Natural decay series isotopes in surface waters, bottom waters, and plankton from the East Pacific

PDF

Download

Open document

Flip pages

Download a page range

Download transcript

Copy asset link

Request this asset

Request accessible transcript

Transcript (if available)

Content NATURAL DECAY SERIES ISOTOPES IN SURFACE WATERS, BOTTOM WATERS, AND PLANKTON FROM THE EAST PACIFIC by Kevin Gibbons Knauss A Dissertation Presented to the FACULTY OF THE GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (Geological Sciences) August 1976 UMI Number: DP28545 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. UMI Dissertation Publishing UMI DP28545 Published by ProQuest LLC (2014). Copyright in the Dissertation held by the Author. Microform Edition © ProQuest LLC. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code ProQuest: ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 UNIVERSITY OF SOUTHERN CALIFORNIA T H E G R A D U A T E S C H O O L U N IV E R S IT Y P A R K LO S A N G E L E S , C A L I F O R N IA 9 0 0 0 7 T h is dissertation, w ritte n by Kevin Gibbons Knauss un d e r the d ire ctio n o f hxs.... D isse rta tion C o m m ittee, and a pp ro ve d by a ll its m em bers, has been presented to and accepted by T h e G raduate S choolj in p a rtia l fu lfillm e n t o f requirem ents of the degree of D O C T O R O F P H I L O S O P H Y Dean A DISSERXATION C O M M IT T E E 7 / / /7 / Chairman CONTENTS Page ABSTRACT .......................................... INTRODUCTION ...................................... 1 Ra AND Th IN THE SURFACE WATERS OF THE EAST PACIFIC ...................................... 7 Introduction ................................. 7 Sample locations ............................. 8 Sampling and analytical procedures ......... 12 Fiber extraction........................ 12 Chemical processing of fibe r s.......... 1 h Chemical y i e l d s ........................ 17 Ra isotopic analysis .................... 18 Th isotopic analysis .................... 21 Blanks.................................... 23 Results...................................... 23 Surface East Pacific open ocean • • • • 23 Surface nearshore East Pacific Ocean . . 26 Discussion.................................... 30 Surface East Pacific open ocean • • . . 30 General oceanography ............... 30 Ra isotopes in open East Pacific • • 3h Page Tli isotopes in open East Pacific • • 45 Nearshore surface East Pacific • • • • • 52 General oceanography •••••••• 52 Ra isotopes in the nearshore environ ment • • • • • • • • • • • • • • 60 Th isotopes in the nearshore environ ment • • • • • • • • • • • • • • 71 Th^"^ and Th^"^ in surface seawater • • 73 Other radioisotopes extracted by Mn— f i b e r ............... 81 Summary • • • • • • • • • • • • • • • • • • 85 Ra AND Th ISOTOPES AND EXCESS Rn IN EQUATORIAL PACIFIC BOTTOM WATERS ........... 88 Introduction........... 88 Sampling and analytical procedures ......... 90 Fiber extraction.................. 90 Chemical processing of fibers • • • • • 96 Ra and Th isotopic analysis..... 96 Blanks............................. 96 222 226 Excess Rn and Ra analysis • • • • 97 Sample locations ............... 98 Results •••••••• .................... 99 Fiber Ra and T h ............. 99 222 226 Excess Rn and Ra ............... 99 iii Page Discussion • • • • • • • • • • • • • • • • • 10k General oceanography .................... 104 r , 222 ^ " d 226 inn Excess Rn and Ra ••••••••• 109 Fiber Ra228 117 Fiber Th isotopes...................... 123 Summary • • • • • • • • • • • • • • • • • • 126 AND Th- SERIES RADIONUCLIDES IN SESTON FROM THE EASTERN PACIFIC OCEAN .................. 129 Introduction ................................. 129 Sampling and analytical procedures ......... 130 Sampling techniqties.................... 130 Sample locations ........................ 132 Chemical techniques .................... 132 Spikes............................... 132 Chemical procedure ••••••••• 139 Isotopic analysis...................... 1^+1 Po (pb) isotopic analysis • • • • • l4l U isotopic analysis ............... 1^3 Th isotopic analysis................ 1^+3 Ra isotopic analysis ................ l44 B l a n k ................................... 1^5 Results...................................... 1^-5 Discussion................................... 151 U isotopes............... 151 Page Ra isotopes.................. 155 Pb isotopes • • • • • ............ ...... 160 Th isotopes • • • • • • • • • • • • • • • 163 Elemental comparisons........... • • • • 166 Summary • • • • • • • • • • • • • • • • • • • 168 CONCLUSIONS........................................ 170 APPENDICES........................................ 174 APPENDIX I. Instrumentation and Calculation Methods............ 175 222 Emanation method for excess Rn and Ra2 2 6 ..................... 176 Equilibrators • ••••••••••• 176 Gas extraction lines ......... 177 Radon counters and cells • ••••• 178 Data reduction • •••••••••• 179 Alpha spectrometry • •••••••••• 183 Proportional counter • • • • ........... 184 Low level beta counter ........... 185 Fiber calculations •••••• .......... 194 _ 228 . 228 1Q, , Ra via Th ...................... 19*+ Absolute Th calculation ••••••• 198 Plankton calculations •••••••••• 199 - r , , 210 ( . - p , 210\ 1 QQ Pb (via Po ) •••• 199 U . 202 Page T h .................................... 202 R a .................................... 205 APPENDIX II. Ac228 Counting for Ra228 . . . 206 Calculations . ............... ••••• 207 Problems with Ac chemistry ........... 209 Potential counting problems ........... 211 a 228 _ 228 _ , o-io Ac versus Th results ••••••• 212 APPENDIX III. Seawater U and U23^/U238 Activity Ratios .................... 216 Seawater U analyses ................. 217 Pore water U and R a .................... 219 BIBLIOGRAPHY ...................................... 22k 5 9 10 11 32 35 39 42 44 48 49 53 ILLUSTRATIONS Naturally-occurring uranium and thorium decay series ............................. 228 Continental shelf supply of Ra to the surface layer of the ocean ••••••• Station locations in the East Pacific • Locations of Mn-fiber samples taken within 1000 km of the coast of Peru . . Surface currents in the Eastern Equa torial Pacific •••• .................. 228 Surface Ra values in the East Pacific for the station locations shown in Figure II-2 ••••• .................. ppQ Surface Ra values within 1000 1cm of Peruvian Coast •••••••• .......... 228 Plot.of log Ra versus distance from coast ........................ .. Plot of apparent coefficient of eddy diffusion, Ka, versus scale of dif fusion, ................................. 228 Values of Th activity measured in the surface waters of the East Equatorial Pacific using Mn-fiber columns on station........................ .. Values of activity ratio Th^^/Ra^^ measured in surface waters of the East Equatorial Pacific .................. •• Surface circulation in the Southern California Bight •••••••••••• 55 57 59 6l 64 6 9 72 74 79 82 89 91 93 108 110 ill Seasonal variation in surface circula tion within the Southern California Bight ......... .................. Location of the California Current with in the study area ........................ Bathymetry of the Southern California Borderland •••• ...................... 228 Surface Ra activities measured in coastal waters off southern California . 228 Vertical Ra profiles taken with sub mersible pumping system 226 Surface Ra activities measured in coastal waters off southern California . Surface values of the activity ratio Th^28y, pa228 £n coastal waters off southern California ...................... Values of the activity ra tio Th228/Ra228 in vertical profiles taken with the sub mersible pumping system . ................ Sample Th alpha spectrum plotted as raw counting data • • •....................... Sample Th alpha spectrum with Pu plotted as raw counting ........................... oo£ 228 Model for supply of Ra"~ , Ra f and pn222 - j ^ o bottom waters from deep-sea sediments ••••••••• ............. Station locations for CARSAT II Expedi tion aboard R.V. T, G. Thomson . . • . • Modified Niskin bottle • • • • Location of Pacific Bottom Water flow predicted by anomalous oceanographic data 222 Bottom water excess Rn profiles • . . Figure XII- 6. Ill- 7. IV- 1. IV- 2. AI- 1. AI- 2. AI- 3. AI- k. AI- 5. AI- 6. Page Vertical profiles of bottom water Ra taken at each station using modified Niskin bottles ........................... 118 Raw data nephelometer trace from Geo secs station 33^ at 0°N, 124.5°k • • • • 124 Plankton sampling locations in the East Equatorial Pacific ...................... 13^ Plankton sampling locations off southern California............................... 133 Types of particles identifiable using the two detectors...................... 187 Plots of counter operating characteris tics: relative counting rate versus applied bias voltage.................... 189 228 Growth of Th ~ towards equilibrium with ppQ x Ra in a Ra solution initially free of T h ........................................ 195 Sample of Th alpha spectrum (plotted as raw counting data) resulting from an analysis of Th contained in a stored Mn-fiber Ra solution after ingrowth of Th228 and spiking with T h ^ O ............ 196 Sample Po alpha spectrum of plankton sample plotted as raw counting data Sample U alpha spectrum of plankton sample plotted as raw counting data 200 203 ix Table Page II- 1. Tli and Ra blank runs on Mn-oxide, acrylic fiber ............................. 24 II- 2. Fiber radium data.............. 25 II— 3* Th data for open ocean • •••••••• 26 ppQ II— 4. Nearshore Ra data •••••••••• 28 II- 5# Nearshore Th data................... 29 22 6 II- 6. Pt. Sal and Pt. Dume Ra 31 228 II- 7# Vertical Ra and temperature at Bartlett Station 2 ................. 38 228 II- 8. Previous Ra measurements.............. 46 II- 9* Fiber intercalibration ................. 47 11—10. Vertical diffusivities ••••••••• 67 11-11. Th isotopes in open ocean surface waters 77 11-12. Summary table of Th2"^2 and Th2*^ results 80 III- 1. CSII Ra228 100 III- 2. CSII T h .............................. 101 III- 3. CSII Excess Rn222 ........................ 102 22^> III- 4. CSII Ra ............................... 103 pop III- 5* Excess Rn' model diffusivities (0-90 m) 116 III- 6. Ra228 model diffusivities (0-700 m) . . . 122 IV- 1. Plankton locations................ 133 IV- 2. U2"^2 (Th228-free) spike calibration . . . 138 IV- 3* Plankton decay times for P o ........ 142 IV- 4. Plankton blanks..................... 146 IV- 5. Plankton U ......................... 147 x Table Page IV- 6. Plankton T h ............................... 148 IV- 7. Plankton P b ............................... 1^9 IV- 8. Plankton R a ............................... 150 IV- 9* Plankton concentration factors for U^~^ • 153 n n / IV-10. Plankton concentration factors for Ra 157 210 IV—11. Plankton concentration factors for Pb 162 IV-12. Plankton concentration factors for Th^"^ 165 AI- 1. Counter characteristics ................. 191 All- 1. Fiber Ra228/Ra226 ........................ 2l4 All- 2. Fiber Ra228/Ra226 ........................ 215 AIII-1. Uranium data from the Pacific Ocean . . . 218 AIII-2. Uranium in East Pacific sediment pore water and core top water ........... 221 AIII-3. ’ ’Core top” water Ra22^ .................. 223 ABSTRACT A fiber extraction technique is used to concentrate Ra and Th isotopes from 1000 liters or more of seawater. Natural Ra—226 and Th-23^ are used as tracers for the other Ra and Th isotopes. In the Equatorial Pacific the Ra-228 activity of surface waters vary from 20 to 1 dpm/1000 kg and generally decrease away from continental shelf areas. Across the Peru Current System, this decrease is modelled as one dimensional diffusion and indicates the possibility of two flow regimes with distinct characteristic mixing lengths c r 7 2 and apparent eddy diffusivities of 10 and 10 cm /sec. The perturbing effects of advection and equatorial upwell— ing west of the Galapagos Islands upon this simple dif fusion model for the surface layer of the Pacific are noted. Off the coast of southern California the vertical Ra-228 distribution is used to estimate mixing rates through the upper thermocline and apparent diffusivities of 1 to 3 cm /sec are obtained. Ra-226 concentrations near the coast delineate the importance of continental shelf supply of this isotope to the surface layer of the open ocean, The insoluble daughter/soluble parent activity ratio Th-228/Ra-228 in the Equatorial Pacific surface waters displays latitudinal trends which may be correlated with productivity variations* Near the coast of California, this ratio reflects the differing oceanographic conditions north and south of Pt* Conception indicating a shorter mean time of chemical removal of Th and other highly re active elements within the southern California Bight* The Th-232 content of seawater sampled was indis tinguishable from blank and is certainly less than 0.1 >ug/l00o/ suggesting most published seawater Th-232 values may be too high* Profiles of Ra-228 in the bottom kilometer of the Equatorial Pacific display to varying degrees the effect of eastward flowing Pacific Bottom Water (PBW). Modelled as one infinite layer the PBW is shown to have apparent vertical diffusivities increasing from west to east, per haps due to its interaction with the East Pacific Rise. The source strength of the sediment for Ra-228 decreases from west to east reflecting the dilution effect of carbon ate sedimentation at shallower depths* Advective processes appear to play an important role in governing the bottom Rn—222 distribution. The standing crop of excess Rn-222 profiles vary and show the same trend as for Ra—228. At two stations the Rn-222 profiles are xiii remarkably similar to those taken in previous years nearby, arguing for a steady state feature. The activity ratios Th-230/U-23^ and Th-228/Ra-228 indicate removal processes for the reactive elements in the deep sea act on a time scale significantly longer than in the surface ocean. The Th-232 and Th-230 content of bottom water is invariably higher than in surface waters. A chemical procedure involving sequential isolation and analysis of several isotopes of geochemical interests U, Th, Ra, Po, Pb has been developed. The relative concen tration factors for these isotopes by plankton indicate the sequences Th^> Pb RaU. Calculation of the degree of association of the above elements with the plankton demonstrates that the plankton uptake is very significant in the case of Th, much less so with respect to Pb, and Ra, and of little significance with respect to the U content of the photic layer. As potential oceanographic tracers of surface water movement, the U and Ra content of plankton are of little use. Of the radioisotopes studied here, the most useful potential traces, those displaying water mass differences and relatively high concentration in plankton, are: Th-228, Pb-210, and Po-210. A comparison of the two indirect methods of deter mining Ra—228 (via Ac—228 and via Th—228) made on 6k sea water samples shows that the time delay required by the xiv Th-228 method is more than compensated by its better analytical simplicity and precision. | The fiber extraction technique utilized here has large potential for future research with trace radio elements. The removal of Pu and Ac from seawater by this technique was demonstrated. xv ACKNOWLEDGMENTS Professor T.-L, Ku provided the original impetus for this work and enkindled in me his enthusiasm for science* His continuing support and encouragement are gratefully acknowledged. Dr. W, S. Moore provided ship—time, equip ment, and advice throughout this study and introduced me to the Mn-fiber extraction techniques he developed. Dr. B.L.K. Somayajulu gave freely of his time and ideas concern ing low-level beta counting. Running conversations and arguments with Quay Nary benefitted this work and provided constant entertainment in the lab and at sea. Many others provided assistance in one form or another and I would particularly like to thank Dave Reid, Blaine Hartman, Dr. J. L. Bischoff, Ed Ruth, M.-C. Lin, Jorge Sarmiento, Dr. Y.-C. Chung, Steve Murray, Ross Horowitz, and Steve Prensky for their help in sample col lection and/or data processing. Craig Todd and John Wilson were instrumental in setting up the low-level beta counting system. Dr. D. E. Hammond assisted in data interpretation. The aid in sample collection of the officers and crew of the following research vessels is gratefully acknow ledged: MENDELEEV (institute of Oceanology, Moscow), xvi 1 BARTLETT (Navoceano), VELERO IV (U.S.C.), MELVILLE (S.I.O.) and THOMSON (U. of Wash.). This research was supported by various grants from the office of XDOE and Oceanography section of the National Science Foundation to Dr. T.-L. Ku. i The understanding and encouragement of my wife Darnell enabled me to carry this study to completion and provided a constant source of strength to me. xvii INTRODUCTION The parents of the three naturally-occurring radio- poo p oQ po c active decay-series: Th , U , and U have existed since the time of formation of the earth and through the process of radioactive decay are continuously generating their shorter—lived radioactive daughters. These radio nuclides were initially contained in the crust of the earth and were then introduced into the oceans by runoff from the continents as part of the weathering products and by pre cipitation from the atmosphere of the daughters of noble gas members of the decay series. If following this initial introduction the daughter products are not separated from their parents, then after a sufficient period of time in any given part of the oceanic system a situation of secular equilibrium should be established where the activity ratio of any parent-daughter pair is equal to one. In the marine environment, however, the concentrations of the members of these series are, in general, lower than in the earth*s crust and their ratios are strikingly dissimilar (Horne, 1969). In the oc eans there is a disruption of the radio active equilibria by the biogeochemical processes operat ing there. 1 In the oceans the distributions of the dissolved radionuclides are effected by the physical-chemical and biological environments. The physical-chemical processes involved include: adsorption, co-precipitation, and/or ion-exchange with the particulate matter in seawater. In the biological cycles the concentrations of these nuclides are altered by: incorporation into the organisms* struc ture, accidental trapping of particulate matter, or feeding on other organisms which had previously incorporated radio nuclides (Mauchline e_t al. , 1964). These processes are such that some elements are more liable than others to be re moved from seawater and incorporated in the solid phases which may ultimately become part of the marine sediments, resulting in radioactive disequilibria among the various isotopes of the decay chains. Assuming that through the passage of time a steady- state situation has been achieved for the isotopes; i.e., the processes of: supply to seawater, transit within the fluid phase (with subsequent radioactive decay), and re moval from seawater are all balanced, then the disequili brium relationships can be used to obtain such information as residence time (hence reactivity) of elements in the ocean, mixing rates, and the biogeochemical pathways of trace elements in the marine environment. Naturally, this information is attainable only if the quantities and vari ations of these isotopes in various parts of the ocean can be measured# For many of these radionuclides the measure ments are quite difficult and require the use of very large samples# In order to study the effect of biogeo chemical cycling causing the disequilibria, the distribu tion of these nuclides must also be determined in marine plankton# The aim of this study was to determine the distri butions of selected members of the uranium and thorium decay series in surface waters, bottom waters, and in marine plankton. Surface water samples were collected from both the open—ocean and near-shore areas of the East Pacific off North and South America# The distribution of these nuclides in these waters is used to study the hori zontal advective-diffusive rates in these areas. Input 226 228 _ 210 . x . sources of certain nuclides (Ra , Ra , Pb , etc.) to the surface layer were also studied# Profiles were taken through the mixed layer and upper regions of the thermo— cline to study vertical mixing rates. Plankton samples taken from the same area were analyzed radiochemically in order to assess their importance in the marine cycles of those isotopes of interest. The analyses provide informa tion on the relative concentration factors (selectivity), degree of association with plankton, and mean removal times for these isotopes. Bottom water profiles were taken in the Eastern Equatorial Pacific in order to study mixing processes near the sediment—water interface. Also, 3 the relative source strength of the sediments for these isotopes was determined* An attempt was made to explain the distributions of these isotopes in surface and bottom waters using simple, one-dimensional mixing models involv ing diffusive parameters. The diffusion coefficients derived from the model calculations bear on the dispersion rate of pollutants in the marine environment. This should be of particular interest in light of the heightened concern over man*s contamination of his planet. The nuclides chosen for this study (Fig. I-l) were selected on the basis of their half-lives and chemical , , _ 232 . _ 232 228 , 228 behavior: from the Th -series: Th , Ra , and Th ; , _ __238 . TT238 T T23h 230 _ 226 222 and from the U —series: U , U , Th , Ra , Rn , 210 and Pb . Some of the isotopes of Ra and Th had to be concentrated free of contamination from several thousand liters of iirater. This was accomplished using a fiber ex traction technique. The plankton analysis involved developing a technique which sequentially isolates: Po, Ra, U, Th, and Pb from the same sample. The determination of a whole suite of radioisotopes on the same sample is useful in understanding the complex inter-relationships be tween the isotopes contained within the natural decay series members. The study is divided into three parts (chapters): surface waters, bottom waters, and plankton. Each part has been written more or less independently with separate Element U-238 Series ♦ Th-232 Series Neptunium I Uranium U-238 4.49 X 10» yrs U-234 2.48 X 10* yrs Protactinium Pa-234 1.18 -.m in Thorium Th-234^ 24.1 days Th-230 73 X 104 yrs Th-232 1.39 X 10l# yn Th-228 1.90 Actinium Ac-228^ 6.13 hrs Radium Ra-226 1622 yrs Ra-228^ s.is y* Ra-224 3.64 days Francium Radon Rn-222 3.825 days Rn-220 5 4 5 sec Astatine Figure 1-1. Naturally-occurring uranium and tiioriura decay series* 5 introductions, techniques, discussions, etc. so as to bring related data into better focus. The models used to calculate mixing rates and removal times using the ob served distributions are described in the appropriate dis cussion sections of each chapter. Several appendices are included dealing with instrumentation, comparisons of analytical techniques, and a related study dealing with the U isotopes in seawater and porewater of sediments. 