.080 O ♦ 4 0.060 o ♦ £ 0.060 O A 4 0.040 ° 0.040 ♦ EXPERIMENT 25 0 EXPERIMENT 26 0.020 " 8 Observed C/0o v. Time 0.020 Observed C/C0 v. Time 0 1 0 * 5 104 1 105 1 .5 1 0 * 2 10* 2 .5 1 0 * 3 10* Time (s) 0 100 5 1 0 4 n o s 1.5 10* 2 10* 2.510* 3 10* 3.510* 410* Time (s) 0.120 i i 7 r r i > | i i i i ■ i >'■ ■ ■ i— r............... 1 , ~ ''1 "i 1 ■ ■ M |"r 'w 1 0.120 1 1 ' ' 1 1 1 T-1-1-| l- l- P I | 1 1 1 • | 1 1 1 1 O Observed Si 9 2 O Observed SI O 0.100 4 Observed Ge 0.100 4 Observed Ge 9 8 0.080 0 ° 0.080 0.060 4 0 1 0.060 9 9 4 1 M A 9 0.040 ° 0.040 V EXPERIMENT 27 EXPERIMENT 28 0.020 - Observed C/CQ v. Time 0.020 F * 8 Observed 0/Co v. Time 0.000 , 0.000 o . . . . . . . . . . . . . . . . . . . . 0100 5 1 0 4 1 10* 1.510* 2 10* 2.510* 3 10* 3.510* 410* Time (s) 0 1 0 ° 5 104 1 105 1.5 10* 2 1 0 * 2 .5 1 0 * 3 10* 3 .5 1 0 * 4 1 0 * Time (s) 0.160 0.140 0.120 0 .1 0 0 0.060 0.060 0.040 0.020 0 .0 0 0 q O Observed Si 4 Observed Ge EXPERIMENT 29 Observed C/C„ v. Time 0 10* 5 10 * 110* 1.5 10* 2 10* 2.510* 3 10* 3.510* 410* Time (s) 0.160 0.140 0 .1 2 0 0 .1 0 0 0.080 0.060 0.040 0 .0 2 0 0 .0 0 0 0 O Observed SI 4 Observed Ge a Observed KCI EXPERIMENT 30 Observed C/C0 v. Time 100 5 1Q4 n o * 1.510* 2 10* 2.510* 310* 3.5 10* Time (s) 108 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Observed C/C. Figure B-1. Normalized Concentration (C /C 0) Verus Time, (cont.) Observed Si Observed Ge Observed KCI EXPERIMENT 31 Observed C/C0 v. Time 010° 5 10« 110* 1.510* 210* 2.510* 3 10* 3.510* Time (s) 109 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix C Diffusion Experiments: Observed C/C0 Versus Numerical Simulation C/C Plots are organized by experiment and then by solute. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. 0.14 Experiment 1 Observed v. Modeled Si C/C, 0.12 o 0.10 o a 5 5 0.0# J | 0.06 M ° 0.04 0.02 0.00 0.12 0.06 0.08 0.10 0.14 0.02 0.04 0.00 0.25 Experiment 1 Observed v. Modeled KCI C/C, 0.20 o O O 0.15 O X •a o w A O 0.10 Error 0.05 0.00 0.25 0.05 0.10 0.15 0.20 0.00 Model Generated KCI CIC0 Model Generated SI C/C„ Experiment 2 Observed v. Modeled Si C/C, Experiment 2 Observed v. Modeled KCI C/C, y*M1»M2’ M0 value* “ 0,012692 0,041493 0.02 0.05 0.07 0.09 0.11 0.14 0.16 0.18 Model Generated SI C/Cc 0.10 0.15 Model Generated KCI C/C( 0.25 Experiment 3 Observed v. Modeled KCI C/C, 0.20 o o _ 0.15 0 X 1 | 0.10 < 0 A o ■ErrST 0.05 Chisq 0.00 0.15 0.20 0.25 0.00 0.05 0.10 0.20 Experiment 3 Observed v. Modeled Si C/C, 1.15 o a o < 0 e w 0 . 1 0 A o Error 0.05 0.00 0.20 0.15 0.00 0.05 0 . 1 0 Model Generated KCI C/CQ Model Generated Si C/C0 0.25 Experiment 4 Observed v. Modeled KCI C/C, 0 . 2 0 e O - 0.15 O X 'S o 0.10 V a l u e " V ) A o 0.05 0 . 0 0 0.25 0.10 0.15 0.20 0.00 0.05 0.16 0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00 Experiment 4 Observed v. Modeled Si C/C0 y«Ul+M2'M 0 Error m l •0.0066466 0.005603 m2 1.062 0.059281 Chisq 0.00032164 NA R* 0.96231 NA Model Generated KCI C/CQ 0.04 0.06 0.08 0.10 0.12 Model Generated Si C/C0 111 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. (cont.) Experiment S Observed v. Modeled KCI C/C0n y m M1+M2 M0 0.0035876 1.0355 0.10 0.15 Model Generated KCI C/C0 0.20 Experiment 5 Observed v. Modeled Si C/C, 0.15 O O V ) o £ e V ) A o 0.05 0.00 0.05 0.10 0.15 0.20 0.00 Model Generated Si C/C0 Experiment 6 Observed v. Modeled KCI C/C0 o 0 1 5 * 0 1 0.10 A o 0.00 0.0 y = Ml+M2’ M0 brror ml -0.017858 0.011451 m2 1.1166 0.08459 Chisq 0.00074635 NA R2 0.97756 NA 0.08 0 .1 2 0.16 Model Generated KCI C/C0 0.15 Experiment 6 Observed v. Modeled Si C/C, < 0 * I © A o 0.05 0.00 0.08 0.10 0.12 0.14 0.04 0.06 0.00 0.02 Model Generated Si C/CQ Experiment 7 Observed v. Modeled KCI C/C0 r E T u H r M O 0.021532 0.99774 0.10 0.15 Model Generated KCI C/C0 0.20 Experiment 7 Observed v. Modeled Si C/C, 0.15 < J o ( 0 1 t © 0.10 “ ETcT A o 0.05 0.00 0.15 0.20 0.05 0.10 0.00 Model Generated Si CICQ Experiment 8 Observed v. Modeled KCI C/C, Experiment 7 Observed v. Modeled Ge C/C, M1+M2-M0 0.0015344 0.020626 0.99921 0.99827 0 .1 0 0.15 Model Generated KCI C/C( 0.10 Model Generated Ge C/C< 1 1 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Observed KCI C /C „ Observed S i CIC Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. (cont.) 0.20 Experiment 8 Observed v. Modeled Si C/C, 