6 Ra AND Th IN THE SURFACE WATERS OF THE EAST PACIFIC Introduction The distributions of the isotopes of Ra and Th in the surface waters of the ocean have potential applications to the study of certain oceanographic processes. They may be used to study lateral and vertical advective-diffusive processes as well as chemical removal processes in the surface layer of the ocean (Moore, 1969; Broecker £t al,♦ 1973)* Their utility has been restricted, however, by the difficulties involved in their measurement. This chapter will focus primarily on the isotopes, Ra^^^ (t 7 5 y ) and Th22^ (t^y2=1.917) which are of use as tracers in studies of surficial processes occurring with a few months to a 30 y time scale. Secondarily, a discussion will be made concerning the distribution in the 22 6 222 surface ocean of the isotopes Ra (t^y2=l620y), Th = x 1010y) , and Th2-^0 x • ^ee Figure 1-1, Early studies by Koczy e_t al. (1937)> Moore (1969a, 2 p Q b) , and Kaufman e^t al. (1973) indicated that Ra is primarily supplied to the surface ocean by diffusion from 2 32 continental shelf sediments containing its parent, Th . After injection along coastlines it is spread across ocean basins primarily by the wind-driven currents. During this 228 process it decays producing its daughter, Th , in the surface ocean (see Fig. XX-l). However, the biologic and chemical processes acting in the surface layer cause a fractionation between the elements Ra and Th, resulting in the preferential removal of Th. The above scheme results in the following relative activities in surface waters: Ra228> Th228 Th232 By modeling the manner in which this observed dis equilibrium is produced, it is possible to derive dif fusive and chemical removal rates. Sample Locations The samples for this study were collected on several cruises. On cruise $137301 of USNS Bartlett 3^ samples were taken from February-April 1973* On Legs J and K of the Pacific Geosecs (Geochemical Ocean Sections) cruise by R/V Melville 11 samples were taken from April- May 1974. On 2 cruises (#1282 and #129l) of R/V Velero IV. a total of 25 samples were taken in May and August of 1975* The locations of the 70 fiber samples are shown in Figures II—2, II-3, and 11-15. 8 <0 CD C V J OJ GO C V J C V J GO C V J C V J GO C V J C V J C V J ro C V J V * ^g g vl fe vP 22 8 Figure II-l. Continental shelf supply of Ra to the surface layer of the ocean. 9 150 120 90 60 347B 3 4 6 a 3 4 5 a . St5-. St.4 St3 —BARTLETT STATIONS ■-GEOPAC* STATIONS a - GULF of CALIF STATION ®- PREVIOUS Ra MEASUREMENTS , 150 120 90 60 Figure II-2. Station locations in the East Pacific. Mn—fiber samples were taken by pumping surface water through fiber-filled columns while on station. 10 90 85 80 75 o PARTLETT STATIONS WITHIN 1000 K m OF COAST 5 10 BH3 St 9 * B-15 M6V B-19 • • ^ ** B-18 W # CALLO, PERU B-42 B“43 B-41 b 42 B-39 St I— • 15 St 7 • C-2 • 90 85 80 75 Figure IX-3. Locations of Mn-fiber samples taken within 1000 km of tbe coast of Peru. B- and C-samples were taken by tow ing a fiber-filled sampler behind the ship while underway. St samples were taken by pumping surface seawater through fiber-filled columns while on station. 11 Sampling and Analytical Procedures Fiber Extraction Prior measurements of Ra and Th. in seawater largely relied on shipboard co-precipitation with Ba SO^ and Fe(OH)^ using large volumes of water. Sample volumes ranged from as little as 25 liters (Sackett ej; al., 1958) to 1500 liters (Kaufman et al., 1973)• However, the very low values observed previously indicate that for most oceanic samples a volume of 1000 liters is required for ^ + • 1 • n r . 228 _ 232 _ 230 228 determinations of: Ra , Th , Th , and Th . The difficulty in processing these large water samples is largely responsible for the limited number of papers deal ing these radioelements. Moore e_t al (1973) developed a method for extract ing Ra from seawater using Mn-impregnated acrylic fibers. Subsequent improvement by Moore (preprint, 1976) of the MnO^ deposition process employing permanganate oxidation of the acrilan enables nearly quantitative extraction of both Ra and Th from 1000 liters or more of seawater onto 100 g of Mn0o fiber. The fiber is prepared by immersing it acrylic fiber (Monsanto "Acrilan”) in 0.5 M potassium permanganate solution heated to J0°C» The permanganate oxidizes the acrylic fiber and deposits MnO^ . The exo thermic reaction is quenched after 10—15 minutes by plac ing the fiber in a de-ionized water wash. After washing 12 and drying, tlie blackened acrylic fiber is ready for use. 226 Since one isotope of Ha (Ra ) and one isotope of o r t l ± Th (Th ) are present in seawater at concentrations easily measured in small volumes (20 liters) of water, they can be used as natural tracers for the other Ra and Th isotopes, minimizing the spike equilibration problems (Kaufman 1969)* Thus, a fiber sample and a small water sample from the same location allow the absolute determina- „ „ 228 _ 226 232 _ 230 228 , _ 23^ tion of Ra , Ra , Th , Th , Th , and Th . Water samples from the surface layer of the ocean were collected in several ways. On station, samples from the upper 3 m were collected by pumping from the ship*s salt water line through two fiber columns in series. The columns used were commercially available (Culligan Flavr- Gard) units with all plastic construction and containing an internal cartridge (initially filled with charcoal) into which 30 to 100 g of Mn-fiber were loaded. After passing through the columns, the flow was directed through a water meter used only to estimate the volume sampled and then passed over the side. Routinely, with flow rates of 5 to 13 liters/min, the pumping time is about 2 to 3 hours for each sample. Using the flow rates and sample volumes mentioned, the Mn-fiber extraction efficiencies were on the order of 90 percent or better. Samples from the upper 200 m were collected on station by using two fiber columns in series as above, but 13 the water was supplied by a submersible pumping system. This system consisted of a domestic water well pump (Sub urban) and several hundred meters of 1 inch ID re-inforced rubber hose. The pump had a 3/k 220 volt motor, plastic impeller, and stainless steel body. The hose was mounted on a motorized reel running on the same 220 volt line. Between stations samples were collected while under way by towing a Mn—fiber filled PVC sampler behind the ship. The sampler was fashioned from PVC pipe covered with a PVC sheet perforated with holes. PVC screening material was used to prevent fiber washout. The sampler was towed using polyester line at a sufficient distance from the ship such that it remained submerged at all times. The ship1s speed while sampling was usually around 8 to 10 knots, and sampling times of 2 to 3 hours were used. These samples thus averaged the Ra and Th content over several tens of kilometers. The fiber samples collected by one of the above means were placed in plastic bags and returned to the lab for analysis. Where applicable 20 liter splits of sea water were collected, stored in 3 gallon plastic cubi- tainers and returned to the lab for analysis of Chemical Processing of Fibers The chemical process used is designed to produce lk pure fractions of Ra and Th. suitable for isotopic analysis. The fiber is boiled in several liters of 6 N IIC1 using a 4 liter pyrex beaker and a glass coffee percolator to prevent bumping (Moore and Reid, 1973)• After several hours, the Ra, Th, Mn, etc, are almost completely in solu tion and the fiber is again white in color. This clear HC1 solution of MnCl^ is then filtered through a fine filter paper and stored in plastic containers until fur ther processing. The dry weight of the clean fiber is recorded. The above solution is put into a pyrex beaker and a magnetic stirrer is used to stir the solution. ¥hile stirring, 3 to h ml of a saturated BaCl^ solution and 50 ml of concentrated H^SO^ are slowly added to effect a BaSO^ precipitation, and the solution plus precipitate are stirred for 60 minutes. This BaSO^ precipitation scavanges both Ra and Th very efficiently. The precipitate is al lowed to sit overnight. After decanting the supernatant, the BaSO^ preci pitate is centrifuged, washed several times with de ionized water, and dilute HC1, and transferred to a glass beaker where it is digested by gentle heating in dilute HC1 for 1 hour. The precipitate is again centrifuged, washed several times with de-ionized water, and filtered onto a vacuum drawn 0.^5 M Millipore filter. The filter plus BaSO^ precipitate is dried at 15 110°C for 1 hour. After igniting the filter with methanol in a platinum crucible, the BaSO^ precipitate is fused with excess anhydrous Na2C0^. The fused pellet is broken up in an ultrasonic bath and washed repeatedly with de —ionized water until the washings have a pH^ 7.5. At this point the remaining BaCO^ is dissolved in a minimum of HC1. Normally a small amount of insoluble residue re mains. Invariably a large portion of the Th is contained within this residue and for all samples a second carbonate fusion step is necessary. The Ra, however, was found to be largely (^90 percent) in solution and if only a Ra analysis is desired, the second fusion can be safely omitted. Several drops of a concentrated FeCl^ solution are added to the combined dilute IICl solutions. The solution is heated and allowed to cool. A Fe(OH)^ precipitation is made by adding NH^OH. This precipitation effectively separates Th (precipitate) from Ra (solution). To the Ra solution, a solution of concentrated Na^CO^ is then added to precipitate BaCO^. This precipi tate is washed repeatedly with de-ionized water to remove excess NH^ and dissolved in a minimum of HC1. The BaCI,-, (with contained RaCl^) solution is dried and then taken up to about a 5 nil solution with de-ionized water and the pH adjusted with HC1 to about 1 to 2. A cleanup extraction for any last traces of Th present is 16 made by emulsifying the aqueous phase with a 0,25 M TTA- benzene solution using a micropipet. The pH is then ad justed to 5*5 to 6.0 with NaOH and a second TTA cleanup is made to remove Ac. The purified Ra solution stored at 226 22f t p H 1 is ready for Ra and Ra analysis. Following Ku (1966), the Fe(OH)^ precipitate is washed successively with hot, 3 N NaOH, and de-ionized water, and taken up in a 4 N HC1 solution. The solution is passed through a cation exchange column (Dowex 50W x 16) about 4 to 5 cm long. After washing with several columns of 4 N HC1, about 40 ml of 0.75 M oxalic acid is used to elute the Th from the column. The oxalic acid is then de composed by fuming with a HNO^-HCIO^ mixture. Several + 3 hundred micrograms of La is added to prevent Th adsorp tion on the glass beaker. After fuming to dryness, several ml of HCIO^ are added to insure complete removal of oxalic acid, and the solution again is taken to dryness. The Th fraction is then ready for isotopic analysis. Chemical Yields Several methods were used to estimate the chemical yields of the above procedure for Ra and Th. 226 An NBS Ra standard solution was run through the entire chemical procedure starting with the first BaSC^ precipitation. The final Ra solution was then analyzed 2^5 for Ra ^ by the emanation method (see Appendix l), Re- 17 coveries for 4 analyses ranged from 25 to 40 percent. Alternatively, the overall chemical efficiency (including the fiber extraction efficiency) could be estimated on all fiber samples pumped on station through the columns. By knowing the total volume passed through the column (flow 226 meter readings), the initial Ra content of the sample 22 6 (20 liter split), and the final Ra content of the Ra solution resulting from the fiber chemistry, the overall efficiencies were found to range from about 25 to 60 per cent, with most samples being about 30 to 40 percent. Xn a similar manner by knowing the total sample 234 234 volume, initial Th content of seawater, and final Th content of the Th fiber solution, the overall Th chemical efficiency was found to range from 20 to 80 percent with 234 most samples being around 50 percent. The Th was deter mined by counting p.a^^ (see Appendix X) . Radium Isotopic Analysis The final Ra solution (pH-^l) was first analyzed for 228 228 Ra by measuring the immediate daughter Ac (half- life = 6.13 hours) which grows into equilibrium within 228 228 about 30 hours. Although both Ra and Ac are beta- 223 emitters, the beta-emission from Ra is relatively "soft” p p Q and difficult to detect (Ra jB = .048 Mev Emax, 228 Ac Js =1.11 Mev Emax). On the other hand, Ac emits betas that are sufficiently hard to be easily detectable 18 (see Appendix l) and its short half-life allows it to be re-analyzed every 30 hours. To accomplish this, the Ra solution is taken to dryness, and taken up in a sodium acetate-acetic acid buffer solution (pH = 5*7). Use of the buffer was found to increase the reprodueability of the TTA extraction which follows, compared to adjusting the pH each time with NaOH (see Appendix II). The TTA extraction was repeated and the benzene solution was mounted dropwise on a stainless steel plate and flamed gently. Ac on the plate was measured by low-level beta 228 counting and thus the Ra activity of the final fiber Ra solution was determined. This Ac measurement was repeated 3 times for each sample and the results averaged. The precision was - 15 percent routinely (see Appendix XX). After completion of Ac measurements, the Ra solu- 2 2 6 tion itfas analyzed for Ra via the emanation method (Broecker, 1965)* The solution (pH-^l) was placed in a 200 ml equilibration flask, flushed with He to remove any 222 Rn present and sealed. After 7 to l4 days, the solu- 222 tion was analyzed for its Rn^ content (see Appendix i). Radon 222 (half-life = 3,82 d) is the immediate daughter 226 of Ra , and by knowing precisely the time in which it is allowed to grow towards equilibrium with its parent, 226 the Ra content of the fiber Ra solution is determined. 226 222 This determination of Ra via Rn was repeated 2 or 3 times and the results averaged. The precisions for these 19 ■ J - analyses were always better than - 5 percent. The Ra solution (pH^l) was then stored in teflon 228 (or nalgene) bottles to allow the next daughter of Ra , ^28 i.e., Th"~ (half-life = 1.91 y) > to grow towards equili- 230 briura. Sometime near the midpoint of storage a Th spike was added to the initially Th-free solution, to act as a 228 yield tracer for Th which was growing towards equilibrium in the fiber Ra solution. Storage times ranged from 230 to 580 days. These times correspond to a range of 20 to 40 percent of equilibrium (see Appendix i). After storage the fiber Ra solution was taken to dryness and taken up with HC1 to a pH 1—2 solution. A TTA extraction was performed (twice) and the benzene phase mounted on a stainless steel 230 plate and flamed. The plate was then analyzed for Th 228 and Th on an alpha-spectrometer (see Appendix l). This provided a second indirect determination of in the fiber Ra solution. In this case precision was always better than - 15 percent and the vast majority of samples had precision better than - 6 percent. From the above Ra analyses the ratio Ra~^^/Ra^^^ contained in the fiber Ra solution is obtained. This ratio 226 when multiplied by the Ra ~ content of the assocxated 228 water sample (20 liters) gives the Ra content of the 226 original water sample. For this reason, the Ra content of the 20 liter splits was determined whenever necessary (only in nearshore samples) by the emanation method des— 20 cribed above# Xn this case 20 liter glass equilibrator bottles were used (see Appendix X). It was found that open ocean surface waters north of the Antarctic Convergence 226 have a rather constant Ra content (Broecker et al#, O 1967)> and so a constant value of 65 dpm/lO kg was used# Analysis of several 20 liter splits associated with "un- . ^ 228 226 ^ . usual" fiber Ra /Ra ratios confirmed this value# Thorium Isotopic Analysis The dry beaker containing Th is taken up in 0#1 N HNO^ and two TTA extractions are made to extract Th. The benzene phase is evaporated dropwise on a stainless steel plate, flamed, and is ready for isotopic analysis for Th. Generally, the time between sample collection and final plating of the purified Th fraction resulting from the fiber chemistry varies from 1 to as long as 3 months# This means that the fiber Th^^ (half-life = 24# 1 d) has decayed to a value about 50 to 10 percent of its original activity# Hence, by necessity the determination of 234 Th must be made as soon as possible# This was done p o U indirectly by counting the Pa (half-life = 1#18 min) which quickly grows into equilibrium on the plate with its 234 immediate parent, Th . Again, although both are beta- o o 2 , 1 o 0)1 emitters, the Pa beta is much more energetic (Th — p O 4 0.19 Mev Emas; Pa =2.29 Mev Emax) and hence easily 21 determined. Depending on the activity encountered, the 234 Pa determination was made either by counting in a gas- flow proportional counter (background^ 2 5 cpm) or a low- level geiger counter (background ^ ,06 cpm) (see Appendix I). The stainless steel plate was covered with an Al foil adsorber to prevent any alpha or weak beta discharge of the gas. The Th plate was analyzed in an alpha spectrometer 232 ?30 for determination of the activities of Th , Th and 228 Th . By knowing the relative beta-counting efficiencies versus the alpha spectrometer (see Appendix l), the result ing fiber ratios: Th232/Th23\ Th23°/Th23^, Th228/Th23^; 23b / and the Th activity of the seawater (associated 20 liter split), it is possible to calculate the activities of Th232, Th23° and Th228 in the seawater. However, in 2 34 practice, Th was not directly determined on the 20 liter splits. As demonstrated by Matsumoto (l975)f the 2 34 median value of 84 analyses for Th activity in the upper 200 meters of the open Pacific ocean is 2.0 - .3 dpm/f . This value has an uncertainty of i 15 percent which, as will be seen, is of a similar order of precision as the alpha determinations of Th isotopes. Thus, a value of 2 34 2 dpmUl was assumed for the Th activity of surface waters. 22 Blanks Blanks were run on 5 MnO^ fibers taken from the several batches prepared. The results are shown in Table XX—1. The fibers were run unspiked so as to check all Th 234 blanks, including Th . Blank Th values were computed by assuming a 30 percent efficiency, which was close to the average for the samples. Similarly, for the Ra blanks no spike was added and an assumed efficiency of 40 percent 226 234 was used. The fiber blank values for Ra and Th were in all cases negligible compared to the sample activities. 228 228 For surficial Ra and Th the blank became of some significance in only a few cases, particularly the open ocean South Pacific Geosecs samples. However, the fiber 232 230 blank was significant for surficial Th and Th . This matter will be discussed later. Results Surface East Pacific Open Ocean The results for the 43 surface open ocean fiber samples are presented as follows: Table II-2 shows the results of the fiber Ra isotopic analysis. The values for 228 226 Ra ~ /Ra illustrated here were computed using only the 228 228 "grow—in" Th determination of Ra , because the precision of this method was considerably better than that 223 obtained by Ac counting (see Appendix II). Table II-3 23 Table II-1, Th and Ra blank runs on Mn-oxide, acrylic fiber Blanlc Number Th-;;3:i(dPm/sxio'') Th-230(dpmZgX103) Th-228(dpm/gX103) Ra-22 8(dpm/gX10^) 1 4.6 1.^3 1.26 3.8 2 2.6 1.16 1.13 1.2 3 6.0 1.80 2.8 1.0 4 7.3 1.88 1.4 1.6 5 6.9 _... 2.03. _ . 1.3 . 0.0 Average 5A - 1.7 1.6 ± .33 . 1.6 i .60 1.5 - 1.2 Note: Fiber weights used for samples varied from about 70 to 120 gms, averaging about 100 gm. Table II-2. Fiber radium data Sample Ra-228 (dpm/10 kg) Ra-228/ Ra-226 Location Sample Ra-228 (dpm/10 kg) Ra-228/ Ra-226 Location Geopac 322 .98 .015 43°01 s, 129"57* Bartlett 15 5.92 .091 07°30 N, 94*30' 320 .98 .015 33°21 S, 128° 24f 17 & 18 5.40 .083 09° 50 N, 94°45' 324 1.14 .017 22° 59 s, 146°04* 19 5.98 .092 10°00 N, 96°45* 326 1.96 .030 14°03 s, 126°16» 20 6.18 .095 12° 40 N, 98°12* 331 2.10 .032 4° 37 s, 125°091 B-13 5.07 .078 10° 55 s, 87°29'W 334 3.58 .055 0°04 N, 124° 34' B-14 6.44 .099 11° 19 S, 81°05* 337 3.09 .048 4° 51 N, 124°05* B-15 6.11 .094 11° 46 s, 79°49* 343 6.81 .105 16° 32 N, 123°01» B-16 7.67 .118 11° 35 S, 78° 38* 345 5.04 .078 22° 32 N, 122°13' B-18 7.80 .120 11°50 s, 78° 20* 346 5.57 .086 25° 29 N, 121° 511 B-19 5.27 .081 11°54 s, 80°05* 347 4.84 .074 28° 31 N, 121° 291 B-22 5.85 .090 10°20 s, 86°33f Bartlett 1 9.88 .152 14° 30 s, 76°30'W B-39 4.03 .062 13°08 s, 85°00' 2 2.41 .037 17°53 s, 76°31’ B-41 4.23 .065 12° 44 s, 82° 55' 3 3.90 .060 18°57 s, 75°03f B-42 5.20 .080 12°32 s, 81°33' 4 4.29 .066 19° 37 s, 76°34f B-43 6.44 .099 12° 18 s, 80°24f 5 4.88 .075 19°15 s, 78°07» C- 1 19.37 .298 12° 32 s, 77°10' 7 5.33 .082 15° 24 s, 80°42» C- 2 6.5 .100 15°41 s, 76° 35' 9 3.84 .059 11°09 s, 88°15' C- 3 7.15 .110 18° 33 s, 75° 35* 10 6.76 .104 04° 25 s, 93° 20* C- 4 2.47 .038 19°30 s, 76°58' 11 3.19 .049 02°00 S, 94° 32* C- 5 7.215 .111 16°42 s, 78° 52' 12 3.38 .052 00° 30 N> 94°30’ C-14 4.62 .071 08° 41 N, 94° 30'W 13 5.33 .082 02° 10 N, 94°35! C-18 6.70 .103 15° 54 N, 99°27' C-19 9.69 .149 16° 10 N, 99°20' 3 226 226 An assumed value of 65 dpm/10 kg for Ra is used, except for Bartlett 2; the measured Ra value of 66 dpm/lO^kg is used. ?o Table II-3. Th data for open ocean Sample 232 Th * (dpm/10 kg) Th23® (dpm/10 kg) Th223 (dpm/10 kg) Th228/Ra228 Geopac 322 .041 (25) .139 (23) .223 (14) .227 320 dl .085 (79) .177 (37) .181 326 dl dl .100 (51) .049 331 dl dl .145 (20) .069 334 dl dl .359 (17) .101 337 dl dl .173 (17) .056 343 dl dl .520 (05) .076 345 dl dl .689 (05) .137 346 .020 (90) .085 (79) .789 (08) .142 Average .030 .103 .353 .117 ( ) = maximum blank contribution in percent, dl = detection limit, i.e., signal ^highest blank value. shows the results of the fiber Th isotopic analysis. Only values greater than the largest blank are shown. As will be explained below, this requirement eliminates most of the an<j Th^~^ numbers, but not the Th^^ numbers. The 228 Ra numbers are significantly greater than the blank as well. As noted by Kaufman et al. (1973) > ‘ the extent of the required blank correction is not precisely known; hence, all values listed are not blank corrected. Only the re striction that all data be greater than the largest blank has been applied. The precision in the final open ocean 228 + Ra number is - 10 percent or better in all cases with the exception of Geosecs Station 33^* The precision in the 228 O o|| final Th number is limited by the assumed Th'' activity, hence, - 15 percent. The few Th^^ and Th^*^ values that are significantly greater than blank have counting errors of i 10 to - 17 percent. Surface Nearshore East Pacific Ocean The results for the 25 fiber samples obtained using the submersible pump-fiber column system off the coast of California are presented in the following tables. Table II-4 shows the results of the fiber Ra isotopic analysis. q 8 22 8 Again, the "grow—in" Th/'‘ ~ determination of Ra was used. Table XI—5 shows the results of the fiber Th isotopic analysis. For these nearshore Ra and Th data the blank 228 Table II-4. Nearshore Ra data Sample No* Location (N,W) Fiber Ra228/Ra226 _ 226 Ra (dpm/lOOkg) „ 228 Ra (dpm/lOOkg) PS1- 0 34°59', 120"43* • 262 9.577 25.1 - 40 • 177 (9.6) 17.0 2- 0 34°51 *, 121°08' .095 9.740 9.3 - 70 .078 (9.0) 7.1 -120 .074 8.452 6.2 3- 0 34°43', 121°35f .096 9.609 9.2 - 80 .089 (9.25) 8.2 -150 .076 8.904 6.7 4- 0 34° 341, 121°08' .095 8.659 8.2 - 60 .076 (8.75) 6.7 -150 .059 (8.9) 5.3 Gulf Cal. 25°42', 109°42» .737 9.0* 66.0 PD2- 0 34°00», 118°571 .116 8.471 9.8 - 60 .127 (8.698) 11.0 -150 .105 8.925 9.3 3- 0 33° 44', 119° 35* .119 7.389 8.8 - 30 .092 (7.594) 7.0 - 75 .076 (7.799) 5.9 -150 .059 8.004 4.7 4- 0 33° 371 , 120°08' .097 7.302 7.1 - 90 .071 (8.182) 5.8 -150 .048 9.062 4.3 5- 0 33°28', 120°40» .096 7.844 7.5 -150 .069 8.739 6.0 - V 00 * 226 = interpolated* All other Ra = Bruland et al* (1974)* results are measured values. Table II-5. Nearshore Th data Sample No. Th232 (dpm/10 /) Th230, (dpm/10 I) ■m.228 Th , (dpm/10 /) Th228/Th232 Th23°/Th232 Th228/Ra228 PS1- 0 .605 (10) .721 (27) 2.27 (08) 3.76 1.19 .09 - 40 1.56* (06) 1.77 (15) 3.22 (06) 2.38 1.13 .21 2- 70 .204 (32) dl 1.69 (11) 8.33 dl .23 -120 dl dl 1.92 (28) dl dl .31 3- 0 dl dl 1.59 (16) dl dl .17 - 80 dl dl 1.38 (18) dl dl .17 -150 dl dl 1.34 (14) dl dl .20 4- 0 dl dl 1.68 (16) dl dl .20 - 60 dl dl 1.56 (10) dl dl .23 -150 dl dl 1.25 (27) dl dl .24 Gulf Cal. .40 (21) .40 (64) 8.22 (03) 20.68 1.01 .12 PD2- 0 .045 (72) dl .602(15) 13.5 dl .06 3- 30 dl dl .903(13) dl dl .13 - 75 dl dl 1.04 (10) dl dl .18 -150 dl dl 1.52 (05) dl dl .32 4- 0 dl dl 1.11 (22) dl dl .15 - 90 dl dl 1.02 (05) dl dl .18 -150 .031 (73) .086 (80) 1.35 (04) 43.6 2.76 .31 5- 0 dl dl .792(25) dl dl .10 - 75 dl dl .779(13) dl dl - -150 dl dl .986(06) dl dl .16 dl = detection limit, i.e., ^highest blank. - = lost Ra. ( ) * contribution of blank in percent, w * a particulate contamination bottom at 45 m. restriction described above was also used. The precision 228 in the final nearshore Ra number is in all cases better + 228 than — 8 percent. The precision in the final Th number, + 226 as above, is — 15 percent. The precision in the Ra analysis of the 20 1 splits is better than - 6 percent. Table II-6 shows the results of the 23 Ra~^° analyses of the 20 / splits associated with the fiber samples, as well as some bottom water samples taken at each station. These bottom samples were taken with a Niskin bottle modified to be tripped when the bottle was 5 m off the bottom. The water samples associated with fibers were taken with the submersible pumping system at the same sampling location as the fiber. Discussion Surface East Pacific Open Ocean General oceanography The surface circulation patterns for the East Equatorial Pacific have been reviewed by ¥yrtki (1966). Figure II-4 taken from this text illustrates the names and locations of the major current systems in the area of this study. The positions of these current systems vary con siderably in response to the shifting of the major wind systems. The circulation is dominated by the anticyclonic gyrals in the North and South Pacific Ocean. These consist 30 Table II-6. Pt. Sal and Pt. Dunne Ra^^^ Sample 226 Ra (dpm/lOO kg) Open Ocean Values at G0G0 I (29°N, 122°¥) from: Chung et al. (1974) IPS 1-0* 26.1 + • 6 1-20 9.6 + .3 2-0 9.7 + • 6 2-120 8.5 + .3 2-bottom (550 m) 15.9 .7 (l4.2 t .2) 3-0 9.6 t .3 3-150 8.9 + .3 3-bottom (786 m) 21.5 + .5 (17.6 i .2) 4-0 8.7 .3 4-bottom (850 m) 24.2 + • 6 (21.9 - .*0 PD 2-0 8.5 + .3 2-130 8.9 i .3 2-bottom (205 m) 7.5 + .3 ( 6.6- .2) 3-0 7.4 t .3 3-140 8.0 + .4 3-700 16.8 + .9 3—bottom (1920 m) 21.5 + .5 (26.3 - .*0 4-0 7.3 + .3 4-130 9.1 + .3 4-bottom (915 m) 21.1 t .5 (2.20 ± .*0 3-0 7.8 + .3 5-150 8.7 £ .5 5-bottom (1550 ra) 22.7 + .5 (23.1 - .*0 * Contained visible suspended material. 31 OCTOBER SURFACE CURRENTS —-Cos t o Rico Coastal C u r rent orth Equoforlol C urrent Sovth Equolorial Current Figure II-4. Surface currents in the Eastern Equatorial Pacific. Taken from VJyrtki (1966). of the California Current and the North Equatorial Current in the North Pacific and the Peru Current and the South Equatorial Current in the South Pacific. The Equatorial Counter Current between these two gyrals is well developed as long as the intertropical convergence is sufficiently far north of the equator. During the months from February to April, the intertropical convergence is in its most southerly position and the positions of the major equatorial currents are all shifted to the south as well. This is precisely the time period during which the 3b samples col lected on the Bartlett cruise were taken. On this cruise 7 Be samples were collected by the author for W. Silker of Battelle Labs. These samples were collected "on-line” with my fiber columns using a Battelle Large Volume Water O9O Sampler (Silker ejb al. , 197l)* For this reason, no Th ^ numbers are available from this cruise; the BLVWS employs adsorption on Al^O^ which removes Th, but not Ra, How ever, the important point to emphasize is that results of 7 Silker*s analysis of Be on this cruise indicate that the intertropical convergence is in fact displaced far to the 7 south. The Be northern hemispheric maximum normally found at 8° to l4°N has been displaced to 4°N (Silker, 197^)• The current patterns illustrated in Figure IX-4 were similarly displaced more to the south than indicated here. During the Geosecs cruise samples were collected later in the year (May—June) when the intertropical con vergence is again near 10°N. Xt should also be noted tiiat in the area studied there exists a subsurface counter-current normally situated underneath the northern edge of the South Equatorial Cur rent flowing to the east. This Equatorial Undercurrent first reported by Cromwell et al. (195^)> although sub merged, greatly influences the surface circulation. Trans port calculations (Wyrtki, 1966) indicate that parts of this Equatorial Undercurrent definitely participate in the equatorial upwelling to the west of the Galapagos Islands. The flow to the north in the eastern South Pacific anticyclone off the coast of South America is made up of the Peru Coastal Current and the Peru Oceanic Current. Coastal current extends to only about 10-15°S (Wyrtki, 1966), however, and north of here the wind drift is still to the northwest. Beneath this northward flowing surface layer there exists the southward-flowing Peru Undercurrent, being strongest at about 100 m depth. Radium isotopes in open East Pacific The radium isotopic data for surface samples are presented in Figure II-5* Only those samples collected on station using the fiber columns are shown here. All of 228 these Ra numbers (except for Bartlett Station 2) are O computed by using an assumed value of 65 dpm/lO kg for ^ * Ra' • The 20 X split for Bartlett Station 2 was analyzed 34 30 ' 0 ' 30 ' 223 Figure IX—5. Surface Ra values in the East Pacific for the station locations shown in Figure II-2. The circled numbers are the previous Ra2^^ measurements made in the study area* 35 o 226 and found to contain 66 dpra/lO kg of* Ra , confirming the validity of this assumption. The data do show the expected 228 general trend of a decrease m the Ra content as one moves away from shelf areas. The highest values observed were 10 dpm/lO^kg off Mexico and 19 dpm/lO^kg off Peru. Moving towards the centers of circulation gyres the values droj3 off rapidly. The open ocean values in the northern hemisphere tend to be slightly higher than those in the O South Pacific where values as low as 1 dpm/lO kg were found. The effects of advection on this simple picture are apparent. At Bartlett Station 10 the high value (7 dpm/lO kg) probably represents advection by the South Equatorial Current. The values to the north and south of this station are significantly lower. There is also the O suggestion that the values of 3 dpra/lO kg maintained at Geosecs Stations 33^ and 337 are due to the effect of the shift in the intertropical convergence (and hence South Equatorial Current) occurring between the Bartlett and Geo secs cruises. The effects of equatorial upwelling are clearly 228 illustrated by the Ra concentrations at Bartlett Sta— O tion 11 and 12 of 3 dpm/lO kg. These stations just to the west of the Galapagos Islands have concentrations dis tinctly lower than stations on either side of them. Sub surface fiber samples confirm these I o t v surface values 36 (Moore, personal communication, 1975)* Temperature data Tor these stations indicates the presence of water 8° to 10°C cooler than the surface at a depth of only 20 m. This probably represents the Equatorial Undercurrent Water which participates in the equatorial upwelling in this area. 228 This water would be expected to have a very low Ra con tent . Bartlett Station 2 near the southern tip of Peru O was found to have the unusually low value of 2 dpm/lO^kg. Subsurface fiber samples (Moore, personal communication, 228 1975) showed a constant and relatively low value of Ra down to about 100 m (Table XI-7)• Moore (personal communi cation, 1975) has shown a positive, linear relationship 228 between temperature and Ra in vertical profiles through the mixed layer. Xt would be tempting to say the observa tion corresponds to a recent mixing of the upper 100 m. This is the depth range in which upwelling was found to take place in this area (Smith, 1968). However, the temperature profile shown by the XBT data indicates the presence of a stable thermal gradient (Table XI-7)• Data for fiber samples obtained using both columns 3 and towed samplers within 10 km of the coast of Peru are plotted in Figure II-6. Values range from 19 to 3 dpm/ q 228 10 kg. This observed gradient in Ra should represent 228 the westward movement of Ra by eddy diffusion perpendi cular to the mean advective flow direction of the Peru 37 228 Table XI-7* Vertical Ra and temperature at Bartlett Station 2. Depth Temperature (°C) Ra228/Ra226 Analyst 2 23.65 .037 Knauss 35 23.29 .042 Moore 65 18.69 .038 Moore 95 17.66 .032 Moore 145 14.55 .009 Moore 245 11.53 .003 Moore 38 75 85 90 Ra (dpm/10 kg) 20 85 80 90 228 Surface Ra of Peruvian Coast Figure II-6 values within 1000 km See Figure XX-3 39 Current System. If tlais is true, then in the simplest case 228 the distribution of Ra should be governed by a one layer model which assumes constant horizontal eddy diffusivity. At steady state this model is described by a balance be tween diffusion and decay (Broeclcer, 1965) ! s & - -A<= (1) The solution is given here: C = Cq exp (-y V^/k) (2) where K = apparent horizontal eddy diffusivity C = concentration y = horizontal distance from coast C = C at y = y o 228 ^ = decay constant for Ra" In applying this model, we ignore all vertical transport effects (upwelling) which would tend to diminish the sur— 228 face Ra value by diluting it with deeper water which, as previously explained, is expected to be impoverished in this isotope. The solution to this equation can be used to derive an expression for the apparent coefficient of eddy diffusion: K y s JL In o/c (3) ko This diffusivity can be determined graphically by 228 plotting the log of the concentration of Ra versus dis tance. Xn Figure II-7* this has been done for the data points shown in Figure II-6. The 12 samples taken more than 150 Inn offshore are falling on a line indicating an apparent coefficient of eddy diffusivity (ACED) of about 7 2 10 cm /sec. However, the five samples taken within 150 km of shore fall on a line indicating an ACED of only 10" 5 cm /sec. Quite possibly this represents the effects of two distinct and different flow regimes, and hence two distinct average eddy sizes are in operation. It is well known that the ACED is a function of the scale of dif fusion (Bowden, 19&4)• Using the empirical diffusion diagram of Okubo (l97l) (see Fig. II-8), the two values of ACED computed above correspond to scale lengths on the order of 10 km for the samples 150 Ion or less from shore, and approximately 1000 km for the samples between 150 and 1000 Ion from the coast of Peru. Perhaps the two flow regimes are represented by the Peru Coastal Current (plus any Peru Undercurrent effects) and the Peru Oceanic Cur rent, respectively, for the nearshore and offshore cases. Alternatively, the steeper gradient nearshore could result from coastal upwelling, and hence artificially in crease the gradient by dilution with deeper water. How ever, as noted by Wyrtlci (1966) coastal upwelling off Peru is of a transient and highly variable nature. For the 41 ' I O f s Figure 11 -7* Plot of lor versus distance Trom coast. Tlio apparent coef ficient of horizontal odcly di fusivity associated v;i th the is indicated. The 5.7 hair- o o ° " " Ixfe Tor ha— ° is used for this c a 1 c u 1 a t i o n. . 1*0 H Hj 20 Ky = 4 x 10 cm /sec ro O E Q. "O Ky * 4 * 10 cm /sec C D C M C M o cr 8 0 0 1000 0 200 4 0 0 6 0 0 Km ® RHENO A 1964 3 Z - □ 1962 n O 1962 n <=>1961 I NORTH SEA — *Z A * 3 OFF CAPE KENNEDY ■ * 5 • *6 O NEW YORK BIGHT ® #0 © *b © *c © #d © »e ® # f O J OFF CALIFORNIA I05 © BANANA RIVER ©e I06 i. (cm) Fxgrure IX—S. Plot of apparent coeffic 101113 of eddy diffusion, Ka, versus scale of diffusion,J • Taken from Olcubo (l97l)* observed gradient to be so well correlated with an "artifi cial” ACED is less likely than the scale of diffusion hypothesis presented above. There have been approximately five papers dealing 228 with the Ra content of oj^en ocean surface waters. Two of the papers (Moore, 19^9 and Kaufman et al.. 1973) pres ent data on six samples from the East Equatorial Pacific (bounded by 20°N, 20°S and 135°T ‘ 0 • These six values (Table II -8) compare favorably with values obtained in this study. Another comparison of the present analyses with those of ¥. S. Moore (then with NAVOCEANO) was made through fiber samples listed in Table II-9 (locations are shown in Figs. II-2 and II—3)• These fibers were all col lected on the Bartlett cruise. The agreement is excellent* Th isotopes in open East Pacific This section will be limited to a discussion of the , 228 228 / 22S isotope Th and the activity ratio Th /Ra . Discus- 232 230 sion of the results for Th and Th will be reserved for a later section. As previously explained, the fibers collected on the Bartlett cruise were unusable for Th analysis because of 7 the Be sampling done concurrently; hence, only the fibers collected on the Geosecs cruise are presented (Fig. II—9)* 228 / 228 The activity ratio Th /Ra for the corresponding fiber samples, which are of more interest oceanographically than 228 the Th concentration values, are shown in Figure 11—10. ^5 228 Table II-8. Previous Ra measurements Source Sample No. Location R 227/ 226 Ra u 228 / Ra / 0 (dpm/lO^l) Moore (1969 ¥55-13 17°30fN, 105o26*¥ .27 23 Kaufman et al. (1973) 889 20°16 *N, 106°10»¥ .23 20 890 jo H 0 O H ll4°34»¥ .16 13 842 * CO o - d - 0 00 * = 2 0 CO .03 6 3 84l 13°201S, 101°00»W .032 3 840 2038*S, 119°511W .028 2 o n Q o Notes All of the Ra (dpm/lO kg) values listed here were calculated by using an assumed Ra^26 concen tration of 8.5 dpm/lOOkg. 46 Table IX-9. Fiber intercalibration. Sample Analyzer „ 228/ 226 Ra /Ra D 228 0 (dpm/10 kg) B-24 Moore .034 2.2 B-35 Moore .020 1.3 B-39 Knauss .062 4.0 b-4o Moore .065 4.2 b-4i Knauss .065 4.2 150 120 60 90 30 30 '8.22 79 *j69 35 30 3 0 22 150 120 90 60 228 Figure II-9* Values of Th. activity measured. in the surface waters of the Fast Equatorial Pacific using' Mn-fiber columns on station. hS 90 120 60 150 30 30 J4 * J4 • OB 0 30 30 226 ^^2ZQ a c t i v it y r a t i o 90 60 120 150 P 23 Figure 11-10. Values of activity ratio T 1 l ~ / Ra228 measured in surface v/aters of th.e East Equatorial Pacific. 49 Determination of the mean residence time of a highly reactive substance in surface seawater is in general dif ficult. It requires a knowledge of the concentration of the element (which is very small) as well as its supply or removal rate. Thorium is one reactive element for which this information can be obtained. The mean residence time of any substance whose concentration is at steady state with respect to supply and removal is defined by the ratio of the amount of the substance contained by the water divided by the rate at which it is being removed (or supplied): — r" c ! = T p T (fc) eft For the element Th, two processes of removal have to be considered: radioactive decay and removal onto particles. Neglecting mixing effects (consider the ocean to be laterally well mixed), the fraction of the Th that is re moved onto particles per unit time, A c * can be computed from the following equation: N r N t ) t + N t } c (5) where: Nr = Ra atomic concentration N,p = Th atomic concentration = radioactive decay constant for Ra = radioactive decay constant for Th 50 > = first-order chemical removal rate constant for Th Let the symbol A denote "activity," i.e., A = /^N V * At / K y ) (6) T or: A t / x r (7) c ' - " ' A ' T where: 'Y' = mean time = ]/^ 028 228 Thus the measured Th‘ ~ /Ra ratios can be used to compute the mean time of chemical removal of Th onto particles. The fiber apparently extracts both the particulate and the dissolved Th in seawater. Therefore, the removal time calculated represents the time required for uptake of Th by the particulate matter as well as for removal of this matter from surface layers by particle settling or organism migration. The evidence for the uptake of particulate material by the fiber will be presented in a section to follow dealing with Pu and Ac uptake by the fiber. ppO p 9 O The Th /Ra ratios in Figure 11-10 suggest a fairly distinct latitudinal trend with equatorial values being somewhat low (<C.l); at higher latitudes values over 0.2 are found. This trend may be due to latitudinal 51 productivity variations (Owen et al,, 19J0). The average value for all these samples is about 0.12. Using this value one estimates a chemical removal time of about 0.4 years for Th (or perhaps other elements of similar re activity). As pointed out by Broecker et al.(l973)> this value is similar to the time scale on which plant matter is cycled within the surface layer of the ocean, and strongly suggests that the Th produced within the surface ocean is efficiently removed by each generation of plant matter. The very high seston concentration factor for Th (to be determined in a following chapter) provides addi tional evidence for this hypothesis. 228 Of the few publications dealing with the Th con tent of seawater (6) only one (Broecker et al.. 1973) con tain data from the East Pacific Ocean. Again, they are quite compatible with those presented here. Nearshore Surface East Pacific General oceanography The surface circulation patterns off southern Cali fornia have been the subject of many studies. Figure 11-11 taken from SCCWRP (1973) illustrates the major features of the surface circulation in this area. The California Cur rent is a typical eastern boundary current (analogous to the Peru Oceanic Current). The Davidson Current shown here 52 SAN LUIS OBISPO POINT CONCEPTION SAN BUENAVENTURA SANTA MONICA NEWPORT BEACH SAN DIEGO U.S. ___ ■* MEXICO 3 2 ° ENSENADA — CABO COLNETT ft SAN QUIN TIN 3000 0 0 0 ' 100 KILOMETERS DEPTHS IN METERS Figure 11-11. Surface circulation in the Southern California Bight. From SCCi/RP (1973). 53 is actually a submerged poleward flowing countercurrent (analogous to the Peru Undercurrent). In the winter months north of Pt. Conception this develops into a nearshore, surface countercurrent as illustrated in Figure 11—12* Thus, it is simply the surface manifestation of the deeper countercurrent that develops when the winds weaken seasonally (Reid et al*, 1962). The California Current (see Fig. H - 13) is about 600 km wide and flows southward with its eastern edge relatively near shore following the coastline until it reaches Pt. Conception, At this point the coastline turns abruptly eastward and the flow of water departs from the coast, generally continuing in a southerly direction. South of Pt. Conception and east of the California Current there exists a large counterclockwise eddy, the Southern California Eddy, that is a nearly permanent feature of the flow pattern although seasonal in nature (see Fig. II—12). The eddy is usually well developed in summer and autumn and weak in winter and spring. The exact cause of the eddy is unclear but is certainly related to the en- trainment of water by the California Current. North of Pt. Conception the California Current is unable to entrain water to the east of it because little water is available. However, xvhen it passes south of this point, it encounters a large body of water to the east. According to Emery (i960), this large eddy requires about 10 to 20 days for a 'igrare 11-12, Seasonal variation in surface circulation witliin tlie Soutliern 0 a 1 ifornia Si all t , ^ S A N LUIS CB'SPO POINT CONCEPTION SAN BUENAVENTURA L jju U . SANTA MONICA NEWPORT BEACH SAN DIEGO s - MEXICO SUMMER (JULY) ENSENADA CABO COLNETT N \ SAN L' Q UINTIN 3000 100 KILOMETERS DEPTHS IN METERS 118° 120° 1 22° 124® SAN LUIS OBISPO POINT CONCEPTION 13°C SAN BUENAVENTURA ^ SANTA MONICA NEWPORT BEACH SAN DIEGO US. ___ " MEXICO w in t e r (Ja n u a r y ) ENSENADA - SAN O UINTIN 3000 !000 100 KILOMETERS DEPTHS IN METERS 120° 116° W SAN LUIS OBISPO POINT CONCEPTION SAN BUENAVENTURA ^ SANTA MONICA 16°C NEWPORT BEACH SAN DIEGO Ka_ - —— - - —7? . MEXICO AUTUMN (OCTOBER) ENSENADA - 18°C CABO COLNETT , SAN r O UINTIN 3000 100 KILOMETERS DEPTHS IN METERS 18°< 120° 1 18° SAN LUIS OBISPO POINT CONCEPTION SAN BUENAVENTURA SANTA MONICA NEWPORT BEACH SAN DIEGO U _ JJjv ___ v y ’ MEXICO ENSENADA SPRING (APRIL) CABO COLNETT SAN OUINTIN 3000 100 KILOMETERS DEPTHS IN METERS 120° 121°W 120°W 119°W 118°W 117°W SANTA VNEZ WTNS ^ l o s a n g c l e s 5? if I L S A N TA A N A , f ; NEWPORT I y OCEANSIDE \ xoo ; SAN CLEMENTE I. e n s e n a o a i t o d o s DRAINAGE DIVIDE c a b o c c l n e t t VJ \ KILOMETERS DEPTHS in m e t e r s Fig-ure 11-13* Location of the California Current within the study area. 57 half revolution. The continental shelf area along this coastline is extremely narrow. The shelf (approximately 100 m deep) is on the average only 5 hm wide (varying from as little as less than 1 km to about 15 km). The narrow shelf implies a relatively small surface area of continental shelf sedi— 228 ments to be available as a source of Ra . This fact presents certain advantages and disadvantages. Xt makes 22 8 the detection of Ra" more difficult and also limits the distance from shore able to be studied by providing only a weak source. On the other hand, it also means that a simple first order decay-diffusion model is an acceptable approximation for the system. In other words, we can rationalize a relatively uniform and restricted source 228 region for Ra at the shoreward boundary in the use of a 228 one dimensional model. The flux of Ra^ from bottom sedi ments to the surface ocean will be restricted to this very narrow, immediate nearshore area (Brewer et al.. 1975)* The bottom topography for this region is presented in Figure II-14. It is readily apparent that the classic continental slope-abyssal seafloor sequence outward from the coastline does not exist off California, south of Pt. Conception. Instead, the mainland shore is bordered by the narrow shelf—slope region mentioned above, followed by a wide, complex region of basins and troughs interspersed with ridges locally piercing the surface as offshore is— 58 1 2 0 ° W rr.^*— CONCEPTION -p— SANTA CR UZ • • LOS ANGELES SAN MIGUEL LO N G B E A C H JL f-v r1 n ic o la s i. 3 3 ° N SAN LA JOLLA ? CLEVEN TE I. t h i r t y r / L I rvi^E ' • < / > . dd'3ANKJ >v.7^fsS S o n ? y ^v/O mshty s, 1 K±-’*SAN OlEGO I A.Q.1 E T . I V i W V i « ■ » CORONADOS f»T/V OESCANSO I E N S E N A D A \ DE TODO! 2 SANTOS ' S ^ T A . ^ s a n c a / Figure II-14. Bathymetry of the Southern Cali fornia Borderland. lands (SCC¥RP, 1973)* Finally, the continental slope, the transition Feature between borderland and deep sea Floor, is Found as much as 250 1cm From the coast. North oF Pt, Conception, the distance to true continental slope From the narrow sheIF is much smaller, and the complex basin top ography is absent. A consideration oF this bottom top- 226 ography in relation to the observed coastal Ra values will be made in a later section. Radium isotopes in the nearshore environment 228 The surFace Ra' data For nearshore samples are presented in Figure 11-15. These values were all computed vising the measured Ra^^^ values oF the 20 Jl splits. The samples were taken on a station track north oF Pt, Concep tion where, again, it might be possible to expect to ob serve a diFFusional gradient perpendicular to a mean ad- vective Flow direction (CaliFornia Current and Davidson Current); and another track cutting across the Southern CaliFornia Bight, where hydrographic conditions are very complex involving the gyring oF water. The decrease away From shelF areas is apparent, but not as consistent as oFF Peru. North oF Pt. Conception there is an immediate drop O oFF From a high oF 26 dmp/lO kg to almost uniForm value oF o about 9 dpm/lO kg. This oFFshore area seems to be relative ly well mixed. South oF Pt. Conception there is again a O dropoFF, but the initial value 10 dpm/lO kg is significant- 60 2 2 8 llif^re 1I — 15 • Surf a c e Ra activities measured in coastal waters off southern C alifornia. 61 228 SURFACE Ra (dpm /IO O O kg) PS2 t 9.6 4 P88 9.6 Pt. Dumi 10.8 PD8 7.8 1 2 1 * 120* 119* G\ i'O ly loweir than the area to the north. Moving offshore from Pt. Dume there Is a progressive decrease, hut not a smooth 228 one. Between the two inner stations, the Ra shows a small decrease and between the two outer stations it de- 228 creases as well. But the major drop in Ra is betx^reen the second and third stations. This region betxveen PD 3 and PD b corresponds roughly to the location of the center of the gyre in the Southern California Eddy and thus is subject to less intense horizontal mixing. As will be shown, some of the stations exhibited intense vertical mix ing, and for this reason (as xv^ell as the small number of \ 228 data points) no plot of log Ra versus distance was made to estimate the horizontal ACED. On these two cruises, I used a submersible pumping system to sample discrete depths doxm to about 200 m. 228 Figure II—16 illustrates the decrease of Ra with depth due to vertical mixing processes through the mixed layer and upper thermocline for a station in the middle of the Santa Cruz basin (PD 3). This distribution appears to be quasi-exponential and, assuming it is due to eddy dif fusive processes, permits a calculation of the vertical ACED. The equation used is analogous to the case for horizontal eddy diffusion (3)s depth (m) Ra dpm/IOOO kg Pt Dume 20 60 Q. a) ■° 100 __ 228 Ra dpm/IOOO kg Pt Sal 140 S t. 2 S t . 3 St. 4 228 Figure II-16. Vertical Ra profiles taken with submersible pumping system. where K = apparent coefficient of vertical eddy diffusion Z S and = vertical distance from surface, z Here the assumption is made tliat the surface mixed layer above any given profile provides a point source of 228 Ra which can mix vertically through the seasonal thermo- cline and upper regions of the permanent thermocline (Trier et al., 1972). For this particular station the 2 calculation yields an average vertical ACED of 1,6 cm /sec for the upper 150 m. All stations, however, did not display a simple exponential profile. The three stations off Pt. Sal and one station off Pt. Duiiie (Fig. XI—16) do show a decrease of 228 Ra with depth, but not in a simple logarithmic manner. It is very likely that these stations have been effected to some degree by horizontal advection and thus only order- 228 of—magnitude estimates of K may be made from these Ra z vertical profiles. Station 2 off Pt, Dume (Fig, II-16) had a mid depth maximum, indicating either a source at depth or some transient distribution due to vertical mixing. This station was very near shore and is probably perturbed by the nearshore currents associated with the eastern edge of the Southern California Eddy and northward flowing long shore currents from Santa Monica Bay, 228 For the vertical profiles where Ra' concentrations did decrease with, depth, between any two points, the de crease was assumed to be exponential and an approximate value for the vertical ACED was calculated using the equa tion (8) above* Over the l^O m depth sampled the values were averaged. These average values (Table II-IO) are all 2 around 1-3 cm '/sec which are consistent with previous esti- ^ 2 8 2 * 7 mates using Ra' , H , Be , and other radiotracers (Kauf man et al. , 1973)* It should be mentioned that these two cruises were made during the months of May and August. As observed by Cairns et al. (1970), the vertical thermal gradients off the California coast are at their maximum and present almost continuously from late April through early September. Ilence the vertical profiles sampled the season al thermocline and the upper region of the main thermocline for most of these stations. Only the station most near shore (Pt. Sal, Station l) possessed no thermal structure at all. As mentioned previously for these nearshore samples, 226 U it was necessary to measure the Ra directly on 20 X splits associated with each fiber sample. For samples col lected in the surface open ocean this is unnecessary, but 226 vertical profiles show Ra to vary considerably with q 2 6 depth (Chung e_t al. , 1973). The Ra"" values measured here are presented in Table IX—6. These include samples in the thermocline, at mid-depth, and samples collected 5 ni off the bottom at each station using a bottom-tripped Niskin. 66 Table XX—10. Vertical dififusivities Sample Vertical ACED (cm^/sec) PS 2 2.1 PS 3 CO • C M PS h 2.0 PD 3 1.6 PD h 2.6 67 For* comparison, the open ocean vaJaies of Chung et al, (l97*0 from the Geosecs intercalibration station Gogo-I (29°N, 122°¥) are included. Li (1973) using a simple box model calcula- 226 tion showed that a balance for Ra in the surface box 22 6 (upper 300 m) required an input of Ra' from continental shelf sediments three times the flux from deep-sea sedi ments. Although it was well known that shelf sediments 228 supply Ra to the surface sea, their importance as a 226 source of Ra as well was not appreciated. The open 22 6 ocean Ra value for the surface Pacific Ocean north of the Antarctic Convergence is 6.3 — *5 dpm/lOOkg. However, no nearshore values are available. The surface values presented here (see Fig. II—17) indicate that as much as 226 50 Ion from shore the Ra' is from 40 to 13 percent higher than the open ocean value, indicating the continental shelf sy ^ is in fact supplying Ra to the open ocean surface waters. 228 226 As with Ra , the surface Ra values north of Pt. Concep tion tend to be higher than those over the continental borderland. This is due to the larger shelf areas along the Santa Lucia Banks. The nearshore bottom water values also indicate 22 6 this source of Ra . North of Pt. Conception where larger 226 shelf areas are available as a source of Ra , the values are somewhat higher than those observed at the same depth in the open ocean (Chung et al.. 197*0* For the stations 68 '>26 Figure 11-17. Surface Ra activities measured in coastal waters off southern 0 al i f o rnis.. 69 226 Pt. 8af SURFACE Ra (dpm / 100 kg) PS2 1 0 . 1 P84 10. 0 P83 10.0 9.0 Pt. Dum« PD2 8.9 PD3 P04 PD5 7.8 7.7 7.6 121* 1 1 9 * 1 1 8 *W -a o over the continental borderland at the outer two stations (Pt. Dume Stations h and 5) the values are comparable to those observed in the open ocean, i.e., 6.5 dpm/lOOkg. Pt. Dume Station 3» however, does indicate the shelf supply. This bottom sample was taken from within the Santa Cruz basin and appears to be lower than that expected at a comparable depth in the open ocean. However, it must be remembered that as mentioned by Emery (i960) the basins tend to have properties identical to open ocean water at the sill depth, because the basins are periodically flushed out with water supplied from sill depth (Sholkovitz et al.. 1975) and the properties of the basin water are thus homogeneous. In this sense the bottom water within the Santa Cruz basin contains significantly more Ra^^^ than that contained by the open ocean water at the depth of the sill. Thorium isotopes in the nearshore environment As with the equatorial samples, Th isotopes were 232 230 also measured. The Th and Th results will be dis- 228 cussed in a later section. The Th^ results and the r>pQ 228 activity ratio Th‘ " /Ra are presented in Table IX—5* pop ppO The surface values of Th /Ra for the two nearshore cruises are shorn in Figure 11—18. The stations most close to shore are the lowest. Within each area the offshore values tend to be similar (values north of Pt. Conception are higher than those within the Bight, averaging 0.1^ and 71 co GO C M CM CC G O M LlI O •O OJ C O CVI CO a. w <=>• a. fO co a. Fi/Trc 11-18# Surface values of the activity ratio T 1 ; V°2 ' c '/yar '^ ^ in coastal waters off southern California. 72 0,10, respectively). As was done for the equatorial samples, this activity ratio may be used to calculate a mean time of chemical removal for Th and other highly re active elements. Using the above average values, removal times of* 0.45 y for the area north of Pt, Conception and 0.30 y for stations within the Bight are obtained. This difference is most likely related to productivity dif ferences between the areas. The productivity in turn depends on the availability of N and P being supplied by upwelling. As mentioned, the occurrence of upwelling is expected to be higher south of Pt, Conception because of possible entrainment of Bight water by the California Cur rent (Calcofi, 1973)* 228 / The vertical profiles of the activity ratio Th / 228 Ra for the two cruises are plotted in Figure XI—19* There is a distinct trend of increasing value with depth, 228 The trend is due to both increasing Th and decreasing 098 Ra* '" ' ' , It seems to indicate either a lower scavenging ef ficiency with depth or perhaps reinjection of Th^^ at depth due to particle dissolution in a manner similar to 234 / 238 that proposed for the Th /U disequilibria observed by Bhat et al, (1969)* 232 230 Th and Th ~ in Surface Seawater 2 34 The fiber extraction technique uses natural Th as a tracer for the other naturally-occurring isotopes of 73 Figure 11-19* Values of the activity ratio 'pj1228/Ra228 in vertical profile taken with the submersible pump ing system. Depth (m) 0.15 0 .20 0.25 0.30 0.15 0 .25 0 .3 5 Pt. Dume Pt. Sal • • St. 5 □ St. 4 • St. 4 □ St. 3 a St. 2 throium: Th282, Th2* 30 and Th228. Thus far I have only dis- 203 cussed results of the Th analyses. Before presenting the 21? 230 results for Th and Th , it is important to consider 232 230 the previous published results on Th*' and Th in open ocean surface seawater as shown in Table 11—11. These re sults show a vague trend of decreasing values with time from 1957 to 1969* Part of this decline is due to improved sampling techniques (less contamination) and more specific analytical techniques (with regard to total Th (Th*~ *~) analysis). It was in 19^9 that A. Kaufman (Kaufman, 1969) o n 9 0 9 proposed the very low value o f^r0.017 dpm/lOJ I for content of seawater. Since then, however, recent results have reversed the trend and gone back to higher values (Miyake et al.. 1970; Imai et al.. 1973)* A recent review of the situation (Cherry et al., 197^0 includes a rationali zation for deleting Kaufman1s work: 232 . . . we must comment on the Th data of Kaufman. He was unable to detect Th232 a- £ the level observed by other works, and from a careful analysis of a composite of 2k analyses decided that the Th.232 oceanic level was ^ 0.07 x 10-9g// # He suggested that previous workers might have grossly underestimated their blanks. Kaufman1 s Th.22o data appear to be of excel lent quality and liis suggestion should not be for gotten: the weight of current experimental evidence is. however, against him and for purposes of the fol lowing discussion we shall omit his Th/~32 value. In all cases the previous workers used sample sizes of about 500 i l _ or less. It was hoped that using the fiber extrac — tion technique, the true value of Th (and Th ) in large volumes of seawater might be determined and this 76 Table XI—11. Tli isotopes in open ocean surface waters Tli2 32 (dpm/lO3/) Th23° (dpm/lO-7 ) No. of Samples Reference <4.9 < 27.6 8 Koczy et al. (1957) . 16 .44 2 Moore ejb al. (1964) 1.06 45 2 Kuznetsov et al. (1965) .081 .55 1 Somayajulu et al. (1966) <.017 - 24 Kaufman (1969) .591 1.29 12 Miyake et al. (1970) .294 1.01 5 Xmai et al. (1973) .014 .051 9 Tliis work 77 controversy resolved. Figure 11—20 shows an example of a Th spectrum from the fiber analysis, plotted as raw count— / ing data. The indicated effective volume of 1350 a . was 2 34 estimated from the Th activity found on the Th plate. This effective volume is about an order—of-magnitude larger than any of the previous workers could obtain. 232 2 30 However, we can see that the signals in the Th and Th energy regions are very small. In fact, they are both exactly equal to the fiber blank (see previous section for 228 table of fiber blanks)• By comparison, the blank for Th 230 is exactly equal to the Th blank, so as this spectrum 228 indicates, Th is significantly greater than blank. The Th results from the open—ocean and near-shore surface fibers are presented in Table II-3 and II-5» re spectively. Only those values significantly different from blank are listed and then are not corrected for blank. Table 11-12 presents a breakdown of the data by cruise and location. 232 For the open ocean Th samples 2 out of 9 are un questionably different from blank. For the near shore (within 30 to 100 km) samples 5 out of 21 are significant. The average values for samples different from blank (but with no blank correction) are calculated as 0.13 and 1.07 J 0s/l03l , respectively, for open ocean and near shore samples. These numbers are about an order—of- magnitude lower than the averages listed in the review by 78 C O U N T S P E R CHANNEL 200 P T DUME 3 -1 5 0 At3 2545 min. effective vo lu m e 3 1348 liters 228 Th 150 IOO 224 Ra 5 0 230 Th 232 Th 5.42 4 . 0 ALPHA ENERGY (MeV) Fifnix'e XI —20. Sample Th alpha spectrum plotted as raw cor:ntinr: data. Effective volunie ^ r , calculated From Th - > ' ■ beta activity was 13 ' ! 8 liters. 79 Table 11-12. Summary table of Th2^2 and Th2"^ results Cruise Samples run Samples 3>blank Ave. Th2“ ^2 blank W i o o o I) _ Surface Th2^2 analyses Nearshore PS PD 11 10 3 2 1.07 Open-ocean Geopac -2 2 0.13 Total 30 Surface Th23° 7 analyses Nearshore PS PD 11 10 1 1 0.10 dpm/lO*?/ Open-ocean Geopac 2 0.10 dpm/lO3^ Total 30 5 00 o Cherry _et al. (197^) • 230 The case of Th is very similar. Only 3 out of 9 open ocean samples and 2 out of 21 nearshore samples are significantly greater than blank. The average values for these samples are again about an order of magnitude lower than the averages listed in the review by Cherry et al. ( 197* 0. 2 32 The values observed here for Th do lend support to Kaufman*s drastic downward revision of the thorium con tent of seawater. Xt is encouraging that two quite dif ferent concentration techniques were utilized: direct ferric hydroxide precipitation (Kaufman) and Mn-fiber ex- 228 traction (Knauss), and that our Th results (Broecker et al,, 1973) are also very similar. Xn addition, with the fiber extraction technique there is less concern with spike 234 equilibration since natural Th is used. The discrepancy must be explained either by spike equilibration problems or else previous workers were measuring some contamination or a higher blank than they estimated. The thorium content in seawater must be less than 0.1 >ug/lO^£ and has yet to be measured unequivocally. Other Radioisotopes Extracted by Mn-Fiber There were some interesting and unexpected results from the fiber thorium analysis. Figure 11-21 illustrates a fiber thorium plate alpha spectrum, similar to the one 81 Figure XI—21. Sample Tli alpha spectrum with Pu plotted as raw counting. Alpha energy for peak at 5*18 Mev determined by linear relationship between channel number and alpha energy. Recounting this same plate 2 years later reconfirmed the existence and alpha energy of* this peak. C O U N T S P E R CHANNEL BO GEOPAC 3 2 4 At * 3900 m in . 228 60 40 224 5.18 m e v 20 230 232 4.0 5.42 c o A LPHA EN ER G Y (m ev) shown previously. In addition to the expected peaks due _ 232 _ 230 . 228 . r ) 224 / . _ , _ 228 to: Th , Th and Th , and Ra (a daughter of Tn with a 3*6 d half-life), there is an additional peak present. This peak was conspicuously present in about 10-15 percent of the thorium spectrums. It always occurred 228 just on the low-energy side of Th peak and has an energy of about 5*18 Mev. for the peak channel, corresponding to 229 240 the Pu , alphas which are at 3.16-3*17 Mev. The chemistry used for thorium should be discriminating against Pu fairly well (Kaufman, 1969) and yet some apparently is still following along and being plated ivith Th. This data indicates Pu is being extracted from seawater by the fiber. Recent work by Kong ejt al. (1976) has established that plutonium is in fact extracted well by MnO^. The presence 2 39 240 of Pu , also indicates that the fiber extraction technique is removing particulate material from the sea water. The association of plutonium with the particulate phase is well known (Miyake et al., 1975)• It was found, also unintentionally, that the fiber is very efficiently picking up Ac from seawater as well. Thorium plates which were made from fibers not worked up for 2 months, and which were not immediately counted for 224 228 alphas (>"2 week delay) had Ra*~ activity Th activity. 2 qo o r>Q 228 If only Th ~, Th^ , and Th were originally on the 228 224 plate, then after 2 weeks Th should be equal to Ra . 224 The excess activity in the Ra region must be due to 84 2^3 224 Ra'" , which, is not resolvable from Ra * The source of ppo 007 the Ra on the thorium mount should be the Ac (half- 227 life = 21,8 y) in seawater. If Ac were extracted by the 227 / fiber, an equilibrium amount of its daughter, Th (half- life = 18,7 d), would grow in after about 2 months, Xf then the Th plate were allowed to sit for 2 weeks before 227 counting an appreciable amount of the daughter of Th poo (i,c., Ra with half-life = 11,4 d) would grow in on the thorium plate. This particular time sequence did occur for 224 223 7 fiber samples which were found to have Ra + Ra ^ 223 p p y Th indicating that Ac was extracted by the fiber. To 227 my knowledge no measurements of Ac in seawater have been made. This fiber extraction technique should make such an analysis possible. Summary Acrylic fibers impregnated with MnO^ were used to concentrate Ra and Th isotopes from 1000 liters or more of seawater from the surface layer ( 0 to 200 m) of the East Pacific Ocean, A total of 70 fiber samples were taken by towing a fiber—filled sampler or by pumping water through two fiber-columns in series using surface or submersible 226 234 pumps. Natural Ra and Th were used as tracers for 228 the other Ra and Th isotopes. On all samples Ra was determined indirectly by two methods: low-level beta- 228 counting of Ac ^ and later by an alpha—spectrometric _ . _ 228 analysis of Th . 228 / In the Equatorial Pacific the activity ratio Ra / q q / ry r y Q Ra^ varied from 0*3 to 0*01 with Ra*'- activities of 20 to 1 dpm/1000 kg* The predicted decrease away from con tinental shelf areas off Mexico and Peru was observed. The observed gradient over a distance of several hundred kilo meters across the mean advective flow of the Peru Current System indicated two flow regimes with apparent eddy dif- 5 7 2 / fusivities of 10 and 10 cm /sec and thus characteristic mixing lengths of 10 and 1000 km, respectively. The ef fect of equatorial upwelling west of the Galapagos Islands 228 was evident in the Ra distribution, as was the effect of the (Inter-Tropical Convergence Zone) displacement as sociated with the 1972 El Nino, 22 8 228 The activity ratio Th /Ra of the open ocean exhibited latitudinal trends perhaps correlative with productivity variation and indicated a mean removal time for Th from the surface mixed layer of about 4 months. Off southern California the activity ratio Ra~‘ ~ / 226 . , 228 Ra varied from 0.3 to 0.04 with Ra activities of 26 to 3 dpm/lOOO kg. Coastal water within the Gulf of Cali- 228 fornia contained 66 dpm/lOOO kg of Ra . Vertical pro- 228 files of Ra extending across the surface mixed layer and upper regions of the thermocline were used to calculate p apparent eddy diffusivities of 1 to 3 cm /sec. Surficial 22 6 / , Ra values were significantly higher (from 15 to 40 per— 86 cent) within 100 km of shore than open ocean valu.es* Bottom water samples collected on the shelf* confirmed this and lend support to the importance of continental shelf supply 2 ?6 of Ra to the surface layer of the open ocean, The activity ratio Th*" /Ra< '*' increased with dis tance from shore and with depth* The differj.ng oceanograph ic conditions north and south of Pt* Conception were evident in both the Ra and Th distributions* With the exception of a few samples very nearshore, 212 the Th content of seawater sampled was indistinguishable from blank and is certainly less than 0.1 ^ig/l000 1. The large number of samples analyzed lends considerable support to the heretofore unpopular downward revision of the sea— 2 00 water Th ^ concentration suggested by Kaufman (1969)* The fiber extraction technique is potentially useful for studying other trace radioelements, in the ocean, such as Pu and Ac, which are shown to be removed from seawater by this method. 87 Ra AND Til ISOTOPES AND EXCESS Rn IN THE EQUATORIAL PACIFIC BOTTOM WATERS Introduction ^ _ 226 _ 228 , _ 222 __ . The isotopes Ra , Ra 9 and Rn are all present in the pore waters of deep-sea sediments in concentrations far larger than the overlying bottom waters (Somayajulu et al.9 1973)• The sources of these isotopes are the in soluble Th isotopes Th2"^2 and Th2^0 (Fig. Ill—l) contained in deep-sea sediments. Hence, there is a net flux into the water column, producing near bottom excesses of these isotopes with respect to their parents. Ra22^ (tj = 1620 y) 2 2 30 remains in excess of Th throughout the entire water 228 2 32 column. Ra (t^ = 5*75 y) remains in excess of Th for 2. 