0.15 0.10 Error 0.05 0.00 0.15 0.20 0.05 0.10 0.00 Model Generated Si C/C0 O 0.12 Experiment 8 Observed v. Modeled Ge C/C0 y * M1+M2*M0 0.04 0.06 0.08 0.10 0.12 Model Generated Ge C/C0 Experiment 9 Observed v. Modeled SI C/C, Experiment 9 Observed v. Modeled KCI C/C, -0.0074973 0.021519 Chieq 0.00015204 0.99698 0.10 0.15 Model Generated Si C/C, 0.12 0.16 0.20 0.24 0.28 0.32 Model Generated KCI C/C0 Experiment 9 Observed v. Modeled Ge C/C, Experiment 10 Observed v. Modeled KCI C/C, •0.0059948 0.005031 •0.0050994 0.026737 0.10 0.15 Model Generated Ge C/C, Experiment 10 Observed v. Modeled Ge C/C, Experiment 10 Observed v. Modeled Si C/C, •0.0012126 0.0025437 0.0058907 0,034163 0.99217 0.10 0.15 0.20 Model Generated Ge C/Ce 0.10 0.15 0.20 Model Generated Si C/C0 113 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C/C0Versus Numerical Simulation C /C 0. (cont.) Experiment 11 Observed v. Modeled KCI C/C, Experiment 11 Observed v. Modeled Si C/C, T n + J H r B T Value I yM i+ U rM o •0,0012549 0.0037151 •0.0107S4 0.95793 0.10 0.15 Model Generated KCI C/C < 0.15 Experiment 11 Observed v. Modeled Ge C/C, o o © o t 6 m A O 0 . 1 0 0.05 Chiaq 0 . 0 0 0.15 0 . 1 0 0.05 0 . 0 0 Experiment 18 Observed v. Modeled Si C/CQ y .U u M ^ M o --------------- Value Error ml 0.0014526 0.0025473 m2 0.95111 0.043052 Chiaq 9.7881e-0S NA R2 0.96556 NA Model Generated Ge C/C0 0.04 0.06 Model Generated SI C/C0 0 . 1 0 Experiment 18 Observed v. Modeled Ge C/C, 0.08 o o © o 0.06 "O © £ © 0.04 Error 0.02 0 . 0 0 0.08 0 . 1 0 0.04 0 . 0 6 0 . 0 0 0 . 0 2 3 0.06 Experiment 19 Observed v. Modeled Si C/C .U u lM lo Model Generated Ge C/Ce 0.04 0.06 0.08 Model Generated SI C/C0 0.50 Experiment 20 Observed v. Modeled Si C/C, 0.40 o 0.30 0 . 2 0 0 . 1 0 Chiaq 0 . 0 0 0.50 0 . 2 0 0.30 0.40 0 . 0 0 0 . 1 0 0 . 1 0 Experiment 19 Observed v. Modeled Ge C/C, o 0 < -> 0.06 © 0 1 S 0.04 M A o 0 . 0 2 "Error" Chisq 0 . 0 0 0.08 0.10 0.02 0.04 0 . 0 6 0.00 Model Generated Ge CrC„ Model Generated SI C/C„ 114 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. (cont.) 0.40 Experiment 21 Observed v. Modeled Si C/C, 0.35 0.30 0.25 ( 0 1 o © £ o 0 . 2 0 in o ° -1 5 0 . 1 0 0.05 0 . 0 0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.50 Experiment 20 Observed v. Modeled Ge C/C, 0.40 O o © 0 ! Q > 0.30 0 . 2 0 V ) n O 0 . 1 0 CNsq 0 . 0 0 0.50 0.30 0.40 0 . 0 0 0 . 1 0 0 . 2 0 Model Generated Ge C/C0 Model Generated SI C/C0 0.25 Experiment 22 Observed v. Modeled Si C/C, 0 . 2 0 (J o 0.15 ( 0 ! e m 0 . 1 0 £ o TrroT 0.05 0 . 0 0 0.25 0 . 1 0 0.15 0 . 2 0 0 . 0 0 0.05 0.40 Experiment 21 Observed v. Modeled Ge C/C, 0.35 0.30 p p o © o T J e £ © 0.25 0 . 2 0 w 5 0.15 O Error 0 . 1 0 0.05 0 . 0 0 0.40 0.30 0 . 1 0 0 . 2 0 0 . 0 0 Model Generated Ge C/C0 Model Generated SI C/C0 Experiment 22 Observed v. Modeled Ge C/C, Experiment 23 Observed v. Modeled Si C/C, •0.0044746 -0.013614 0.053221 0.056944 Chisq 0.00052619 0 . 1 0 Model Generated Ge C/C( 0.04 0.06 0.0* Model Generated Si C/C( 0.15 Experiment 24 Observed v. Modeled SI C/C, O P U 0 . 1 0 < 0 ! « .© o 0.05 0 . 0 0 0.15 0 . 1 0 0 . 0 0 0.05 0 . 1 0 Experiment 23 Observed v. Modeled Ge C/C, 0.08 o y p 0.06 © o •o © 0.04 & o 0 . 0 2 0 . 0 0 0 . 1 0 0.06 0.08 0 . 0 2 0.04 0 . 0 0 Model Generated Ge C/CQ Model Generated Si C/Cc 115 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C /C 0Versus Numerical Simulation C /C D. (cont.) Experiment 25 Observed v. Modeled Si C/C, Experiment 24 Observed v. Modeled Ge C/C, 0.0033754 0.037641 0.99376 0.04 0.06 0.08 Model Generated Si C/C0 0.02 0.04 0.06 0.08 0.10 Model Generated Ge C/C0 Experiment 25 Observed v. Modeled Ge C/C « 0.08 S 0.06 2 0.04 y = M1+^2'M0 Value Error ml •0.0027562 0.0013082 m2 1.0349 0.01929 Chisq 1.7096-05 NA R2 0.99627 NA 0.04 0.06 0.08 Model Generated Ge C/CQ 0 .1 2 Experiment 26 Observed v. Modeled Si C/C, 0 .1 0 o ° o 0.08 ( 0 ” 2 0.06 0 1 0) TrroT 0 . 0 2 0 . 0 0 0 . 1 2 0.04 0.08 0 . 1 0 0 . 0 0 0 . 