222 a distance on the order of 1 km. Rn (tj = 3*8 d) re— 2 22 6 mains in excess of Ra for distances on the order of 100 m. For near bottom waters the decrease of the excesses of 222 228 Rn and Ra can be used to determine vertical mixing rates if their distribution away from the sediment-water interface is governed by vertical eddy diffusion (Broecker, 1963; Moore, 1969)* 228 228 In the water column Ra decays producing Th and 23A 230 similarly U is a continuous source of Th . The in- 88 230 234 > U 226 222 228 228 W ATER SEDIMENT .* 2 3 0 ^ 226 222 Th -> R a —> R n -» 228 228 232 Ra ^ „ 226 _ 228 Figure III-l. Model for supply of Ra , Ra , and Rn^2^ to bottom waters from deep-sea sediments. 2 20 2 24 soluble daughter/soluble parent ratios: Th /U and 228 228 Th /Ra provide informat ion concerning near bottom scavenging of reactive elements# In this study we make measurements of the isotopes: 228 226 _ 222 _ 232 ™230 , _ 228 . _ _ Ra , Ra , Rn , Th , Th , and Th xn near bottom waters (+0 to +700 m) at four stations in the East Pacific (Fig# III-2). These stations lie more or less in an east- west direction, commencing near the crest of the East Pacific Rise (EPR) at 3°N, 101°¥ (water depth = 3200 m) and progressing to the west to 2°N, 122°¥ (water depth = 4600 m). It is hoped that from these measurements near bottom dif fusive and chemical removal processes and source strength of sediments for these isotopes could be studied. Sampling and Analytical Procedures Fiber Extraction The Mn—fiber extraction technique of Moore (1973) as 228 described in the second section was used to measure Ra and the Th isotopes: Th.2" ^ 2, Th2"^, and Th22^. Previous efforts to measure these isotopes in bottom waters (Moore, 1964, 1969; Trier ejb al# , 1972; Kaufman ejb al. , 1973; Kuznetsov ejb al. , 1966; and Imai jet al. , 1973) had relied on large volume sampling (100 to J00 liters) and cumbersome shipboard precipitation. Standard 30 liter Niskin bottles were modified as 90 20 1 0 0 10 110 100 90 80 Figure IXX—2. Station locations for CARSAT II Expedition aboard R.V. T, G. Thomson, 91 shown in Figure III-3 (Moore, 1975)• The nylon mesh bags filled with Mn-fiber were clipped inside the bottles. Xn order to minimize water contact with the Mn-fiber during descent the ends of the "cocked" bottle were plugged with plastic funnels held temporarily in place with surgical rubbing tubing. The tubing was tied securely to the lower funnel, passed through the bottle, and looped through a hole in the upper closure around a soluble alum rod. The solution rate was slowed by coating the rod with acrylic lacquer and then making a hair-line scratch with a razor. Dissolution times at sea (Moore, personal communication, 1975) varied from 1.1 to 2.7 hours. The total elapsed time between the start of a cast and the start of flushing was on the order of one hour (for water depth of about k km)• When the soluble link dissolved at depth the funnels floated away allowing water to flush through the sampler and deposit Ra and Th on the Mn—fiber. This was ac complished by continuously raising and lowering the cast over a 10 m interval for 3 hours. The effective flushing times were thus between 1.3 to 3 hours. The depth of the cast was continuously monitored using a pinger at the bot tom of the cast and the ship1s depth recorder. After flush ing the bottles were closed with their normal endcaps by sending a messenger and the cast was returned to the sur face. The Mn—fiber was spun inside the nylon bag to remove excess water and then placed in a plastic bag for return to 92 Figure III-3. Modified Niskin bottle, A, Mn—fiber bags B, Funnel C, Stretch tubing D, Ring of tubing around soluble link E, Funnel F, Niskin rigging G, Permanent closure (omitted in "descending” figure) H, Messenger 93 DESCENDING ,B / SOAKING ASCENDING t the lab* Each cast consisted of 6 samples. A separate cast was made with normal Niskin bottles 226 to collect 20 liter water samples to be analyzed for Ra 228 as a yield tracer for Ra (preceding section). Cor responding depths were sampled at each station providing a 284 fiber—water pair for each sample. No Th analysis were needed because bottom waters contain a constant ratio of Th2'^/U2'^ (Amin e_t aJL. , 197^0 > and hence a constant Th2"^ content (Appendix III). As noted by Moore (1975) there are several ad vantages and disadvantages to this technique. It concen trates these elements in situ and samples very large volumes of water varying from about 800 to 2000 liters 226 judging from the Ra activity on the fibers. However, there is some question about the actual time of release of the temporary closures. If the expendable closures re lease prematurely, the fiber will initially sample water 228 226 well away from the sea floor producing Ra /Ra ratios on the fiber which are too low (intermediate waters are 228 226 devoid of Ra but contain significant Ra ). This ap parently did happen to most of the samplers on one station (CS II St 5)» rendering the entire cast useless. The first four stations, however, appear to display reasonable 226 activity ratios and Ra contents on the fibers. 95 Chemical Processing of the Fibers Mn-fibers used for sampling bottom waters were pro cessed exactly like the surface fiber samples (previous section). Ra and Th Isotopic Analysis The final Ra and Th fiber solutions for these bottom waters were analyzed exactly like the surface fibers* For surface samples it is possible to use a constant value of 226 Ra (6.5 dpm/lOOkg) for seawater. For the bottom waters, 226 however, Ra can vary somewhat from station to statxon. An associated 20 liter split of seawater was analyzed for every fiber sample. For the Th isotopic analysis it is 23 h possible to assume a constant bottom Th content (2.4 dpm/l; Th2"^/u2"^ = 1.0); this value differs from the one used for surface samples (2.0 dpm/l; Th2"^/U2'^ = 0.8). Blanks The fiber blank values were previously presented in Table II-l. As mentioned, only a few of the 70 surface fiber samples had Ra2^^ and Th.22^ activities as low as the 2 32 largest blank value. The observed surficial Th and 230 Th were almost entirely blank. For the 2k bottom water fibers the Ra"'2^ and Th22^ activities were not as high as surface values, hence the 96 blanks become of* greater importance. Once again tbe re striction of considering only those values larger than the largest blank as being significant has been applied. Generally speaking, the blanks are of greater significance 228 228 (with respect to Ra' and Th ) for these bottom water 2 32 samples than for the surface samples. The Th and 2 30 Th values appear to be generally higher than the surface values, so the blanks are of less significance for these isotopes in bottom water samples than in surface samples, o o n o o / ' Excess Rn and Ra Analysis Water samples were collected in Niskin bottles (6 per 222 profile) for the excess Rn measurements, A pinger was attached to the hydrowire a few meters above the bottom weight and the lowest bottle was attached a measured dis tance above the pinger. Other bottles were then placed at measured intervals up the hydrowire such that the bottom 100 m might be sampled. As the cast descended the direct and reflected signals from the pinger were constantly monitored on the depth recorder. In this manner the pinger could be located a precise distance off the bottom and the distance of each bottle off the bottom was thus known. The bottles were tripped with a messenger and re turned to the surface. The water from each sampler was drawn into an evacuated 20 liter glass bottle (Broecker, 1965) and 97 222 analyzed for total Rn on board ship using the procedure described in Appendix I* Correction was made for the decay of Rn between the time of collection and the time it was analyzed. Afterwards the water samples were transferred into 5 gal. plastic cubitainers and returned to the lab 226 for analysis of Ra via the emanation procedure (Ap pendix X) . Sample Locations The samples for this study were taken on the CARSAT II expedition aboard the R.V. Thompson from the University of Washington. The station locations are shown in Figure III—2. At each of the h stations a total of three casts were made. One cast (6 samples) was for near bottom excess 222 Rn samples. Another cast (6 samples) was for near bottom Ra and Th isotopes using the Mn—fiber extraction 226 technique. The third cast samples) was for Ra analysis of the 20 liter splits at depths not covered by 222 the excess Rn profile. At station 5 (l5°l8*N, 123°23*W) a single cast for near bottom Ra and Th isotopes using the modified bottles was attempted. However, heavy seas resulted in the cast being aborted and returned prematurely. Since it was suspected that the expendable closures had released prematurely, these fibers were analyzed in spite of the 98 above problems to see how the Ra and Th results were ef fected. The fiber Ra solution for all samples in this 226 cast were 3 to 10 times lower in Ra than any of the 228 other stations. No Ra measurements were attempted on the samples. Results Fiber Ra and Th The Ra^^^ and Th^^/Ra^^ results for the 24 fibers are presented in Table III-l. The results of the fiber Th analyses are presented in Table XIX—2. Only values signi ficantly greater than the blank are shown. Because the exact blank correction required is not clear, all values listed are not blank corrected. The precision for the 228 4 * Ra analysis is -11 percent or better based on counting statistics; that of Th ^15 percent or better. 222 226 Excess Rn and Ra 222 The results for the 24 excess Rn and 15 addi- 226 tional Ra samples taken are presented in Tables III—3 and III—4. As explained in the previous section, the 226 Ra data has a precision based on average counting error or deviation in the mean of replicates, whichever is larger. V This stated precision is —4 percent or better in all cases. 99 ? 28 Table III-l. CSII Ra Sample D 228/D 226 Ra /Ra _ 226 Ra , (dpm/10 kg) _ 228 Ra ^ (dpm/10 kg) Th228/Ra228 1- 50 0.003992 333.0 1.329 0.278 -100 0.004372 348.3 1.523 - -175 0.004525 338.2 1.530 0.301 -300 0.003687 336.9 1.242 - -500 0.004031 320.8 1.293 - -700 0.003754 314.2 1.179 - 2- 50 0.003742 384.5 1.439 0.264 -100 0.002895 374.3 1.083 0.397 -175 0.002689 383.6 1.031 - -300 0.007309 366.1 2.675 - -500 0.002100 377.6 0.793 - -700 0.002312 310.8 0.718 - 3- 50 0.003230 398.1 1.286 0.365 -100 0.004048 391.5 1.584 0.284 -230 0.005469 368.9 2.017 0.283 -430 0.003624 348.3 1.262 0.230 -630 325.1 - -830 0.002030 (304.0) 0.617 - 4- 50 0.011299 385.1 4.351 - -100 0.005326 388.6 2.070 0.367 -175 0.004753 388.8 1.847 0.428 -300 0.003859 399.3 1.541 0.448 -500 0.003879 386.9 1.501 0.386 -700 0.005318 393.6 2.093 0.392 100 Table III-2. CSII Th Sample 232 Th (dpm/lOOOkg) Th230 (dpm/lOOOkg) Th228 (dpm/l000kg) 1- 50 dl .52 34) .37 (46) 100 dl .48 44) dl 175 dl .55 37) .46 (43) 300 dl .49 35) dl 500 dl .40 46) dl 700 dl .40 62) dl 2- 50 dl .54 31) .38 (44) 100 dl .56 32) .43 (41) 175 dl .48 37) dl 300 dl .48 44) dl 500 dl .57 34) dl 700 dl .62 37) dl 3- 50 .064 (51) .65 15) .47 (21) 100 .074 (43) .64 15) .45 (21) 230 .162 (49) .77 31) .52 (41) 430 .080 (47) .53 21) .29 (38) 630 .087 (50) .56 23) dl 830 dl .61 34) dl 4- 50 dl d dl 100 .092 (64) .84 21) .76 (23) 125 .095 (49) .89 15) .79 (17) 300 .095 (48) .82 17) .69 (19) 500 .091 (52) .78 18) .58 (24) 700 .114 (49) .89 19) .82 (20) dl = detection limit. ( ) = contribution of blank in percent. 101 Table III-3. CSII Excess Rn222 Total Rn Radium Rn Ex Rn Sample (dpm/l00kg) (dpm/l00kg) (dpm/l00kg) 1-10 97 6 + 8 35 7 X 7 61 8 + £ 1 0 Depth = 3250 m -20 92 6 34 5 + + + + + + + + X 8 58 1 1 0 Location = 3°11.2!N -30 87 i 2 35 7 8 51 9 T X 1 4 101°50.3!W -50 80 8 33 3 1 0 47 0 T + x 1 2 -70 56 9 35 1 6 21 3 1 1 -90 43 4 34 8 8 8. T + X 0 9 2-10 101 9 37 5 6 64 3 1 1 Depth = 3700 m -20 97 7 38 5 6 58 8 T i 9 Location = 2°38.8»N -30 105 i 0 38 1 6 67 0 + x 1 2 5 C s l • o o o -50 70 5 38 4 7 32 2 T x 9 -70 68 3± li 6 39 4 T + + + + + x 7 28 9 T X 1 0 -90 75 7 37 4 6 37 7 T X 1 0 3-10 84 8 37 6 1 0 46 3 T X 1 3 Depth = 3850 m -20 75 6 37 5 9 37 7 T X 1 1 Location * 1°27.3!N -30 58 5 40 3 8 18 4 T i X 9 113°46.6,W -50 47 4 39 8 6 7 5 7 -70 47 6 37 3 T + + X 6 10 6 T + x 9 -90 43 4 39 1 6 4 5 8 4-10 141 i 4 38 6 1 2 103 0 T x 1 8 Depth = 4610 m -20 103 z + 8 38 6 + + + + 6 64 6 T x 1 0 Location = 2°29.3!N -30 101 i 0 38 7 7 62 3 T + x 1 2 121°41.1'W -50 112 i 0 38 5 6 74 0 1 2 -70 111 i 1 38 6 7 72 8 T x 1 3 -90 123 7 38 8 7 84 5 T 1 0 H O ro Table III-4. CSXI Ra226 Ave (dpm/ Sample ___________ _____ ________________ _________ . 1-175 33.8 + .6 -300 33.6 + • 6 -500 32.0 + • 6 -700 31.4 + . 6 2-175 38.3 + .7 -300 36.6 + .8 -500 37.7 + .7 -700 31.0 + .6 3-230 36.8 + .8 -430 34.8 + 1.1 -630 32.5 + .8 4-175 38.8 + 1.5 -300 39.9 + .7 -500 38.6 + .7 -700 39.3 + .8 103 Discussion General Oceanography The line of stations shown in Figure IIX-2 should pass from an area above the Carbonate Compensation Depth (CCD) at station 1 to an area definitely below the lyso- cline and probably close to the CCD. The lysocline is the level above which no dissolution of calcareous tests is observed microscopically and the CCD is the depth at which the flux of calcareous tests is compensated by their re moval. Station b is definitely below the lysocline be cause this surface is shoaling toward the equatorial region as one moves northward in the Central Pacific Basin while the CCD is deepening in this direction, i.e., the trends are in the opposite sense. The CCD is deeper under the Equatorial belt due to the increased surface productivity of foraminifera and coccolithophorids which is controlled by the availability of nutrients resulting from upwelling associated with the Pacific Equatorial Divergence. Edmond (1973) notes that the lysocline also shoals to the east up the flanks of the Southeast Pacific Rise. The sediments at these locations are predominately calcareous oozes. But as one progresses to the west from the EPR one expects carbonate dissolution to increase and hence net sedimentation rates to decrease. This has im portant implications for the fluxes of Ra and Rn into the 104 overlying: water column* As shown by Broecker (1965) the steady -state flux of Ra, Rn through, the sediment-water interface (x = 0) is given by: F = C , ( \D )y sed m where: 2 F = flux (dpm/cm * time) C , = concentration of Ra, Rn in the sediment below sed zone of diffusion, i*e», potential concentra— O tion (dpm/cm ) X = decay constant of Ra, Rn (time D = effective coefficient of molecular diffusion m ( cm" Vt ime ) The total amount of Ra, Rn in the overlying water column supported by this flux is known as the standing crop, M, and this would be (Moore, 1969): Dm/ - , 1 M = C . ( ^)2‘ sed v ' Csed» however, is a function of the integrated production rate and the sedimentation rate: C , = P/S sed ' P = integrated production rate of mobile Ra, Rn 2 in the core (dpm/cm *time) S = sedimentation rate (cm/time) Hence, we have: M = I (0 r r , / x f We should thus expect to see the fluxes of Ra and Rn (and 105 standing crops) to increase in the direction of St 1 to St 4 as we progress east to west from the EPR to the deeper central basin. The same argument would hold for an in- 226 222 creasing source strength for Ra , and hence, Rn . If vertical mixing in the deep ocean is due to eddy diffusive processes, then this standing crop of Ra, Rn will be mixed through some volume of water overlying the sea floor as a function of this vertical eddy diffusion rate. The concentration, C, will vary with the distance above the sediment—water interface according to (Broeclcer, 19^5)5 c = co exp [—x ( V D ) 2 J e where: C = concentration at distance X C = concentration at X = 0 o = apparent coefficient of eddy diffusion (ACED) At each station one could calculate an ACED for the interval o O O p q o 0 to 100 m using excess Rn and 0 to 700 m using Ra • The above simple one-layer model assumes constant with depth and no horizontal advection. As will be seen in the discussion to follow, in many instances we cannot assume the deep sea to be devoid of strong currents (Neumann, 1968). Lonsdale (1976) points out that Pacific Bottom Water (PBW) branches off the circumpolar current southeast of New Zealand, moves north along the Tonga- Kermadec Trench through the Samoan Passage into the central 106 basin of the Pacific, around the southern edge of the Line Islands Ridge, and eastward into the equatorial Pacific to the EPR. This deep zonal flow has been identified via anamolous oceanographic properties as far east as 130°¥ by Wong (1972) who concludes this PBW has split into two major branches: one centered at 5°S but extending from the Samoa— Tokelau Rise, Tuamotu archipelago and Marquis Rise at 10°S to the Line Islands and Galapagos Islands exten sion; and the other centered at 10°N but extending from the Line Islands Rise and Clipperton extension to the Marcus— Necker and Hawaii Rise (Fig. Ill—h). There is a possibility of the southern core of eastward moving PBW having some ef fects along the station line. In addition, Mantyla (1975) notes that the Guatemala Basin seems to receive PBW cross ing the EPR at 7°N through the Sequieros Fracture Zone and near the Galapagos triple junction (2°N). Lonsdale (1976) notes that recent geological surveys of these sites on the Scripp's expedition COCOTQW show little evidence of fast bottom currents, suggesting the inflow is more dispersed. The suggested inflow at 2°N is quite close to the location of St 1 on the EPR. The physical processes responsible for this apparent vertical mixing of Ra, Rn away from the sediment—water interface have been studied by several groups. As pointed out by Garrett jet al. (l972), it is possible to explain certain oceanic mixing as a result of internal wave shear- 107 20*N G om e! Ridge CHILE B A S IN S O U T H E A S T PACIFIC B AS IN 'TRENCH 140° W I20°W IOO°W 0O°W Figure III-4. Location of Pacific Bottom Water flow predicted by anomalous oceanographic data. From Lons dale (1976). 108 ing. This breaking results in a buoyancy flux that can be interpreted, in terms of an eddy diffusivity. Bell (1975) suggested that the interaction of near bottom currents with abyssal hills generates an internal wave field that supports a momentum flux. This wave stress acts on the lowest kilometer of the ocean because of deep critical layers arising from the decay of steady currents with height above the bottom. Wave-induced mixing result ing from this deep critical layer dissipation can provide an apparent vertical mixing coefficient on the order of 10 cm /sec in the lowest kilometer of the ocean. As will be shown this agrees fairly well with the ACED calculated from the observed vertical distributions of Ra, Rn. 222 226 Excess Rn and Ra 222 The plots of excess Rn versus depth for St 1 through St h are shown in Figure III-5. Using the terminology of Chung et al. (1972), only St 3 may be des cribed as quasi-exponential. Using the simple, one- infinite layer model described above one can calculate an ACED for the bottom 100 m at this station to be 26.8 2 , cm /sec. In the literature review to follow, this will be within the range of observations. It also agrees well with the diffusivity predicted above by Bell (l975)* The St 3 profile (l.5°N, 113.8°w) looks very similar to a profile taken by Chung ej; al. (l972) on the SCAN expedition 109 222 Figure III-5# Bottom water excess Rn profiles* Curves are predicted values based on non-linear regression analysis of measured values at each, station* Depth of water column at each station is given in Table III-3# 110 10 20 50 40 SO SO TO EXCESS RADON (dpm/IOOkg) 10 20 90 40 SO SO 70 EXCESS RADON (dpm/IOOkg) BO csn st. 2 70 SO BO 20 10 20 80 4 0 8 0 60 70 EXCESS RADON (dpm/100 kg) • 0 " - csn st.4 70 “ / ** 80 - I w 30 — * * * \ 20 - M \ . to ... ..1 . 1 1 1 1 1 1 1 10 20 30 40 80 80 70 80 90 100 1 1 0 EXCESS RADON (dpm/100 kg) (X-^9 at 1#5°N, 113»9°t ‘ 0* Although taken h years apart, they are both quasi-exponential, This argues for steady- state conditions at this location. However, the other profiles all differ markedly from a simple exponential curve. 222 Xf the vertical excess Rn distribution is con trolled only by vertical mixing with a constant flux across the sediment-water interface and by radioactive decay, then an exponential profile will result when the distribution reaches steady-state. For a non-exponential distribution one might suspect non- steady state, i.e., a transient phenomenon. Examples of such a case were presented by Chung (197^0 for the Geosecs I intercalibra tion station (28.5°N, 121.6°W) and for the Santa Barbara Basin (Chung, 1973)• This explanation, however, is not likely for the profile at St h (2.5°N, 121.6°W) when it is compared with the Rn22^ profile at Geosecs St 337 (4.8°N, 124°W) taken in May 197^« Although taken one year apart, the two fairly close stations look remarkably similar. The 222 Rn m both cases does not undergo a continuous decrease with distance from bottom; superimposed on the rapid de crease is a secondary maximum at about +100 m. These profiles cannot be explained by upward eddy diffusion from a planar source; such a process could not produce a secondary maximum. The St h profile could be explained by horizontal 112 transport of Rn from an adjacent topographic high (Broecker jet al, (l970). Chung e_t al, (1972) define a finite layer model for two finite boundary values in which a fixed concentration of excess Rn, C , is maintained at an ’ u ’ upper level (z = z ) by a steady advective process. As shown previously, the existence of bottom currents in this region is quite probable. The steady state solution for the diffusion in this case is: r C „ s/nh &■ (y^)KJ + 5inhC(zu~z ) ( V & Y 1! sink The measured profile can be fit to this equation by a non linear, least squares iterative program. A diffusivity calculated from this fitted curve can then be referred to as , to indicate a value based on the finite upper boundary model. The profile at St 4 gives = 34.2 cm / sec using this model. The profile at St 2 corresponds to a cascade profile 222 (Chung et al., 1972) with local excess Rn maximums at several depths. This station is located just on the western flank of the EPR and thus the cascade pattern may be a re sult of horizontal advection associated with a rough bot tom topography. The profile at St 1 taken just to the east of the EPR crest exhibits two distinct quasi-exponential distribu tions, i.e., there are two distinct layers with different 113 eddy diffusivities. Such a situation could be caused by a change in the density gradient at +50 m, Below 50 m the density gradient is lower and hence eddy diffusivity is higher than above this point, A two—layer model given by Sarmiento et al, (197^) uses the same equation as the one layer model. The solutions for each layer are given below: upper layer (layer l): -(-z-o)Cx A)k C , = C 0 e lower layer (layer 2): where: 222 C = excess Rn (in dpm/lOOkg) 2 K = vertical eddy diffusivity (cm /sec) 222 1 X = decay constant for Rn (sec” ) 2 = distance above bottom (cm) D = level of interface (cm) 1 & 2 = upper layer and lower layer, respectively. These solutions are based on the following boundary condi tions : 11^ The solutions given above contain 4 unknowns: D, K^, , and Cq# These are determined by a non—linear least squares iterative process. When this procedure is applied to the profile at St 1, the following model diffusivities result: 2 2 upper layer = 13 cm /sec; lower layer = 33 cm /sec. These indicate a more rapid mixing in the lower 50 m than in the interval 50 to 100 m. The model diffusivities calculated from the bottom 100 m at each station are summarized in Table III-5# Also included are the model used in each case: one layer, two layer, or finite boundary. These values compare quite favorably with previous values determined in the Pacific Sarmiento et al. (1976) found values from K = 13 to 121 cm /sec. Broecker et al. (1968) found values ranging from 15 to 30 cm /sec. Chung et al. (1972) found values ranging from K = 19 to 263 cm /sec. It should be mentioned here that the Geosecs program has obtained 119 bottom Rn profiles. Of these only 15 could be modeled in a straight forward manner, i.e., only 12 percent. The concept of a stagnant homogeneous abyssal plain situation where only pure, one-dimensional diffusion occurs is just not realis tic. It appears that the deep sea must be a very dynamic Table III-5. 222 Excess Rn model diffusivities (0-90 m) Station Model Diffusivity (cm2/sec) 1 two—layer K, = 13, K2 = 33 3 one layer K = 26.8 k finite boundary kb = 3^.2 116 environment which has not been studied closely enough yet. 222 Before the Geosecs program most bottom water excess Rn profiles consisted of 3 or 4 points at most within the bottom 100 m, whereas this study was aimed at determining fine structure within this region and 6 points were sampled within this interval. It is conceivable that the fine structure observed recently (departure from simple, one- dimensional diffusion, no advection) was not as apparent previously only because of the lower sample density. There does seem to be a trend of higher 21 Rn in the bottom 100 m as one progresses from east to west with the exception of St 3* As mentioned previously, this probably reflects the effect of decreasing sedimentation rate (Broecker jet aJL. , 1968) . 22 6 At each station the Ra is essentially constant over the 100 m, although there can be significant dif ferences between stations, i.e., St 1 = 35 dpm/lOOkg and St 4 = 39 dpm/lOOkg. These values compare well with those of Chung et al. (197^) at nearby stations. _ 228 Fiber Ra 228 The Ra profiles at each station are shown in 228 Figure IXI-6. As mentioned, these Ra concentrations are low and thus the blank becomes a fairly significant part of the observed activity. Some of the scatter observed in the 228 data may be due to this fact. The Ra flux is low in the 117 Figure III-6. Vertical profiles of bottom water Ra228 taken at each station using modified Niskin bottles. Curves are predicted values based on non linear regression analysis of measured values. Depth of water column at each station is given in Table XXI-3. 118 119 csn _ 228 700 600 500 C O k . 0 ) 1 400 2 300 200 100 L 2 L3 1.4 1 .5 1 .6 1 . 0 U dpm/IOOO kg csn st.3 _ 228 700 600 300 200 100 0.6 0.8 1 .0 1 .2 1 .4 1 .6 1 .8 2.0 dpm/IOOO kg cs n st. 2 228 Ra 700 600 500 400 300 200 100 0.6 0.8 1.0 1 .2 1.4 1 .6 1 .8 2.0 2.2 2.4 2.6 dpm/IOOO kg cs n st. 4 ~ 228 700 600 500 400 300 *H 200 100 1 .5 2.0 25 3.0 3.5 4.0 4.5 dpm/IOOO kg equatorial regions because the terrigenous material con- 2 op 223 taining Th (the source of Ra ) is diluted by high carbonate sedimentation under this highly productive region. Even with these problems some useful information may be gleaned from these profiles. Station 4 shox^rs a quasi exponential distribution (neglecting for the moment the value at +700 m), Using the one layer model the interval +100 to +500 m can be fit with K = 12 cm /sec. The point at +700 m may be real and permanently maintained at this high value, reflecting the effect of horizontal advection on this simple model. At station 3 the distribution resembles the "cusp" distribution described by Chung et al. (1972). There is a 228 pronounced maximum at +200 m and at this point Ra is diffusing both upward and downward in the profile. Since there is no maximum at the bottom, the profile is either transient or there is an advective layer close to the bottom. Xt is most likely a transient phenomenon because 222 the excess Rn in the bottom 100 m at this station ap pears to be quasi—exponential. Above the maximum the profile fits the one layer model with K = 10 cm /sec. The profile at station 2 is a very unusual one. Ignoring the point at +300 m, the interval +100 to +700 m can be fit to a one layer model curve corresponding to / - 2 / K = 7o cm /sec. However, considering this point the profile 120 resembles the "cusp” distribution with a pronounced secondary maximum at +300 m, indicating either a transient profile or the presence of an advective layer at +300 m 228 carrying Ra • There is no reason to suspect this sample of being contaminated, the bottom sampler apparently functioned properly (the fiber Ra solution corresponds to an effective volume of 700 1.) and so the value may be real • The profile at station 1 displays relatively low 223 Ra values and a cascade type of vertical profile. This station is located near the crest of the EPR so that hori zontal advective processes associated with the rough bottom topography may be responsible for this pattern. As noted by Lonsdale (1976) this location may be near a point where PBW is supplied to the Guatemala Basin by spilling over the EPR. If, however, a smooth curve is forced through the points (neglecting the bottom point) a one layer model calculation gives K = 212 cm /sec. The model diffusivities calculated for these 228 stations using the Ra distributions are summarized in Table III—6. These values are only approximations of the rate of vertical mixing. However, it is interesting that there is a trend of increasing vertical mixing rate (coupled with a lower standing crop) as one approaches the EPR. This may result from the interaction of the eastward flowing PBW with the EPR. 121 O p Q Table XII-6* Ra model diffusivities (0-700 m) Diffusivity Station Model (cm^/sec) 1 one layer K = 212 2 one layer K = 76 3 one layer K = 10 h one layer K = 12 122 A review of existing bottom Ra profiles (3) by Kaufman et al, (1973) showed K values ranging from 5 to 21 cm /sec in tbe Atlantic and one profile in the Pacific (8°N, 139°w) based on 2 points had the high value of K = 103 cm /sec. There were several profiles taken on the Geosecs Pacific cruises, but only one (St 331 at 4.5°N, 125.2°w) is in the equatorial Pacific. None of the Geopac profiles, however, have been completely analyzed to date (Feely, personal communication, 197^)• Fiber Th Isotopes The Th^^^, Th^^^, and Th^^ concentrations from the Mn—fiber analyses are listed in Table III-2. As was done with the surface fiber samples, only those values larger than the largest blank are listed. No blank corrections have been made, however. 232 Those Th values that are significantly greater than the blank (10 out of Zb analyses) averaged 0.39 Aig/lOOO 1. This compares with the surface open ocean fibers (2 out of 9 analyses) which averaged 0.13 ug/1000 1. This bottom value is about 3 times the surface value. Per haps the increased value at depth is due to increased suspended material content associated with the deep zonal flow of PBW; a benthic nepheloid layer. A raw data nephelometer trace (Fig. Ill—7) from Geosecs station 33^+ (0°N, 12^.5°W) does in fact show a discontinuity at 3700 m 123 Y « DEPTH X - NEPHELS STATION 334 CAST 1 Down Trace 0.00 0.00 500.00 500.00 1000.00 1000.00 .1 5 0 0 .0 0 1500.00 2000.00 2000.00 2500.00 2500.00 3000.00 3000.00 3500.00 3500.00 4000.00 4000.00 4500.00 4500.00 5000.00 5000.00 0. 12000. 24000. 36000. 48000. 60000. Figure XII—7* Raw data nephelometer trace from Geosecs station 33^ at 0°N, 124*5°W* 124 (about 1 km off* the bottom) with higher values below this point* However, the turbidity within 200 m of the surface is everywhere higher than the turbidity in the bottom 232 kilometer* If the higher Th content of the bottom fibers is due to this benthic nepheloid layer, then the 232 Th content of suspended material in the bottom kilo meter must be significantly greater than that in the upper 232 200 m. Perhaps this difference in Th content of part iculates between surface and deep layers is due to the nature of the particulates* A benthic nepheloid layer would consist primarily of silicate materials while the surface particulates would be largely of biogenic origin. 232 There is no apparent trend of increased Th content with distance from the sediment-water interface, although this fact in itself does not invalidate the particulate argument because the nephelometer trace is constant in the bottom kilometer. Alternatively, the higher value at depth could 232 be due to a higher dissolved Th content in bottom waters * 2 30 Those Th values that are significantly greater than the blank (23 out of 2k) averaged 0.6l dpm/1000 1. tr (1.4 x lCT^g/lOOO 1). This compares with surface open ocean fibers (3 out of 9) which averaged 0.10 dpm/1000 1. The bottom value is about 6 times the surface value. As 232 with the Th above, the explanation for this is not clear. There is no apparent trend with distance above the 123 2 30 sediment-water interface. This Th content results in an average Th^^/U^^ activity ratio of 2.1 x 10 Calculat ing a mean time of chemical removal (as was done for 228 Th in surface waters in the second second of this dis sertation) results in 'X = 23 years. 228 The Th values that are significantly different from blank (13 out of 24) averaged 0.54 dpm/1000 1. The 228 Th activities at St 4 appear to be somewhat higher than ?28 at the other three stations, mirroring the Ra differences 228 between the stations. The deep water values for Th agree very well with some previous work done in the At lantic by Moore (1969) and Feely et: al. (l973)« Of p p Q ppQ interest is the activity ratio Th /Ra (Table XIX-l) which varies from O.23 to 0.45 (averaging 0.34) and re flects that the removal processes in the deep sea are act ing on a longer time scale than in the surface ocean. This p o n 2 ^ 4 is also indicated by the Th /U ratios. Surface waters scavenge Th by direct biological uptake as well as surfi- cial adsorption while in the deep waters only the latter is operative• Summary Modified Niskin bottles containing MnO^-impregnated acrylic fibers were used to concentrate Ra and Th isotopes from large volumes (>1000 l) of seawater from the bottom kilometer of the East Equatorial Pacific Ocean. A total of 126 2k fiber samples at k stations were taken in tbis manner. 226 23k Natural Ra and Th. were used as tracers for the other 222 xsotopes of Ra and Th. Also, at each station excess Rn profiles were taken in the bottom 100 m to study mixing rates close to the sediment—water interface. The station track started near the crest of the East Pacific Rise (EPR) and progressed westward some 2300 km into the Central Pacific Basin covering water depths from 3200 to k&00 m. Thus near bottom diffusive and chemical removal processes and the source strength of these sedi ments for Ra and Rn could be studied. 228 The Ra concentrations varied from k.k to 0.6 dpm/lOOOkg. The profiles displayed to varying degrees the effect of advection on the diffusional gradient and sup port the suggestion of Lonsdale (1976) concerning the existence of eastward flowing Pacific Bottom Water (PBW). To a first approximation the profiles were modeled as one infinite layer and produced apparent diffusivities of 10 to 200 cm /sec. The diffusivity values increase from west to east indicating the effect of the interaction of PBW with the EPR as it spills over the rise crest into the Guatemala 228 Basin. The standing crop of Ra shows the opposite trend (decreasing towards the east) reflecting the dilution effect of carbonate sedimentation on the source strength of the sediment for this isotope. 222 The excess Rn profiles were modeled as either: 127 one infinite layer, two layer, or finite boundary distribu- tions and produced apparent diffusivities of 9 to 34 cm /sec, 222 The trend in standing crop of excess Rn is similar to 223 that of Ra , increasing from the east to the west with increasing water depth* 232 230 The Th and Th content of bottom water, al though quite low, appear to be higher (3 to 6 times, re- 2 30 spectively) than in surface waters. The average Th / 234 U activity ratio of bottom water indicates an approxi— 230 mate mean time of chemical removal for Th of about 20 years. The activity ratio Th^2^/Ra^2^ is also significant ly higher than that in surface waters, indicating that removal processes in the deep sea are acting on a longer time scale. 128 U- AND Th— SERIES RADIONUCLIDES IN SESTON FROM TIIE EASTERN PACIFIC OCEAN Introduction In the recent past there have been many efforts to derive mass balances Tor trace metals in the hydrosphere* This has been accentuated by a growing concern about the effects of man on his environment* However, with improve ments in sampling techniques and chemical analysis, it has become apparent that most of the transition elements and many of the other heavy metals occur at extremely low con centrations in sea water (Benninger, 1976)* In many cases this is due primarily to a rapid loss from the seawater solution after injection* Because the solubility products of the least soluble salts of many trace elements are not exceeded in seawater, the loss from solution requires a particulate carrier of some kind* At the same time it has been known that plankton have relatively high concentra tions of trace elements (Martin ejt al. , 1973)# Thus geo chemists have invariably appealed to plankton as agents of removal for the trace metals. There are some inherent problems, however, in study ing the importance of plankton in marine biogeochemical 129 cycles* Primarily these deal with the collection of suf ficient material for analysis and the inability to intro duce the element of time necessary for studying rates of removal* These problems can be overcome, however, through the use of natural radiotracers contained within the uranium and thorium decay series. They have relatively few contamination problems and have very low detection limits (one literally counts atoms). They should provide a measure of the biogeochemical pathways followed by trace metals * In this work the concentrations of members of the uranium and thorium decay series in marine plankton from the East Pacific (near shore and open ocean) were studied. Simultaneous measurements of uranium, thorium, radium, and lead isotopes in the same plankton sample are preferable to those made on selected individual isotopes contained in plankton (Cherry et al,, 197^0* In addition, measurements of these isotopes in seawater from the same location should be made. It is hoped that from these measurements informa tion regarding surface ocean circulation patterns and sources and sinks for radioisotopes (and trace metals) in the ocean can be gained. Sampling and Analytical Procedures Sampling Techniques Because phytoplankton are the primary producers of 130 organic carbon In the sea, it was decided to select a net mesh size and sampling conditions which should favor their collection. On the other hand, it was desired to minimize contamination of the sample with silt-sized detrital material. Martin and Knauer (1973) used a net size of 7 to collect phytoplankton plus small zooplankton and a 360;u net to collect zooplankton. Since it is well known that nets collect particles with minimum cross-sections less than the nominal mesh size, a y meter diameter, lOQn net was used. In order to maximize the efficiency of phyto plankton collection, all tows were made during daylight hours at a depth interval of 0.5 to 2.0 m. Tows were made using the ship1s hydrowire for 20 to 30 minutes at a speed of approximately 1 to 2 knots and care was taken to keep the net submerged at all times, preventing any floating material from entering the net. After rinsing down the sides of the net with seawater, the plankton sample was transferred from the cod end to a nalgene bottle. The seawater—plankton mixture was immediately frozen and brought back to the lab for analysis. The samples collected consist of a mixture of living and non-living particles which are not easily separable. They are thus mixed samples and perhaps should more properly be referred to as "seston" (Benninger, 197^). This mixture problem is minimal for open ocean samples, but may be signi ficant for coastal samples. 131 Sample Locations The plankton samples for this study were obtained on several different cruises. On cruise $137301 of USNS BARTLETT 4 samples were taken from March—April 1973# A total of 13 plankton samples were taken on 4 cruises of the R/V VELERO IV: cruise #1213 (October 1972), cruise #1230 (February 1973)* cruise #1282 (May 1974), and cruise #1291 (August 1974). The sample numbers and corresponding loca tions are listed in Table IV—1 and Figures IV-1 and TV-2 show their locations geographically. Chemical Techniques Spikes Previous efforts to measure a suite of radionuclides in plankton samples (Kharker et al,, 1973) made use of _ 208 . _ 230 232 228 _ , . . . , Po and Th or U —Th spike solutions in order to study the isotopes of the elements uranium, polonium, and thorium. However, this choice of tracers allowed only part 230 of the isotopes to be studied. When Th spike was used only ratios of the uranium isotopes (u234/xj238) obtain ed. When U^'^-Th^*'^ spike was used, the plankton Th^^ 208 content was unobtainable. Also, the use of Po requires significant tail corrections to be made because the alpha- spectrometer used does not have sufficient resolution to 208 completely separate the alphas of Po (3#H Mev) and that 132 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 IV-1* Plankton locations. Sample Latitude 18027H ri O 0 H ri N 18031H 30°35 N Bart P5 10°00 N Bart P3 18°57 S 18439 33°25 N I8O3OH 30°18 N 18028H 29°52 N Bart P4 17°o4 S 18435 O 33 13 N 18449 34°l6 N Bart P6 09°57 N PS 1 34°59 N PS 2 34°51 N PS 3-1 34°43 N PS 3-2 34°43 N PS 4 34°34 N PD 2 34°00 N PD 3 33°44 N Catalina 33°25 N (OPf* Long Pt.) 150 120 90 60 8* 4* 30 PLANKTON LOCATIONS 150 120 9 0 6 0 Figure IV—1. Plankton sampling locations in the East Equatorial Pacific. 13k co CD O n • _l Ql CM O CM tO Figure IV—2. Plankton sampling locations off southern California. 135 ? 1 o of Po (5.305 Mev). In order to avoid all of the above problems, it was 2 32 228 decided to use the following spike solutions: U (Th - free), Po2°9 and Th23**. The alpha-spectrometer used for this study could completely resolve Po2* " * 3 (^energy = 4.88 Mev) and Po2^^ (see Appendix I). The spike solution has an activity of 3.75 dpm/ml. 2 32 228 The U (Th -free) spike solution was prepared from a XJ232-Th22^ spike solution. The XJ232-Th*'2^ solution was taken up in 8 NHC1 and passed through an anion column (Dowex AG 1 x 8). After washing the column with several column volumes of 8 NHC1, U was eluted from the anion column with 0.1 NHC1. This solution was dried, taken up in 4 NIIC1, and then passed through a cation column (Dowex 228 50W x 16) to further purify against Th . The entire process was then repeated: anion exchange followed by 228 2 32 cation exchange to ensure elimination of Th from XJ . 232 The purified XJ spike was made to a 10 NHNO^ solution. 2 32 2 2 8 It was calibrated against the old U /Th spike solution 232 by comparing the XJ activity of each solution against 238 232 a common U solution. From the two runs of XJnew versus tt238 , 2 32 TT238 XJ and XJoXd versus XJ we have: 136 u238 = U n e v x U o l d (d p m /tn l) = U2 3 2 (d p m /m l) UplI Voll U238 Five such, calibrations (Table IV-2) give an activity of 8.12 i .19 dpm/ml for the U2^2 (Th22^ free) solution. This spike solution was dispensed with a precision pipet (Medical Laboratory Automation) at 200 X capacity. 234 The Th spike solution was prepared by isolating 234 Th from a uranyl nitrate solution using the Th purifica tion procedure described previously (second section) and the anion exchange procedure described above to further purify against uranium. The anion exchange and cation exchange steps were repeated. The final dry beaker was 2 34 taken up in 10 NHNO^ acid and this became the Th spike solution. 2 34 Calibration of the Th spike solution was done by carefully evaporating a known weight of the spike solution on a platinum plate. The platinum plate was then counted for alpha and beta radiation in a gas flow proportional counter of known efficiency (see Appendix X). It gives no p oQ alpha activity, indicating that all uranium (U and 234\ U ) had been removed from spike solution. Two such spike calibration plates were made and their activities monitored during the period of plankton analysis showed Table IV-2. U2^2 (Th228-:free) spike calibra tion. Run Number U232 (Th228-free) in dpm/ml 1 8.20 2 00 • O VO 3 8.4l 4 7.65 5 8.24 x = 8.12 <5 = - .19 (approximately 2.4 percent) decreases with, time with a half—life of 24 days. Thus, 23k the spike solution contained pure Th and its short- 234 lived daughter Pa • Chemical procedure The procedure described below to sequentially isolate pure fractions of polonium, uranium, thorium, radium, and lead is a modification of a method originally developed by Kharltar, er f c al. (1973)* The thawed plankton samples were washed five times with deionized water; each time by shaking the plankton plus deionized water mixture in a 150 ml centrifuge bottle and then centrifuging at 14,000 rpm. The salt-free plank ton was then dried (llO°C) and homogenized in an agate mortar. An aliquot of 100—200 mg was taken for major and trace element analysis. For radiochemical analyses, about 0.3 to 4.0 g was used. The sample was placed in a Teflon beaker, wetted with deionized water, and then HNO^ was added slowly. Spike solutions of Po20^, U2^2 (Th22^-free), and Th.2^ as well as a stable lead carrier (where needed) were added. The sample was treated with aqua regia and taken to dry ness. The dry residue was heated to dryness again with HC1, and leached with 1 NHC1. The residue (silicates) was treated with HF—HCIO^, taken to dryness, treated with HC1 to convert to chloride form and again heated to dryness* 139 This residue was dissolved in 1 NHC1 and combined with the first leach. About 100 mg of ascorbic acid was added to the + 3 clear solution to complex any Fe present. The solution was heated to 80°C and a l/2” diameter Ag disc was immersed in the solution. The beaker was covered with a watch glass and gently stirred for 2—3 hours while the polonium plated out spontaneously on the disc. The disc was washed with deionized water, dried with acetone, and was ready for isotopic analysis (see Appendix i). After the above Po separation, the plankton solu tion was taken to dryness with HNO^-H.CIO^ to destroy the ascorbic acid. The dry residue was taken up in dilute warm HC1. On some occasions, a white-fluffy precipitate developed. In these cases the solid was removed by centri fugation, treated by fuming with HNO^-HCl and HF-HC10^, and taken up in dilute, warm HC1. This dilute HC1 solution was then combined with the original solution. +3 After addition of Fe , U, Th, and Pb were copre cipitated with ferric hydroxide whereas ~ 9h percent of the Ra remained in solution (Applequist, 1975)• The precipi tate was removed by centrifugation and washed with de ionized water. This wash, combined with the original supernatant, was acidified with IICl to a pH<l, and its volume adjusted to about 200 ml with deionized water to prevent NH^Cl precipitation. It was ready at this point 140 226 228 for the Ra analyses. If desired, the Ra content of 228 this solution can be determined indirectly via a Th content in a manner exactly analagous to that used for Mn fiber analysis. However, as will be shown later, by know- 226 ing the Ra activity ratio of the water from which the 228 plankton was taken, it is possible to estimate the Ra content of the plankton. The ferric hydroxide precipitate was made up to an 8 NHC1 solution and the procedure of Ku (1966) was used to isolate pure fractions of U and Th utilizing ion exchange and solvent extraction techniques. Thin sources of U and Th were made by mounting dropwise a TTA-benzene extraction on a stainless steel planchet. These sources were then subjected to alpha-spectrometrie and beta-analysis (see Appendix I). Isotopic Analysis Po (Pb) isotopic analysis In Table IV-3 are listed samples and their respec tive time elapsed between collection and analysis of Po 210 expressed as Po half-lives (138.4c|). For samples Pl-1 to PI—11 and PI—19> the elapsed time varies from 6.7 to 210 9.9 half-lives. For these samples the analysis for Po 210 yields the Pb content of the plankton. For samples PI—12 to Pl-18, the elapsed time is equivalent to 2.8 to 210 210 3.5 half-lives of Po and thus some of the Po is 1^1 19 1 6 2 7 9 5 10 4 8 3 11 12 13 14 13 16 17 18 Plankton decay times for Po. Sample Po210 half-lives) Catalina Island 9.94 18027I - I 7.64 18030H 7.64 I8O3IH 7.64 18028H 7.64 18435 6.81 18439 6.81 18449 6 • 81 Bart P—3 6.65 Bart P—4 6.65 Bart P-5 6.65 Bart P—6 6.65 PS-1 3.53 PS-2 3.53 PS-3-1 3.53 PS-3-2 3.53 - 3 - 1 CO d. 3.53 PD—2 2.79 PD-3 2.79 142 210 probably unsupported by Pb (Kharkar et al. . 