0 2 0.06 Model Generated Si C/C0 Experiment 26 Observed v. Modeled Ge C/C y = Ml+M2'M0 0,0011151 0.0093804 Experiment 27 Observed v. Modeled SI C/C y .U l+ U 2‘ Uo--------------- Value Error m l 0.0031623 0.0031826 m2 0.96463 0.041009 Chlaq 7.91246-05 NA R3 0.99104 NA 0.04 0.06 Model Generated Ge C/C0 0.04 0.06 0.08 Model Generated Si C/C0 Experiment 28 Observed v. Modeled Si C/C, Experiment 27 Observed v. Modeled Ge C/C, TR TffiW T T T S r M c T •0,0037126 0.00061144 0,013336 0.96643 0.033306 NA 5.13e-05 0.04 0.06 0.01 Model Generated Si C/C« 116 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. (cont.) Experiment 29 Observed v. Modeled Si C/C, Experiment 28 Observed v. Modeled Ge C/C, 0,0026142 •0,0019382 0.00093768 0.015671 0.08 0 .1 2 Model Generated Si C/CQ 0.04 0.06 0.08 Model Generated Ge C/C, Experiment 30 Observed v. Modeled KCI C/C, Experiment 29 Observed v. Modeled Ge C/C, 0.0012515 •0-0021191 0.017244 l.819e-0S 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 Model Generated KCI C/C0 4 0.06 0.08 0.10 0.12 0.14 Model Generated Ge C/C„ Experiment 30 Observed v. Modeled Si C/C, Experiment 30 Observed v. Modeled Ge C/C, M h U z 'U q •0,0018467 0.043649 0.04 0.06 0.08 Model Generated Si C/C0 0.04 0.06 Model Generated Ge C/C, Experiment 31 Observed v. Modeled KCI C/C0 y * M1+M2‘M0 brror m l •0.0040408 0.0021732 m2 1.0336 0.020521 Chisq 4.21696-05 NA R2 0.99803 NA 0.05 0.10 0.15 Model Generated KCI C/C0 0.14 Experiment 31 Observed v. Modeled Si C/C, 0 . 1 2 o 0 .1 0 O u V ) ■o a > 2 0) 0.04 ChUq 0 . 0 2 0 . 0 0 0 . 1 0 0 . 1 2 0.06 0.08 0 . 0 0 0 . 0 2 0.04 Model Generated Si C/C0 117 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Observed S i C/C Figure C-1. Observed C /C 0Versus Numerical Simulation C /C 0. (cont.) 0.14 Experiment 31 Observed v. Modeled Si C/C, 0 .1 2 - 0 . 1 0 0.08 0.06 0.04 Chlsq 0 . 0 2 0 . 0 0 0.06 0 . 1 0 0 . 1 2 0.04 0.08 0 . 0 0 0 . 0 2 Model Generated Si C/CQ 1 1 8 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix D Diffusion Experiments: Relative Diffusivities of Solutes Plots are organized by experiment and then by ratios of solutes (Si/KCI, Ge/KCI and Ge/Si). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure D-1. Relative diffusivity of Solutes. 0.14 EXPERIMENT 1 Si v. KCI C/C0 0.12 0.10 .08 <0 0.06 Error 0.04 0.02 0.00 0.15 0.20 0.25 0.10 0.00 0.05 0.20 EXPERIMENT 2 Si v. KCI C/C0 0.15 O 0.10 0.05 0.00 0.30 0.15 0.20 0.25 0.05 0.10 0.00 0.20 EXPERIMENT 3 Si v. KCI C/C0 0.15 5 o .io 0.05 Chlsq 0.00 0.20 0.25 0.10 0.15 0.00 0.05 0.16 0.14 0.12 0.10 0 0.08 0.06 0.04 0.02 EXPERIMENT 4 Si v. KCI C/Cc 0.10 0.15 KCI C/C0 0.20 EXPERIMENT 5 Si v. KCI C/C0 0.15 <3 0 .1 0 0.05 Chlsq 0.00 0.15 0.20 0.25 0.10 0.00 0.05 0.15 EXPERIMENT 6 Si v. KCI C/C0 0.10 o y o » 0.05 0.00 0.25 0.15 0.20 0.05 0.10 0.00 EXPERIMENT 7 Ge v. KCI C/C 2.0fl96e-05 0.10 0.15 KCI C/Cc 0.20 EXPERIMENT 7 Si v. KCI C/C0 0.15 o O u 0.10 < 0 Error 0.05 Chlsq 0.00 0.15 0.20 0.25 0.10 0.05 0.00 120 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure D-1. Relative diffusivity of Solutes, (cont.) EXPER MENT 7 Ge v. Si C/C SiC/C EXPERIMENT 8 Ge v. KCI C/Cfl 0.00 0.05 yaM1+M2’ M0 vaue terror m l 0.0011114 0.0026300 m2 0.64274 0.016644 Chlsq 3.3829e-09 NA Ri 0.90726 NA 0.10 0.15 KCI C/C0 EXPERIMENT 8 Si v. KCI C/C 5 0 .1 0 M1+M2‘M 0 0.003307 0.021172 0.10 0.15 KCI C/CQ EXPERIMENT 8 Ge v. Si C/C U 0.10 ----------------y a Ml+M2 M0--------------- Value terror ml •0.0020709 0.00093234 m2 0.60029 0.0043506 Chlsq 1.6676e4>6 NA R* 0.00968 NA Experiment 9 Ge v. KCI C/C •0.0074771 KCI C/C 0.25 Experiment 9 Si v. KCI C/C0 0 . 2 0 0.15 O o t o 0 . 1 0 0.05 Chlsq 0.00 0.35 0.25 0.30 0.05 0 . 1 0 0.15 0 . 2 0 0.00 Experiment 9 Ge v. Si C/C yTMl+MrMO 0.0030072 Experiment 10 Ge v. KCI C/C0 O 0.15 y.UuM a'Uo 0.10 0.15 Si C/C0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 KCI C/C0 121 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure D-1. Relative diffusivity of Solutes, (cont.) 0.30 Experiment 10 Si v. KCI C/C0 0.25 0.20 O 0.15 0.10 0.05 Chieq 0 . 0 0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Experiment 10 Ge v. Si C/C U 0.15 U i + U r U o 0.10 0.15 0.20 Si C/C0 0.16 EXPERIMENT 11 Si v. KCI C/C0 0.14 0 . 1 2 0 . 1 0 O o 0.08 5 5 0.06 Error 0.04 0 . 0 2 0 . 0 0 0.25 0.15 0 . 2 0 0.05 0 . 1 0 0 . 0 0 0.15 EXPERIMENT 11 Ge v. KCI C/C0 0 . 1 0 o y o © © Error 0.05 0 . 0 0 0.25 0 . 2 0 0 . 1 0 0.15 0 . 0 0 0.05 KCI C/C„ KCI C/C0 EXPERIMENT 18 Ge v. Si C/Co EXPERIMENT 11 Ge v. Si C/C0 k l i . l f e 'k l o <■0067136 0.001423 1.0243 0.014564 0.013535 EXPERIMENT 20 Ge v. Si C/Co EXPERIMENT 19 Ge v. Si C/Co T X u t X T U o Value I 0.0012535 0.00017405 0.96577 1 2 2 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure D-1. Relative diffusivity of Solutes, (cont.) EXPERIMENT 21 Ge v. Si C/Co o 0.20 ^ ^ hT+m? mo S i C /C EXPERIMENT 22 Ge v. Si C/Co y»M l+ M 2*Uo Vakis fcnor ml •0.014254 0.001275 m2 0.98178 0.0099453 Chlsq 1.9664s-OS NA R* 0.99949 NA 0.10 0.15 si c/cn 0 . 1 0 EXPERIMENT 23 Ge v. Si C/Co 0.08 o 0.06 o © 0.04 0 .0 2 Chlsq 0 . 0 0 0.08 0 . 1 0 0 . 1 2 0 . 0 0 0 . 0 2 0.04 0.06 0.14 EXPERIMENT 24 Ge v. Si C/Co 0 . 1 2 0 . 1 0 9 y o o © 0.08 0.06 E fT O t 0.04 0 . 0 2 0 . 0 0 0 . 0 0 0.05 0 . 1 0 0.15 0 . 1 2 EXPERIMENT 25 Ge v. Si C/Co 0 . 1 0 0.08 e O O « © 0.04 0 . 0 2 Chlsq 0 . 0 0 0.04 0.06 0.08 0 . 1 0 0 . 1 2 0 . 0 0 0 . 0 2 EXPERIMENT 26 Ge v. Si C/Co U 0.06 y ■ Ml+Uz*Uo Error m l •0.011692 0.0037366 m2 0.97826 0.046524 CNsq 8.16S4S-05 NA R2 0.98682 NA 0.06 0.08 S i C /C a EXPERIMENT 27 Ge v. Si C/Co 9 0.06 yTM lTw TO Talus 0.98697 0.99799 0.04 0.06 S i C /C Q 0 . 1 2 EXPERIMENT 28 Ge v. Si C/Co 0 . 1 0 0.08 o O o o © 0.06 0.04 0 . 0 2 0 . 0 0 0 . 0 0 0 . 0 2 0.04 0.06 0.08 0 . 1 0 0 . 1 2 123 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure D-1. Relative diffusivity of Solutes, (cont.) 0 . 1 0 EXPERIMENT 30 Ge v. KCI C/Co 0.08 O 0.06 O e o 0.04 0 . 0 2 Chtsq 0 . 0 0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.14 EXPERIMENT 29 Ge v. Si C/Co 0 . 1 2 0 . 1 0 O o « O 0.06 'Error 0.04 0 . 0 2 0 . 0 0 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 S i C /C Q K C I C /C 0 EXPERIMENT 30 Ge v. Si C/Co EXPERIMENT 30 Si v. KCI C/Co T n + J 3 ? i 3 r -0,0045362 0.002856 EXPERIMENT 31 Si v. KCI C/Co EXPERIMENT 31 Ge v. KCI C/Co 0,0033241 0,0094592 0.71054 0 . 1 0 EXPERIMENT 31 Ge v. Si C/Co 0.08 o o 0.06 0.04 0 . 0 2 Chlsq 0 . 0 0 0.08 0 . 1 0 0 . 1 2 0 . 0 2 0.04 0.06 0 . 0 0 124 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix E Determination of Boundary Layer for Core Incubations To determine the optimal speed for the motors that stir the cores during incubation, the dissolution of alabaster is used. Alabaster is made of C aS 0 4-2 H2 0 (gypsum) and as the water in the core is stirred, the alabaster dissolves. The mass of alabaster that dissolves is controlled by the stir rate, temperature, solubility of gypsum, water column height, and length of time stirred. A boundary layer (8) is a thin layer assumed to be present between the sea floor sediment and the overlying water column. This layer is not well mixed like the water column above, causing it to develop a concentration gradient due to the flux of material out of the sediment pore waters. If it is assumed that the interface is at equilibrium and that mass transport through the stagnant boundary layer is diffusive transport only, then the solubility, mass of alabaster dissolved, average diffusivity of Ca+ 2 and SCV2, and the time it took to dissolve can be used to calculate a boundary layer thickness for the core environment as follows: S = (A t)Dav(Cea- C W )A„ Am 125 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. where: 8 = boundary layer in pm At = time in seconds Dav = average diffusivity of Ca+2 and SO4'2 in cm2/s Ce q = equilibrium concentration of gypsum (g/cm3) Cw = concentration of gypsum in the water column (g/cm3) Ap = the area of the alabaster plate, 7.3 cm2 Am = mass loss in g Note: C... = Am A h where: At = cross-sectional area of the core (cm2) h = water column height (cm) Because the core incubation simulates the sea floor environment, it would be ideal to create a boundary layer thickness in the core that is similar to that observed in the ocean. A desirable boundary layer thickness for core incubations for these stations is 100-200 microns, about 1/3 the thickness of the boundary layer in the ocean at M ANOP site H (3750 m). Two experiments were conducted. The first was a variable column height and a fixed stir rate. The second was a fixed column height and a variable stir rate. Experiments were run in a water bath at 25°C. The diffusivity mean for C a+2 and S 0 4" 2 was taken as 9.27 x 10"5, activity 126 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. coefficients were assumed to be unity, and for a molecular weight of 172 g/mol, Ce q is 28.9 pg/l. The procedure used is as follows: 1. The alabaster plate was soaked in D IW for approximately one hour then air dried for one hour and weighed on a top loading scale. 