1973)* For these samples Pb carrier was added to the plankton solution 210 for later direct analysis of Pb*" . Xt should be noted, however, that since they are all nearshore samples whose 210 210 original Po /Pb activity ratios have been found to be 210 low (Turekian et al,, 1973)» even the uncorrected Po' 210 activity of these samples could be ^2 times the Pb activity. The Ag disc was counted on an alpha spectrometer 209 210 for determination of the activities of Po and Po . By knowing the Po^^/Po^^^ activity ratio and the initial Pq209 Sp-jj£e activity, one can determine the Po~^^ activity 210 (and hence Pb activity) of the original plankton sample. A sample spectrum and calculations are presented in the Appendix I• U isotopic analysis The U mount was counted on an alpha spectrometer for 238 23b the determination of the activities of U , U , and u232. By knowing the U238/U232, U234/U232 activities and 2 32 228 the initial U (Th ~ free) spike activity, one can p o Q p p | ( calculate the U and U activity of the original plankton sample (Appendix X). Th isotopic analysis 23b For the thorium analysis the isotope Th was used as a yield tracer. In a manner exactly analogous to the 1^3 Mn fiber chemistry, this was counted indirectly on a gas flow proportional counter (Appendix X) by counting 212 2 20 The Th mount was then analyzed for Th , Th and q 2 0 Th'' in an alpha spectrometer. By knowing the relative counting efficienties ( — Appendix l) the resulting ratios Th232/Th23\ Th23°/Th23\ Th228/ ^ 234 and the 2 24 original activity of the Th spike, it is possible to a. . . . • a. ^ ^,232 ™ 230 _ ^ 2 2 8 calculate the original activity of Th , Th , and Th in the plankton. As with the case for Po210, Th^2^ in the plankton 228 was not immediately isolated from its parent, Ha , and hence the activity determined is not exactly the original 228 Th activity of the plankton. However, the effect is not 2<?8 nearly as great, because the Th half-life is 3 times 210 that of P o . Ra isotopic analysis 226 The radium solution was analyzed for its Ra content (Appendix l) indirectly via the emanation method (Broecker, 1965)* Prior to isolating this radium fraction from the original plankton solution, only the Po self- deposition had been done. The subsequent ferric hydroxide precipitation should, on the average, result in approxi mately a 6 percent loss of Ra. Thus, the Ra‘ ~ ^ determina tion when corrected for this loss should be the initial 226 Ra content of the plankton. 144 Knowing the content of seawater from 226 which the plankton was taken and the Ra content of the 228 plankton, one can predict the Ra content of the plankton assuming no fractionation of the two isotopes occurs be tween seawater and plankton. The possibility of fractiona tion is very small because they are both heavy isotopes of an element that is quite soluble at the concentrations found in seawater. Blank A blank was run by starting with about 50 ml of de ionized water and carrying it through the entire chemical procedure from acid dissolution to final plating and 209 232 isotopic analysis. Spike solutions of: Po , U / 222 x 234 (Th free), and Th were used just as for samples. The results are shown in Table IV-4. Results The results of isotopic analyses for uranium, thorium, lead, and radium are presented in Tables IV-5 through IV-8, respectively. They are all presented on a dry weight basis. Since the samples are mixed plankton as well as plant debris, etc., an appropriate conversion factor (to a wet basis) might be 25 or perhaps as high as 50 (Cherry e_t al. . 1974). In the di scussion to follow, the dry weight figures will be used. In order to be directly 145 Table IV-4. Plankton blanks. U238(ug) .026 + .008 U23Z|'(dpm) .018 + .006 u23^/u238 .683 + .315 Po2^(dpm) .056 + .005 Ra.226 ^pujj .080 + .008 Th232(dpm) .014 + .003 Th23°(dpm) .139 + .014 TbL238(dpm) .251 + .020 Table IV-5. Plankton U Sample_______U234/U238_________U238 (ug/g) ,026________U234 (dpm/g).018 1 1*42 + .18 4.15 + .54 01) 4.39 4 - .52 .9) 2 1*26 4 - .07 1.79 + .11 01) 1.68 + .09 .9) 4 1.16 4 - .09 1.35 4 - .09 02) 1.12 4 - .08 02) 5 1.11 + .05 7.70 + .42 ♦ 7) 6.36 + .34 .6) 7 1.41 + .04 8.32 4 .37 .2) 8.72 + .37 • 1) 9 1.26 + .05 5.99 + .31 *5) 5.60 4 - .28 .3) 10 1.08 + .04 4.15 + .20 .5) 3.32 + .21 .4) 13 1.17 + .11 1.73 + .27 .3) 1.50 + .23 .2) 14 1.22 + .05 0.81 + .04 .7) 0.73 + .04 .5) 15 1.17 + .04 1.09 + .05 • 4) 0.94 + .04 • 3) 16 1.16 + .04 1.15 + .05 .4) 0.98 + .04 .4) 17 1.34 + .07 3.62 + .21 • 8) 3.61 + .20 .5) 18 1.28 4 - .04 3.72 + .14 .9) 3.54 + .13 .7) 19 0.82 4 - .04 2.48 + .13 .8) 1.50 + .08 .9) Ave 1.20 3.38 3.08 ( ) = Contribution of blank in percent* 1^7 Table IV-6. Plankton Th. 230, 228 - Sample Th232 (dpm/g) .014 Th230 (dpm/g) .139 Th228 (dpm/g) .251 232 Th Th?«9/ Tti 1 .309 t .136 t .026 - 023 (9) .592 | .493 - .125 | .371 - .454 J .646 - .267 1 .979 - 061 (50 dl , 1.92 dl 2 021 (9) 035 (24) .770 I 044 (26) 3.62 5.65 3 004 (12) 010 (24) .149 t .481 I 010 (37) 4.88 5.83 4 I1 .156 | .313 1 .079 1 .148 J 301 (51) 042 (62) dl dl 5 041 (20) 054 (64) .974 J 1.103 T .818 - 075 (56) 2.92 6.26 6 065 (03) 079 (16) 101 (17) 2.06 3.52 7 021 (14) 030 (41) 050 (24) 3.38 10.35 8 035 (20) 085 (30) I1 .647 | 6.59 dl 9 .124 | 018 (13) I1 .402 | .247 1 .729 I .108 | .no I .262 I 1.052 t .549 J .501 - 041 (44) dl 5.21 10 .209 1 020 (06) 028 (31) .596 7 .722 I 1.255 1 034 (38) 1.92 2.85 12 .133 I .884 J .058 | 010 (03 014 (17) 025 (10) 1.85 5.41 13 020 (.3) 018 (04) 026 (04) .82 1.42 14 005 (05) 008 (30) .478 1 017 (12) 1.84 8.18 15 .048 J .046 t .759 | .288 J .592 - 003 (05) 005 (25) .478 I 012 (10) 2.26 9.85 16 004 (06) 010 (11) .634 - 015 (08) 5.69 13.35 17 054 (02) 069 (15) 1.082 i 1.059 I 080 (26) 1.39 1.42 18 026 (06) 037 (35) 053 (31) 1.90 3.67 19 073 (02) .067 (22) 1.555 - 109 (13) .85 2.63 Ave. .253 .461 .799 1.82 3.16 , 1 - . ui d 239, 240 x = detectable Pu h i ) * contribution of blank in percent. oJdl = detection limit. Table IV-7. Plankton Pb# Sample Pb21° ( dpm/ _ 210/ 226 Pb /Ra 1 16.26 + .80 .7) 7 2 11.13 + .36 .*0 18 3 5.08 + .22 .2) - 4 6.09 + .31 01) 10 5 22.47 + .54 .5) 6 6 0 • H H + .50 .3) 8 7 13.58 + .36 .3) - 8 3.15 + • 16 03) 2 9 12.07 + .48 .5) - 10 9.71 + .26 .5) 4 11 H 00 H + .49 .9) 13 12 S.77 + .23 .1) 14 13 6.66 + .19 .1) 15 14 6.18 + .19 .2) 16 15 7.37 + .28 .1) 35 16 6.01 + .15 .2) 20 17 17.85 ±i .03 .3) 69 18 11.86 + .35 .6) 23 19 14.28 + .61 .3) 6 Ave . 10.96 17 - = Ra was detection limit. ( ) = contribution of blank in percent. 149 Table IV-8# Plankton Ra. Sample Ra226 ( dpm/g) .080 Predicted Ha228 _(dpm/g)* 1 2.42 + .11 (07) .29 2 0.627 + .034 (11) .08 3 .059 + .005 (30) .007 4 .676 + .053 (1 6 ) .04 6 1.57 + • 06 (03) .19 8 1.28 + .07 (1 2 ) .13 IO .268 + .020 (24) .03 11 1.36 + .16 (18) .12 12 .613 + .024 (03) .11 13 .442 + .026 (04) .04 14 .381 + .019 ( o4) .03 15 .210 + .010 (07) .02 16 .295 + .013 (05) .03 17 .258 + .043 (31) .03 18 .507 + .032 (2 2 ) .06 19 2.53 + .07 (02) .30 Ave. .849 .10 * = see text# ( ) = contribution of blank in percent. 150 comparable to the few previous measurements, the numbers are not blank; corrected, however, the contribution of the blank (in percent) has been indicated in all cases. It can be seen that the blank is unimportant in terms of U and Pb content, but it is of some significance in the Ra and Th analyses. For these later elements the values less than blank are listed as "dl,” i.e., at or below detection limit. Although the Ra content of the plankton is quite 228 low, an approximation of the Ra content is attempted as explained previously and these values are included in Table IV-8. Discussion U Isotopes The U data shown in Table TV-5 provide indirect evidence that the lOQu net is in fact sampling mainly p o Q phytoplankton. Berminger (1976) found the U content of seston collected with a net (phytoplankton fraction) to be significantly higher than that collected with a 333A1 net (zooplankton fraction). The 35a1 seston contained from 2 0.6 to 1.8 dpm/g U compared to our values of 0.6 to 6.0 p p D dpm/g U . This is somewhat higher than the values reported by Kharkar et al. (l973) but they were using a coarse net and sampling zooplankton. The scatter is pre— 151 sumably due to compositional variations in the mixed sample. Very few papers dealing with the U content of plank ton are published, but the ones available do agree well. Miyake el; ad. (1970) found an average of 0.6 dpm/g U238 in plankton from the western North Pacific. Kharkar et ad. (1973) found .3 to .6 dpm/g XJ238 in plankton from the Caribbean. Benninger (1976) found from .6 to 1.8 dpm/g O O 0 U in plankton from the Long Island Sound. With the exception of a few unusual values, the o o h p o Q majority of the U /U ratios cluster around the oceanic value of 1.14 (Thurber, 1962; Koide et ad., 1964; Ku et al., 1976). This fact argues against a large detrital com ponent being present in the plankton sample. Detrital 90/1 p oQ material would be expected to have a U*~ /U activity ratio of unity. As noted by Miyake et al. (1970), there is no fractionation of the uranium isotopes by biologic up take from seawater. 238 Table IV—9 lists the concentration factors for U C.F. = ^S//g TJ238 in dry plankton Mg/ u238 in seawater S for the plankton samples analyzed. The values range from 2 3 10 - 10 , and when compared with the concentration factors of the other radioisotopes studied, are relatively low. If one wishes to consider the effect of plankton on 152 Table IV-9# Planlcton concentration factors Tor U238# Sample C. F. (x 103) 18027H 1.28 18031H 0.55 Bart P3 0.42 18439 2.39 18030H 0.82 18028H 2.58 18435 1.85 18449 1.28 PS 2 0.53 PS 3-1 0.25 PS 3-2 0.33 PS 4 0.35 PD 2 1.12 PD 3 1.15 Catalina O.76 Notes U238 in "dry" seston = C. F. Ms/ U238 in seawater & the uranium content of the ocean, the following simple cal culation can be made: According to Strickland (1965) "the average areal primary standing crop in the photic zone is -4 / ^ 2 x 10 g Carbon/cm . Assuming plankton collected is 30 percent carbon, this corresponds to a standing biomass of 6.7 x 10”^ gplankton/cm^ * This means 2.3 x 10 ^ g 003 2 U /cm is associated with the plankton population* How— q o Q ever, the average U concentration of seawater is 3*3 x -9 / 10 g/g» hence a 100 m thick photic zone would contain 3*3 x 10**^ g/cm" of U2"^. The U contained in plankton represents only a small fraction (.007 percent) of the p o Q total U in the photic zone. The change in uranium content in seawater due to the difference of uptake rate by living organisms from year to year or place to place is very small* Several papers by Miyake ert al. (1964, 1966) purported to show variations (up to 25 percent) in uranium content of seawater and alluded to the possible effects of biologic activity, however, from the argument above this is not possible. The mean time of biological removal of U from the photic layer can be calculated by assuming removal per year equals production. According to Owen et_ al. (1970) , the average annual production in the Eastern Pacific is — o 2 ^9 7*5 x 10 g Carbon/cm /y, or about 2.5 x 10”" " g plankton/ cm~/y• Using the content of plankton above, this —S 238 2 means 8.4 x 10~ g U /cm /y could be removed by plankton. 15^ Thus: X c N = 8.4 x 10“8 g U2^8/cm2/y where Ac = probability per unit time of a U atom being biologically removed, and N = concentration of U in the q photic layer. This yields Ac = 2.5 x 10 y or r = mean time of biologic removal = A “1 = 400 years. This number is a lower limit because the entire yearly production of organic material was assumed to be lost and not just a portion of it. The system can be envisioned as a "flow- through" model where supply (productivity) and removal of organic material are balanced, i.e., a steady state exists. Ra Isotopes Similar to the case for U, the Ra results listed in Table IV—8 indicate that a mixed plankton sample of pre dominantly phytoplankton was collected. The activity varies by a factor of ten and does show possible regional trends. The values along the coast of California are all fairly low. These samples were the ones dominated by plant fiber and algae. The samples collected off Mexico are highest, with the equatorial samples being intermediate. However, there is considerable scatter in the data and these trends are probably due entirely to the compositional variations in the samples and not water mass Ra22^ dif ferences. 155 A recent review paper (Cherry el; al, , 197^) has shown the marked difference between the Ra content of phytoplankton and zooplankton. My values seem to be inter mediate between the two. Shannon et al, (l97l) note that —12 in normal oceanic regions phytoplankton have 1-2 x 10 g pn/T - I p Ra /g whi1e zooplankton have 0.3 x 10*" g/g* Szabo 226 ~ J (1967) found very low Ra contents of 8 x 10" g/g in his mixed plankton (but obviously zooplankton rich) sample col lected with a 200 m net. Shannon et al. (l97l) also noted that phytoplankton collected in the Agulhas Current, where 226 upwelling dominates, had extremely high Ra contents reaching 38 x 10~1^g/g and averaging 8 x lO^^gRa^^/g. 226 Algae are also known to concentrate Ra from seawater as shown by Edgington et al. (l970)# The mechanism of en richment is not known, but Szabo1s data (1967) indicate it may be a process of ion exchange, at least for the alkaline +2 +2 earths. The Ra ion can also substitute for Ca in the aragonite struction. Alternatively, there may be a cor relation between Ra and Si profiles in seawater and Ku et al. (1970) and Edmond (1970) have supported this view. 226 The concentration factors for Ra listed in Table Ll Ll XV-10 range from 0.2 to 3*7 x 10 and average 1.1 x 10 for 226 Ra • This compares fairly well with the concentration factor for the Caribbean plankton calculated by Kharkar o et al. (1973), i.e. , 3 x 10 , although they sampled pre dominantly zooplankton. The Ra concentration factor is 136 Table IV-10. Plankton .concentration factors for Ra226. Concentration Sample Factor (x 18027H 3.72 18031H 0.96 Bart P3 1.04 18439 0.22 I8O3OH 2.41 Bart P4 1.97 18449 0.41 Bart P6 2.09 PS1 O.63 PS2 0.45 PS3-1 0.39 PS3-2 0.22 PS4 0.34 PD2 0.30 PD3 0.68 Catalina 2.63 Note: dpm/g Ra in dry seston. C. F. dpm/gRa^Sb in seawater about 10 times higher than that for uranium, indicating a biologic preference for Ra over U (accidental or other wise) . Assuming the standing crop is 6.7 x 10 ^g plankton/ 2 226 cm and that the Ra content of plankton is 1.8 x 10*”^^g/g, then 1.2 x 10 "^gRa^^/cm^ is contained in the plankton. Assuming the 100 m photic zone has an average —17 226 concentration of 3.1 x 10 gRa /g, this amounts to 3*1 x 10”^^gRa^^^/cm^. Thus, the amount of Ra^^^ as sociated with plankton represents 0.4 percent of the total Ra226 jLn the photic zone. One can calculate the mean time of biologic removal 226 for Ra from the photic layer by assuming the entire production per year is removed: ^cN = 4.5 x 10-^^gRa^2^/cm2/y where N = 3»1 x 10 ^gRa^^/cm. Thus, Ac = 1.3 x 10 ^y ^ and = 6.6 y. This means on the average an atom of 226 Ra should be expected to reside 6 years in the photic zone before being removed by plankton. This number is a lower limit because Ra^^^ content of plankton has been maximized (no blank correction) and total annual production has been removed. This residence time is significantly shorter than that calculated by Szabo (1967) who estimated 'Y = 930 y. This discrepancy is due to the fact that 226 Szabo*s Ra plankton values were much lower, probably be cause he sampled predominantly zooplankton with his 200 ja 138 net and zooplankton have much lower Ra contents than phytoplankton. Since carbon production rates refer to plant(primary) productivity, phytoplankton and not zoo plankton data should be used. Xn addition, Szabo assumed only 10 percent of the annual production was lost to deep water layers (his estimate of the standing crop). The value of 'Y = 6 y calculated here does, however, agree well with that of Shannon et al. (l97l) who calculated 'Y = 8 y with the model used in this work and based on plankton samples from the South Atlantic and Indian Oceans off South Africa. As was mentioned previously, it is possible to pre- 228 226 diet the Ra"” content of the plankton by knowing the Ra content of the plankton, the Ra228/Ra22( * activity ratio of seawater, and assuming there is no fractionation. These 228 predicted Ra values are listed in Table IV-8. Natural ly, they simply reflect the Ra228/Ra22^ activity ratio of 228 seawater. Samples nearshore (near the source of Ra ) 228 appear to have a higher Ra content than plankton samples offshore. This analysis of Ra isotopes in plankton does lead to the conclusion that, at least for Ra, the plankton are not very useful indicators of water movement. Techniques exist to measure the Ra content of the water directly and more easily (Mn-fiber concentration) than collecting plank ton. The Ra content of plankton is simply too low to be 159 used as a tracer. Pb Isotopes 210 The values of Pb concentrations contained in Table XV-7 illustrate some interesting trends. The coastal samples off southern California show the highest values. The samples collected in coastal waters, but north of Pt. Conception, are lower than those in the Bight. The samples off Mexico are somewhat lower than those within the Bight but higher than those north of Pt. Conception. The cause of this pattern seems to be two-fold: latitude and, more importantly, the relative aridity of the adjacent land. 210 Tsunogai et al. (l97l) have shown that the input of Pb varies with latitude and this source variation is reflected 210 in the concentration of Pb in surface Pacific waters. Air masses derived from arid lands (high Rn emanation) are 210 good suppliers of Pb to the surface ocean. The plankton samples off southern California and Mexico are taken from 210 waters enriched in Pb for this reason. O I A Benninger (1976) found 2 to 9 dpm/g Pb in his 35A* fraction and significantly lower values in the 333 M zoo plankton fraction. Kharkar et al. (1973) reported values 210 of 1 to 10 dpm/g Pb in plankton from the Caribbean. The 210 values observed here ranged from 3 to 22 dpm/g Pb • The values from off the coast of California and Mexico were about 2 times higher than the previous values primarily 160 due to the proximity of arid lands* The concentration factors listed in Table IV-11 were 210 calculated using an average Pb content for the eastern Pacific ocean surface water of 8 x 10 dpm/g (Bruland et al. , 1974). The average concentration factor of ^10 compares very well with that calculated by Kharkar et al* (1973)* This factor is about 10 times higher than that 226 calculated from Ra and about 100 times higher than that o qQ calculated for U , indicating that Pb is very efficiently being concentrated by marine organisms. The values of the 210 226 daughter/parent activity ratio Pb /Ra listed in Table XV-7 vary from 2 to 69 and average about 17* These simply reflect the ratio of the relative concentration factors for _ 210 _ 226 Pb and Ra • ^ -4 / 2 Assuming standing crop is 6.7 x 10 g plankton/cm 210 and that the Pb content of plankton is 10 dpm/g, then r o 210 2 6.7 x 10 dpmPb /cm is contained in the plankton. If the 100 m photic zone averages 8 x 10 dpm/g, then this 210 2 210 amounts to 0.80 dpmPb /cm . The Pb contained in 210 plankton represents 0.8 percent of the total Pb in this photic zone. 210 The mean time of biologic removal of Pb from the photic layer, calculated from A A = 0.23 dpm Pb^^/cm^/y where A = 0.8 dpm Pb^^/cm^ is 'X — 3*2 y. A Pb^^ atom 161 210 Table IV-11# Plankton concentration factors for Pb . Plate Number Sample Concentration Factor (x 103) 1 18027H 2.03 2 18031H 1.39 3 Bart 5 0.64 4 Bart 3 o# 76 3 18439 • 00 H 6 18030H 1.46 7 18028H 0 • H 8 Bart 4 0.39 9 18433 1.51 10 18449 1.21 11 Bart 6 2.27 19 Catalina 1.79 12 PS1 1.09 13 PS2 .83 l4 PS3-1 .77 13 PS3-2 .92 16 PS4 .75 17 PD2 2.23 18 PD3 CO * j H j where: CF = dpm//sc dry plankton dpm/g seawater 162 should reside three years in the photic layer before being biologically removed. This value agrees very well with value of about 5 years calculated by Shannon et al. (l970) based on a simple two-layer oceanic model using water values only. As mentioned before, the 3*2 year value is a 210 minimum because the Pb associated with plankton has been maximized and the entire production is removed annually. Th Isotopes The range of values for the Th isotopes are: .03 to .88 dpm Th^^/g, .11 to 1.05 dpm Th^^/g, and .15 to 1.55 ppQ dpm Th /g (Table XV-6). Previous reports (Cherry e_t al. , 1969) have deter- 228 mined an association between Th and water mass, however, the data presented here scatter widely and present no clear cut trends with water mass or distance from shore. Data of Kharkar et; al. (1973) on Caribbean plankton give 2 22 6 values of Th ranging from .Ob to .2 dpm/g and b 228 values of Th between .07 and .1 dpm/g. Because these 230 230 workers used a Th spike, no Th numbers were deter mined. Edgington et al. (1970) found marine algae to con- 2 T2 tain .01 to .15 dpm Th /g. The results presented here thus compare well with these previous data. Six plankton samples out of 19 analyzed were found 239 240 to have conspicuous Pu * peaks in the Th alpha spec— trums (second section) reflecting concentration of Pu by 163 plankton, as noted by Cherry et al. (1974). The concentration factors listed in Table XV-12 were calculated using two different seawater values. For -7 open ocean plankton samples a seawater value of . 