2. The plate was then taped to the base of the core tube, the “plug”, and the piston was adjusted to the appropriate height in the core tube to yield the specified water column height. It was then filled with DIW, and the plug was inserted and the motor started. 3. The plate was kept in the core tube and stirred for various time lengths as indicated below. It was then removed, air dried for one hour, and then weighed. 4. The weight was recorded and subtracted from the initial weight recorded in step one. This is the mass loss (Am) or dissolved due to the circulating water. 127 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 5. This procedure is repeated for each of the different experiment criteria listed below. 6. The mass loss is recorded and the time duration in seconds is recorded. The area of the alabaster plate is 7.3 cm2. The height of the water column is variable depending on experiment; equilibrium concentration depends on Ke q and the concentration of gypsum in the water column depends on the mixing in the water column. Variable Water Column Height The variable water column height experiment used a 115 rpm motor with an actual stir rate of 76 rpm, dissolution was measured for water column heights of 15, 10, and 6 cm. Results are as follows: Table E-1. Results of variable water column height experiment._____________ M ass Tim e W ater colum n ht S tir rate Boundary Layer loss (g) (seconds)_________________ (cm )______(rpm )________________ (pm ) 0.09 8400 15 76 48.6 0.11 10200 10 76 43.8 0.15 7800 6 76 18.0 Variable Stir Rate The second experiment was conducted using variable speed stirring and a fixed water column height of 10cm. A 25 rpm motor was used and the 128 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. voltage varied to obtain the stir rates indicated below. The results are as follows: Table E-2. Results of variable stir rate experiment. Mass loss (g) Time (seconds) Water column ht (cm) Stir rate (rpm) Boundary Layer (urn) 0.0042 45000 10 19 6150 0.1106 86100 10 21 367 0.2275 86400 10 29 136 0.1507 79200 10 24 266 The ideal water column height (10 cm) was determined by varying the water column height and plotting it against the resulting boundary layer thickness (Figure E-1). The stir rate was determined by plotting variable stir rates at a fixed water column height of 10 cm against the boundary layer (Figure E-2). At a water column height of 10 cm, and a stir rate of 32 rpm, it was determined that a boundary layer of about 120 |im would be produced. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 60 E 50 k . 0 ) % 40 ■ g 30 C 3 O 00 20 10 — i — 1 — 1 — 1 — 1 — < — 1 — . Variable Water Column Height (Stir rate 76 rpm) 11111111 8 10 12 14 W ater Colum n Height (cm) 16 Figure E-1. Boundary Layer v. W ater Column Height. This plot shows the change in boundary layer thickness with increasing water column height. 500 E 400 3. 0 ro 300 _i c? ■O 200 c 3 o 03 100 Variable Stir Rate (Water column height 10 cm) 11 1 ■ ■ 11 1 1 * ■ ■ ■ 1 ■ » 1 20 30 40 50 60 70 Stir Rate (rpm) 80 Figure E-2. Boundary Layer v. Stir Rate. This plot shows the change in boundary layer thickness as the stir rate varies. 130 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Appendix F Incubation Nutrient Flux Plots by Station. Plots are organized by station number, nutrient (NO3, PO4, Si, and Ge/Si), then core (A or B). Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Si ((iM) P 0 4 ftiM) N O (fiM) Figure F-1. Station 1, San Pedro (897 m). STATION 1 San Pedro Basin Core A 32 "ErroT 28 Chi»q y a m1‘8Xp(-m2'mO) 24 Error Chtoq 2 0 1 0 2 0 25 30 0 15 5 32 28 24 2 0 y a m t '«xp(-m2'm0) value Error ml 32.666 0.784 m2 0.016776 0.0021554 Chlaq 2.057 NA R 0.98528 NA STATION 1 San Pedro Basin Core B yaM l+M 4‘ Mo Error ml 32.392 0.95639 m2 -0.44235 0.070174 Chlaq * 3.3731 NA R 0.97675 NA days/m 10 15 days/m 2 0 25 STATION 1 San Pedro Basin C oreB STATION 1 San Pedro Basin Core A w i.w rM c 0,21093 0.085246 0.012936 Chiaq 0.033133 0 5 10 15 20 25 30 35 40 Days/m 10 15 20 25 30 Days/m ” ~ value Error 117.63 0.99804 0.97065 STATION 1 San Pedro Basin Core B STATION 1 San Pedro Basin Core A 10 15 20 25 days/m days/m STATION 1 San Pedro Basin Core A “ 7alu#" Error 45.677 132 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure F-2. Station 2, Santa Monica (904 m). 3 32 STATION 2 Santa Monica Basin Core A 28 Value 24 Chiaq 2 0 16 30 35 1 0 2 0 25 0 5 15 y • m1'e*p(-m2'm0) ml 28.695 1.2766 m2 0.021122 0.0031802 Chiaq 7.8812 NA R 0.97212 NA STATION 2 Santa Monica Basin Core B yTMl+TBTO 27.862 days/m 15 20 25 days/m STATION 2 Santa Monica Basin CoreB STATION 2 Santa Monica Basin Core A MI+M2-M0 0.042017 0.0022671 0 5 10 15 20 25 30 35 40 Days/m 5 10 15 20 25 30 35 Days/m 145 E 135 A 125 115 *m2’ M0+m3 value MO’ M O m l 119.32 0.4S m2 1.42 0.07 m3 •0.0165 0.0022 Chiaq 0.4529 NA R 0.99957 NA y b M1+M2*M0 brror ml 121.39 1.50 m2 0.88 0.08 Chiaq 12.70 NA R 0.98786 NA STATION 2 Santa Monica Basin Core A 10 15 20 days/m 25 30 35 155 145 135 125 115 y » ml+fitf'MO+mS'MO'MO Value fc ffO f ml 119.39 0.90 m2 1.66 0.13 m3 •0.0215 0.0036 Chiaq 1.9916 NA R 0.99856 y b M1+M2‘ M0 Value brror ml 122.76 2.46 m2 0.91 0.13 Chiaq 37.716 NA R 0.97245 NA STATION 2 Santa Monica Basin Core B 1 0 15 20 25 days/m 30 35 40 STATION 2 Santa Monica Basin Core A 0.10053 133 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure F-3. Station 3, Catalina Basin (1328 m). y b mlexp(-m2 m0) STATION 3 Catalina Basin Core A 1.7953 y s M1+M2'M0 10 15 days/m STATION 3 Catalina Basin Core B rxtT+nrMo 40.34 0.32735 •0.23142 0.022324 0.56455 0.93597 y b mr«xp(>m2*m0} 0.0061851 10 15 days/m STATION 3 Catalina Basin Core A 3.4 3.2 S ^3.0 O o . 2 .8 2 .6 25 15 2 0 1 0 0 5 Days/m STATION 3 Catalina Basin C oreB 2 .8 2 .6 y a M1+M2*M0 trior ml 3.2504 0.03063 m2 •0.0060466 0.002117 Chiaq 0.0052056 NA R 0.65507 NA 10 15 Days/m 2 0 25 150 145 140 135 130 125 y » ml+in2*MO+m3*MO*MO Value ferror ml 127.24 0.34 m2 0.60 0.07 m3 •0.0020 0.0029 Chiaq 0.26691 NA R 0.99936 NA y s M1+M2’ W o trror ml 127.39 0.25 m2 0.76 0.02 Chiaq 0.33273 NA R 0.99921 NA STATION 3 Catalina Basin Core A 1 0 days/m 15 2 0 25 145 140 130 125 y a m 1 +m2,M0+m3'M0‘M0 Value trror ml 129.35 0.42 m2 0.76 0.09 m3 •0.0055 0.0034 Chiaq 0.4223 NA R 0.99659 NA y a M1+M2*M0 Value fcrror m l 129.76 0.42 m2 0.62 0.03 Chiaq 0.9656 NA R 0.99676 NA STATION 3 Catalina Basin C o reB 1 0 days/m 15 2 0 25 STATION 3 Catalina Basin Core A 0.24187 134 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S i (fiM ) P04 (nM ) NOa (fiM) Figure F-4. Station 4, Tanner Basin (1539 m). STATION 4 Tanner Basin Core A 42 40 Error 38 36 25 30 1 0 15 2 0 5 O z 36 y s m1‘exp(-m2*m0) Value trror m l 41.082 0.41315 m2 0.0055286 0.00080457 Chiaq 1.1461 NA n 0.97019 NA STATION 4 Tanner Basin Core B y ■ M1+M2'M0 Value t n of ml 41.048 0.42085 m2 -0.21285 0.031712 Chleq 1.2209 NA n 0.98822 NA 25 days/m STATION 4 Tanner Basin Core A y » M1+M2'M0 D .65731 10 15 20 □ays/m 3.8 3.6 STATION 4 Tanner Basin Core B y « Ml+M2*Mo trror m l 3.1252 0.027945 m2 0.0081258 0.0021067 Chleq 0.0053855 NA R 0.91225 NA 10 15 Days/m m W m 2'M&*mrM0'M0 0.0061 STATION 4 Tanner Basin Core B STATION 4 Tanner Basin Core A 1 0 days/m 10 15 days/m STATION 4 Tanner Basin Core B 28.208 135 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. G e (p M ) S i(n M ) P 0 4 (|iM ) N 0 3 (p M ) Figure F-5. Station 6, Morro Bay (100 m). S T A T IO N 6 Morro Bay 6 Core B y.M 1+M 2‘M0 S T A T IO N 6 Morro Bay 6 Core A 0 . 2 9 0 7 4 • 0 , 5 4 5 1 1 0 , 0 2 3 2 1 0 . 6 0 4 6 5 | Vatu* \ 0 , 1 7 3 4 4 0 . 3 4 7 6 7 0 . 0 0 1 2 4 4 9 0 . 5 1 7 5 1 d a y s /m d a y s /m S T A T IO N 6 Morro Bay 6 C oreB S T A T I O N 6 Morro Bay 6 Core A 0 . 0 7 7 9 8 2 0 . 0 1 2 3 1 2 0 . 1 7 0 1 2 Days/m D a y s /m 140 MO'MO y * ml+m2 Mo+m3 0 . 9 9 9 4 9 y a M 1 +M2 M0 S T A T I O N 6 Morro Bay 6 Core A d a y s /m Error” 1 2 0 Chteq 80 S T A T IO N 6 Morro Bay 6 Core B 40 10 12 14 16 4 6 2 8 d a y s /m 65 S T A T IO N 6 Morro Bay 6 Core A 60 55 50 45 40 35 Chiaq 30 25 140 1 0 0 1 2 0 40 60 80 S i (|iM ) 136 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure F-6. Station 7, Morro Bay (198 m). 28 £ 20 i 5.0 4.5 4.0 E 3 3.5 O “■ 3.0 2.5 2.0 130 110 2 ■3 9 0 55 70 50 60 55 50 2 3 4 5 t o O 40 35 30 50 60 70 80 90 100 110 120 130 Si (nM) 137 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. S T A T IO N 7 Morro Bay 7 Core A y*m l'exp(-m 2‘m0) Value trror ml 29.637 0.2312 m2 0.022673 0.00079603 Chlsq 0.20916 NA R 0.9962 NA y * M 1 +M2'M0 Value brror ml 29,596 0.18696 m2 •0.56387 0.016634 Chisq 0.15508 NA R 0.99667 NA 2 =L o z 24 20 S T A T IO N 7 Morro Bay 7 Core B y = mrexpt*m2"m0) Value brror ml 30.121 0.1009 m2 0.036662 0.00038674 Chisq 0.038156 NA R 0.99984 NA y-U l+ Jntt’Uo value brror ml 29571 0.37446 m2 •0.62257 0.034756 Chlsq 0.82692 • NA R 0.99733 NA 20 15 S T A T I O N 7 Morro Bay 7 Core A brror Chlsq "NT 20 10 15 5 5.0 S T A T IO N 7 Morro Bay 7 Core B 4.5 4.0 3.5 o Q. 3.0 0.062492 0.017514 2.5 2.0 15 20 10 0 5 Days/m Days/m brror S T A T I O N 7 Morro Bay 7 Core A 15 20 10 0 5 ^^n U m 2 iM O *?n3? M O j M O 120 2 3 . C O S T A T IO N 7 Morro B ay 7 C oreB 15 5 10 20 0 days/m days/m STATION 7 . Off Coast Morro Bay (200m) Core A ysU l+M 2'M 0 Value brror ml 22.053 3.3994 m2 0.27767 0.035026 Chisq 13.158 NA R2 0.95444 NA Figure F-7. Station 8, Oxygen Minimum Zone (704 m). STATION 8 Oxygen Minimum Zone Core B STATION 8 Oxygen Minimum Zone Core A raisw -0.55638 0.065533 -0.49937 0.043366 0.017767 0.0035861 10 15 20 25 30 days/m 10 20 30 40 50 60 days/m 5.0 STATION 8 Oxygen Minimum Zone Core A 4.5 4.0 S A 3.5 O a . 3.0 2.5 2.0 40 60 30 50 0 10 20 Days/m 4.5 4.0 2 3.5 O a . 3.0 2.5 2.0 STATION 8 Oxygen Minimum Zone Core A V = Wh M2‘ W o value brror ml 3.1663 0.14787 m2 •0 004432 0.0086608 Chlsa 0.12113 NA R 0.28334 NA 10 15 Days/m 20 25 30 260 220 Chisq s ^ 180 < 7 5 ^hisq 140 STATION 8 Oxygen Minimum Zone Core A 100 60 30 40 50 20 0 10 180 ^ r^ r!3 'M 0 + r!U MO'MOI value' trror 160 brror s r 1 4 0 u > 1 2 0 STATION 8 Oxygen Minimum Zone . Core B 100 30 15 20 25 0 10 5 days/m days/m 110 105 1 0 0 2 9 5 — 90 o 80 75 70 100 110 120 130 140 150 160 170 180 Si (tiM) STATION 8 Oxygen Minimum Zone Core B 138 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Ge(pM) SiftiM) P04(pM ) NO^pM) Figure F-8. Station 9, Central Calif. Margin (1500 m). 46 44 42 40 38 36 STATION 9 Central Calif. Margin (1500m) Core A y = M1+M2*M0 value brror ml 41.633 0.54176 m2 •0.035562 0.039703 Chiaq 1.5206 NA R 0.45956 NA 10 15 days/m 20 25 46 44 42 > ■a 40 O z 38 36 34 STATION 9 Central Calif. Margin (1500m) Core B ------------ y .U .k r M o ------------ Value brror ml 41224 1.1535 m2 •0.10978 0.090666 Chiaq 6.9156 NA R 0.57287 NA 10 15 days/m 20 25 5.0 STATION 9 Central Calif. Margin (1500m) Core A 4.5 4.0 3.5 3.0 0.0052415 2.5 Chiaq 2.0 25 20 0 5 10 15 b.O j- 4.5 ; 4.0 : 3.5 - o £L 3.0 * 2.5 - 2.0 - STATION 9 Central Calif. Margin (1500m) Core B y . U i + U r M o brror ml 3.0441 0.066075 m2 6.0053371 0.0051945 0.02269f NA " f f 0.51019 NA Days/m 10 15 Days/m 20 25 2 1 0 190 170 150 yTf?iT+m?M0Tm?M0,M 0 y = M1+M2'M0 m l Value 157.24 brror 1.29 m2 2.04 0.09 Chiaq 8.6772 NA R 0.99676 NA STATION 9 Central Calif. Margin (1500m) Core A 15 20 25 2 1 0 190 170 150 y a m 1-fm 2‘^ A 0 -H n 3 *M 0 *M 0 ml 156.97 0.65 m2 2 .6 1 0.15 m3 •0.0378 0.0065 Chlaa 0.9144 NA R 0.99959 NA y m M1+M2'M0 STATION 9 Central Calif. Margin (1500m) C o reB 10 days/m 15 20 25 140 135 130 125 1 2 0 115 110 105 150 160 170 180 190 200 210 Si (nM) STATION 9 California Margin Slope (1500m) Core A M1+M2'M O ‘ J .U 0,50367 0.070646 0.94397 139 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Figure F-9. Station 10, Central Calif. Margin (3444 m). 39 2 A STATION 10 Central Calif. Margin (3444m) Core A 38 37 36 y»M1+M2'M0 brror ml 38.366 0.27016 m2 0.014758 0.018999 Chlsq 0.16904 NA R 0.46142 NA 10 15 days/m 20 25 40 39 2 A 38 37 36 STATION 10 Central Calif. Margin (3444m) Core A y = M1+M2'M0 Value brror ml 38.557 0.19131 m2 •0.0039322 0.015685 Chisq 0.094529 NA R 0.17455 NA 10 days/m 15 20 5.0 STATION 10 Central Calif. Margin (3444m) Core A 4.5 4.0 3.5 o a 3.0 2.5 Chisq 2.0 15 20 25 30 10 5 0 5.0 4.5 4.0 3.5 3.0 2.5 2.0 STATION 10 Central Calif. Margin (3444m) Core B .M iv x ra~ Value I 2,?67S 0. Days/m 10 Days/m 15 20 220 2 1 0 3 200 190 STATION 10 Central Calif. Margin (3444m) Core B 180 10 15 20 0 5 220 2 A m 200 STATION 10 Central Calif. Margin (3444m) Core A 180 20 days/m y ■ M1+M2*M0 Value brror ml 66.243 27.766 m2 0.3192 0.1359 Chlsq 46.549 NA R2 0.64775 NA “ 126 1 2 0 115 STATION 10 California M argin Slope (3444m) Core B 180 185 190 195 200 205 210 215 220 S i UiM ) 140 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Numerical simulation of whole core squeezer radon pore water profiles: Methodological considerations and evaluation of benthic fluxes and rates of bio-irrigation and advection

PDF

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

PDF

Petrologic and geochronologic study of Grenville-Age granulites and post-Granulite plutons from the La Mixtequita area, state of Oaxaca in southern Mexico, and their tectonic significance

PDF

Oxygen isotopic evidence for fluid infiltration in the Mount Stuart batholith, Washington

PDF

Magma mingling/mixing in a heterogeneous, multi-pulse magmatic system: An example from the Jackass Lakes pluton, central Sierra Nevada batholith

PDF

The origin of enigmatic sedimentary structures in the Neoproterozoic Noonday dolomite, Death Valley, California: A paleoenvironmental, petrographic, and geochemical investigation

PDF

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

PDF

Radium isotopes in San Pedro Bay, California: Constraint on inputs and use of nearshore distribution to compute horizontal eddy diffusion rates

PDF

Management of large earthquake data using the Antelope Relational Database and seismicity analysis of the 1999 Turkey earthquake sequences

PDF

Multi-proxy studies of climate variability in central China: Subdecadal to centennial records in stalagmite from Budda Cave

PDF

Reinterpreting the tectono-metamorphic evolution of the Tonga Formation, North Cascades: A new perspective from multiple episodes of folding and metamorphism

PDF

Modeling of continuous tiltmeter data from the 1984 rifting event at Krafla Volcano, Iceland

PDF

Late Holocene depositional history of western shelf margin, Gulf of California, Mexico

PDF

Paleoecology and paleoenvironments of early Triassic mass extinction biotic recovery faunas, Sinbad Limestone Member, Moenkopi Formation, south-central Utah

PDF

Investigating the influence of atmospheric changes on the variability of the North Pacific using paleoproxy data and a fully coupled GCM

PDF

Fourier grain-shape analysis of quartz sand from the Santa Monica Bay Littoral Cell, Southern California

PDF

Large epifaunal bivalves from Mesozoic buildups of western North America

PDF

Paleoenvironments and paleoecology of the disaster taxon Lingula in the aftermath of the end-Permian mass extinction: Evidence from the Dinwoody Formation (Griesbachian) of southwestern Montana...

PDF

Oxygen-related biofacies in slope sediment from the Western Gulf of California, Mexico

PDF

Grain-size and Fourier grain-shape sorting of ooids from the Lee Stocking Island area, Exuma Cays, Bahamas

Asset Metadata

Creator Cummins, Kathleen Marie (author)

Core Title Silicic and germanic acids: Laboratory determination of their molecular diffusivities and field study of their benthic fluxes along the California margin

Degree Master of Science

Degree Program Geological Sciences

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

Tag geochemistry,OAI-PMH Harvest

Language English

Contributor Digitized by ProQuest (provenance)

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c16-301018

Unique identifier UC11341842

Identifier 1414874.pdf (filename),usctheses-c16-301018 (legacy record id)

Legacy Identifier 1414874.pdf

Dmrecord 301018

Document Type Thesis

Rights Cummins, Kathleen Marie

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