3 x 10 pop dpm Th /g was used. This corresponds to the average 2 32 value of open ocean Th determined in a previous section with the restriction that only values significantly dif ferent from blank (but with no blank correction) are used. —7 With a similar restriction a seawater value of 2.48 x 10 p Op dpm Th /g was used for plankton collected within 50 km 7 of shore. The factors range from .02 to 1.02 x 10 and 7 average .27 x 10 . This compares well with a factor of .44 x 10^ of Kharkar ejb al. (1973)* These concentration factors are much larger than those determined for Pb, Ra, 232 and U. Because the surface concentration of Th is so low, this concentration factor for Th may be due primarily to trapping of Th-rich particulates (Turekian et al,, 1973) rather than to adsorption of an ionic species. -4 / Assuming a standing crop of 6.7 x 10 g plankton/ 2 232 , cm and a Th content of plankton of .25 dpm/g, then —4 232 / 2 1.7 x 10 dpm Th ~/cm is contained in the plankton in p op the photic zone. Xf the photic zone (100 m) had the Th content of open ocean water, then it would contain 3*1 — 4 2 32 2 x 10 dpm Th /cm . Thus plankton contain 53 percent of 232 the Th in the photic zone. This leads to the conclusion 232 that most of the Th in the surface layer must be as— 164 Table IV-12. Plankton concentration factors for Th232# Plate Number Concentration Factor (x 107) 1 1.02 2 0.43 3 0.09 5 0.06 6 1.04 7 0.26 8 0.49 9 0.03 10 0.08 11 - 12 0.03 13 0.33 14 o • o IP 13 0.02 16 o * o IP 17 0.31 18 0.12 19 0.24 where: CF = dpm/g dry plankton dpm/g s e awa t e r sociated with, the particulate phase. From what is known +4 about the physical chemistry of Th , this high degree of association is not unreasonable (Kraus et al* , 1954) • Because of this high association we would expect that the mean time of biologic removal of Th should be approxi mately equal to the cycling time of plant matter in the surface layer. As shown by Broecker et al. (l973) from a consideration of the P cycle, this plant cycling time is 2 32 about 0.3 years. We would predict that Th should be cycled at just this rate. In fact, as was shown in a 228 228 previous chapter by a consideration of Th /Ra ratios 223 in surface water, the isotope Th is being removed on a time scale of 0.4 years which is about the plant cycling time scale. Elemental Comparisons The concentration factors can be seen to vary in the sequence: Th Pb ^ Ra >U 10 > lO-^lO *^10^ 232 With the possible exception of Th , all the nuclides measured are present at least initially as ionic species (Kharlar et al., 1973) and thus these factors reflect the relative adsorption or trapping efficiencies of these ionic species. Thus, the Pb is most efficiently concentrated and the U the least. This selectivity sequence is compatible 166 with, the known chemical behavior of* these elements with respect to adsorptive processes* The above selectivity sequence is contrasted with that of Miyake et al* (1968)5 U > Th > Ra This sequence was arrived at by defining a biological activity index ( ) as follows: ft= regeneration rate from * organic material — surface water content input + radiogenera— • biological uptake tion- radiodecay Estimates were made of the value for each term in this equation. Using the values given in their paper resulted in the biological activity sequence above. 226 Xn one of the terms concerning the Ra content of plankton, there are some large discrepancies with this work and others (Shannon et al., 197l)* Miyake jet al. (1968) -9 226 / use a value of 10 Mg Ra /g while this work found an average of 10 ^ jag Ra^^/g. There is also a discrepancy 282 between our values for the Th content of plankton, al though it is smaller. It appears that these discrepancies are the cause for the very different selectivity sequences presented. The direct measurements on the concentrations of these elements in plankton and surrounding seawater presented in this work lend support to the sequence given here . 167 Summary A chemical procedure was developed for tlie sequential isolation and analysis of the isotopes of geochemical interest of the elements: Po, U, Th, Ra, and Pb contained in the natural decay series, A total of 19 mixed samples (phytoplankton and small zooplankton) were collected from nearshore and open ocean areas of the East Pacific, Sur face seawater samples were collected at these sites and analyzed for the same suite of radioisotopes, A comparison of these results allows a calculation of the concentration factors, the degree of association with plankton, and the mean time of biological removal from the photic zone for each element, Xn addition the geographic distribution observed for these isotopes may allow their concentration in plankton to be used as a tracer for water movements. To my knowledge there are no previously published results dealing with the determination of a suite of natural decay series isotopes in the same sample from the Pacific, The relative concentration factors were as follows: Th> Pb > R a > U as 10^;> lO^^lO^^ 10*^ on a dry weight basis. With the possible exception of Th, these reflect the relative adsorption or trapping efficiencies of these ionic species. This selectivity sequence is compatible with the known chemical behavior of those elements. Maximizing the productivity estimates results in 168 mean times of removal from the photic layer of 400, 7* 3* and 0.5 years from U, Ra, Pb, and Th, respectively. These are thus minimum estimates. As potential oceanographic tracers, U and Ra hold 228 little promise. Although earlier workers felt Ra might prove useful, the concentrations observed are extremely low, making detection very difficult. The most useful tracers, those displaying geographic water mass differences 228 21O and also easily detected, appear to be Th , Pb , and 169 CONCLUSIONS In the East Equatorial Pacific surface waters the 228 Ra activity generally decreases away from continental shelf areas. Across the Peru Current System this decrease can be modelled as one dimensional diffusion and indicates the possibility of two flow regimes with distinct character istic mixing lengths and apparent eddy diffusivities of * 5 7 2 10 and 10 cm /sec. The perturbing effects of advection, equatorial upwelling west of the Galapagos Islands, and displacement of the Inter-Tropical Convergence Zone (iTCZ) upon this simple diffusion model for the Equatorial Pacific surface layer have been noted. Off the coast of southern California the vertical 228 Ra distribution is used to estimate mixing rates through the surface mixed layer and upper regions of the thermocline. Apparent eddy diffusivities ranging from 1 to 3 cm /sec are obtained. Frequently, however, the ef fects of horizontal advection complicate the profile. The 226 observed Ra concentrations near the California coast delineate the importance of continental shelf supply of this isotope to the surface layer of the open ocean. The more extensive shelf north of Pt. Conception provides a 170 a stronger source of Ra tiian the continental borderland to the south. The insoluble daughter/ soluble parent activity 228 228 ratio Th /Ra in the Equatorial Pacific surface waters displays latitudinal trends which may be correlated with productivity variations. Near the coast of California this ratio reflects the differing oceanographic conditions on either side of Pt. Conception. Compared with areas to the north, the more extensive upwelling (and hence higher productivity) within the southern California Bight result in lower ratios, indicating a shorter mean time of chemical removal of Th and other highly reactive metals. 2 32 The Th content of seawater sampled was indis tinguishable from blank and is certainly less than 0.1 jag/ 1000 1 in support of Kaufman*s (1969) contention that most 232 of the published seawater Th values may be too high. 228 Profiles of Ra in the bottom kilometer of the Equatorial Pacific display to varying degrees the effect of eastward flowing Pacific Bottom Water (PBW). Modelled as one infinite layer the PBW is shown to have apparent vertical diffusivities increasing from west to east, per haps due to its interaction with the East Pacific Rise 228 (EPR). The source strength of the sediment for Ra de creases from west to east reflecting the dilution effect of carbonate sedimentation at shallower depths. Advective processes appear to play an important role 171 222 in governing the bottom Rn distribution* The standing 222 crop of the excess Rn profiles vary and show the same 028 222 trend as for Ra^ • At two stations the Rn profiles are remarkably similar to those taken in previous years nearby, arguing for a steady state feature* The observed distribu tions were modelled as one layer, two layer, or finite boundary problems* .. .. . . 230 / 23^- , rp, 228 / 228 The activity ratios Th /U and Th /Ra indicate removal processes for the reactive elements in the deep sea act on a time scale significantly longer than in 232 230 the surface ocean* The Th and Th content of bottom water is invariably higher than in surface waters* A chemical procedure involving sequential isolation and analysis of several isotopes of geochemical interest: U, Th, Ra, Po, and Pb has been developed. The relative concentration factors for these isotopes by plankton indi cate the sequence: Th > Pb > R a > U, Calculation of the degree of association of the above elements with the plankton demonstrates that the plankton uptake is very significant in the case of Th, much less so with respect to Pb and Ra, and of little significance with respect to the U content of the photic layer* As potential oceanographic tracers of surface water movement the U and Ra content of plankton are of little use. Of the radioisotopes studied here, the most useful potential tracers, those displaying water mass differences and relatively high concentration 172 1 228 T> 210 , _ 210 m plankton, ares Th. , Pb , and Po . A comparison of the two indirect methods of deter— . . _ 228 / . A 228 _ . 228x _ mining Ra (via Ac and via Th ) made on 64 seawater 228 samples shows that the time delay required by the Th method is more than compensated by its better analytical simplicity and precision. The fiber extraction technique utilized here has large potential for future research with trace radio- elements, The removal of Pu and Ac from seawater by this technique was demonstrated. 173 APPENDICES 11 h APPENDIX I Instrumentation and Calculation Methods APPENDIX I INSTRUMENTATION AND CALCULATION METHODS 222 2^6 Emanation Method for Excess Rn and Ra Seawater, plankton, and Mn-fibers were analyzed for 226 222 Ra via Rn by the emenation method of Broecker (1965). Bottom water samples were analyzed directly for excess 222 Rn via this method. Equilibrators For seawater samples 20 1 glass (Arrowhead Puritas Glass Products Division) bottles were used with bubbler heads made of leucite, tygon tubing, and a plastic Bel- Art s gas dispersion tube. Outlet tubes were thick-walled tygon, doubled—over and clamped securely with hose clamps when equilibrating. For Mn-fiber and plankton Ra solutions 250 ml Pyrex flasks with ground glass stopcocks and plastic gas dispersion tubes were used. The small equilibrators used for plankton and Mn- fibers were thoroughly cleaned with Alconox and a hot chelating agent (Micro Cleaning Solution). Flask blanks were run after each sample using Millipore Super Q de— 176 ionized water. Blanks varied from 0.1 to 1.0 dpm for flasks used with Mn-fiber Ra solutions. The flasks used for plankton Ra solutions had blanks ranging from 0.03 to 0.16 dpm. The large (20 l) glass bottles were cleaned by rins ing with distilled water between samples. Bottle blanks run periodically ranged from 0.03 to 0.13 dpm and averaged 0.08 dpm. Gas Extraction Lines Lines constructed by the LDGO Geochemistry Group (G—5 and G-13) were used for all seawater and plankton samples. These boards were made of stainless steel, glass, and a short section of tygon tubing. They were continuous ly leak tested by allowing the board to sit at evacuation overnight and monitoring the pressure rise. The procedure (Broecker, 1965) employs stripping the solution with He, adsorbing Rn at liquid nitrogen temperatures, and removing the and vapor with an ascarite column. The Rn gas is then transferred into the evacuated scintillation cell by a ' ’ milking” process following a second adsorption onto a coiled stainless steel tube at liquid nitrogen tempera tures. System blanks were run periodically by short- circuiting the boards with tygon tubing; they ranged from 0.07 to 0.2 and averaged 0.1 dpm. The 20 1 samples were stripped for 90 min and the 200 ml samples were stripped 177 for 15 minutes* The second liquid nitrogen adsorption was for 10 min in all cases. The Mn-fiber Ra solutions were extracted on a board of similar design, but with brass tubing, constructed by T.-L. Ku. System blanks for this line varied from 0.1 to 0.7 dpm and averaged 0.25 dpm. Radon Counters and Cells The seawater samples and plankton Ra solutions were 222 run for Rn using ZnS(Ag)—lined, acrylic detector cells fitted with Svagelock quick-connecters. These cells had backgrounds of 0.1 to 0.2 cpm. The cells were placed in light-tight wells fitted with RCA 6655A photomultipliers and TC 155A tube-base and preamps. Two different Rn counting systems were used. The older system was assembled by LDG0 and consisted of Canberra Dual HV Power Supply and Counter/Timer Model 1790. The new system xvas a Radon MK XX design by the GEOSECS Operations Group/NSF coupled with an IT Silent J00 model 733 data terminal. A total of 6 channels were available for Rn counting with the detachable cells. The Mn-fiber Ra solutions were counted in a stainless-steel, ZnS(Ag)-lined cell permanently attached to the old, brass extraction board. A photomultiplier and preamp were also attached. A Baird—Atomic Abacus Scaler Model 123B and HV power supply was used in conjunction with 178 this system. This cell had a background of 0.5 cpm. The overall extraction + counting efficiencies of all board, cell, and counter combinations were carefully monitored over the 4 year period during which samples were analyzed. These efficiencies were very stable and ex hibited a slight decreasing trend over the study period. A total of 7 standards solutions (2 small flasks and 5 bottles) were used in determining efficiencies. These standard Ra solutions were carefully prepared by Guy Mathieu of LDGO and used for intercalibrations among the Geosecs Ra analysis groups (USC, LDGO, and SXO). Ef ficiencies for the detachable acrylic cells, extraction systems, and counters varied (from ce 11 to cell) from 87 to 93 percent. The fixed stainless steel cell and ex traction system had an efficiency of 70 percent. Data Reduction The output from any of the above systems is a re cord of the total number of alpha decays (scintillations) detected over a given time interval: cpm. In computing 226 the Ra activity of the original sample of seawater, plankton, or Mn—fiber solution the following must be con sidered : 222 1. Extent of Rn growth toward equilibrium after initial purge (in—growth factor). 179 2. Alpha amplification due to the daughters of 222 Rn (alpha factor). 3* Extraction and counting efficiency. 4. Cell background. 5. System blank. 6. Equilibrator blank (flask or bottle). Considering these separately: 1. Ingrowth factor: after the initial He purge 222 which completely strips Rn from the sample) then at time t: ; H t w ' 22^ 222 \ where: 1 = Ra , 2 = Rn and the N and A are respective ly the number of atoms and decay constant. However, ^ ^ 9 so e-^ec' t the ratio Rn222 ^/Ra22^)Q is given by: ft, NX. _ , _ - (I N, 1 222 Usually 7 to 14 days of Rn ingrowth were allowed. 2. Alpha amplification: the decay scheme of „ 222 . Rn is: Rn222 "•> Po218" > Pb21^ Bi21^ Po21^ Pb210 3.8d 3.1m 26.8m 19.7m ,0001s 22y The activity of the Pb210 alpha emitting daughter, Po21°, 180 210 is limited to the growth of Pb and hence is negligible for the counting period used. The solutions to the Bateman equations for the activities of the alpha emitters are: . .0 -2,£ ^■2. — - A- ^ a o 7 ^ / -'X .'t "Ait a2 = A1 (e. ' - e * aa,) aftifaX) a3 - 4 Y 4 - ; o p ~ X * - t + ----%--— ------r -f- e (X;),)fX; I) a3 J a ; m , - X t ) 222 218 where: A^, , A^ are the activities of Rn , Po , and 214 o 222 Bi , respectively; A^ is the activity of Rn at the time the cell was filled: 2 9 ^3* /14 are the 4 - 4 - -r 222 x > 218 o^21^ , ^ • 214 decay constants for Rn , Po , Pb , and Bi , re- 222 spectively (Friedlander e^t al. , 1964). Usually the Rn + He carrier gas are flushed into the chamber and the cell is allowed to sit for 2 to 3 hours. From the decay scheme it can be seen that all the daughters down to and including 214 Po will essentially have grown to equilibrium by then. Then the cell is monitored for total alpha activity for a period of 3 to 4 hours. The gross alpha activity will 222 be approximately three times the Rn activity because 218 214 Po and Po alpha decays are indistinguishable from 181 J w J WJ n o n Rn decays (Sarmiento e_t al. , 1976)* The original Rn activity in the cell is then calculated using the Bateman 222 equation which is essentially correcting for Rn decay during the counting interval and time since filling the cell, 3. Efficiencies: these were monitored as described previously and applied to the net count rates: dpm = net cpm efficiency Cell backgrounds: these were determined by counting cells filled with 1 atm He for ^ 1000 min prior 222 to filling the cell with the Rn from a sample. This was done before each run and the activity (cpm) was sub tracted from the observed gross alpha activity, 5 and 6, System and equilibrator blanks: these were run as described previously and the resultant activities (dpm) subtracted from the calculated (dpm) 222 Rn activity. All the above corrections are applied to the ob served gross alpha activity (cpm) as follows: t cpm — cell background___________H — [""system b 1 anl^ — ALPHA FACTORX effXingrowth factorj * /"equilibrator I j^blank J 226 = Ra activity of the initial solution xn dpm. The error associated with the above analysis should in theory be determined by the counting statistics. For each run a counting statistic error was calculated 182 based on tbe total number of* events recorded. For each sample tbe Rn extraction and counting was repeated 2 or 226 more times. All the data presented here on Ra quote an error based on the standard deviation in the mean of* the replicates or the largest individual counting error, which ever is larger. Xt should be mentioned that Sarmiento ejt al. (1976) has pointed out that calculating individual errors based on the counting statistics should be determined by only 222 the Rn activity and not the gross alpha activity as was done here. However, in almost all cases the standard deviation in the mean is larger than the counting error of individual runs, so the quoted errors here are fairly good estimates of the precision. 222 In determining the excess Rn concentration of 226 bottom waters, all of the above steps utilized in the Ra analysis were applied with the exception, of course, of the ingrowth factor and the bottle blank. Alpha Spectrometry Two alpha spectrometers were used in this study, providing a total of six Ortec h^O mm silicon surface- barrier detectors. One system (k channels) consists of: Nuclear Data preamps and linear amps, Canberra linear gates and sum-inverter, Nuclear Data mixer-router and a 102k channel multi-channel analyzer. The typical sample 183 resolution of this system is about 80 kev FWHM (full-width at half—maximum)* The other system (2 channels) consists of: Tenelec preamps and linear amps, Xnotech mixer and IT-5000 multichannel analyzer. Resolution of this system is typically 70 kev FWHM. Detector backgrounds were carefully monitored ap proximately once each month during the study. In the regions of interest there were generally - 21,007 cpm/per peak. These systems were operated in a small, air condi tioned counting room and were very stable electronically allowing good resolution to be maintained over long count ing periods (occasionally as long as one week), Proportional Counter For the Mn-fiber Th determinations it is necessary 23 4 to measure the Th activity of the final Th mount. This is accomplished indirectly by counting the equilibrium P o/l activity of its daughter, Pa (ti = 1.18 m). These fiber 2 34 mounts usually have several hundred dpm or more of Th / 234\ (Pa ) activity. They were counted in an NMC Model PC - 3T proportional counter which had a beta—alpha background of 20 to 30 cpm. The efficiency was given by the manufac turer as 52 percent. 184 Low Level Beta Counter* ^ . 228 / . . 228\ This instrument was used m Ra (via. Ac ; 284 / analysis of* Mn-fiber Ra solutions, as well as Th (via 234x Pa ) analysis of* seawater samples* The "home-made" system consists of 4 sample gas — flow Geiger detectors operated simultaneously in anti- coincidence with a larger guard gas-flow Geiger detector. The sample detectors were disc—shaped (l.l cm radius X 0.6 cm thick) thin-walled Geiger counters (Lai et al.. i960). These detectors consist of two thin conducting films of Au mylar (cathode) and a thin, straight stainless steel wire (anode) strung across the center of the circular chamber. The counters were operated in the Geiger region with "Q"—gas (a commercially available mixture of 98.7 per cent He and 1.3 percent butane). The 1 mil anode wire was supported by clamping inside thin (25 g) stainless steel hypodermic needles which were glued firmly to the acrylic plastic body. The length of the hypodermic needle protrud ing into the chamber was kept as small as possible to in crease counting efficiency. The other end of the needle made contact with the center pin of a SHV connector. The two 0.25 mil Au mylar films (Schjeldahl Co., Minn.) were in electrical contact with the SHY connector base. The gas inlet — outlet tubes were mounted asymmetrically to insure efficient flushing. 183 The one, large guard counter was rectangular in shape (27*9 cm x 33 cm x 2.3 cm) and consisted of a total of 10 straight stainless steel (2 mil) wires (anode) strung across the counter using hypodermic needles for support and two stainless steel (l.2 mm) sheets (cathode). The anode wires were spaced symmetrically spaced 2.3 cm apart. The electrical connections were made exactly as for the small counters. All of the counters were assembled at the University of Southern California by the author. The large guard counter was placed under the small sample counters and a coincidence-anticoincidence circuitry logic was used. All the electronics were designed and con structed at the University of Southern California by Craig Todd with the exception of the DC power supply and the HV Detector Bias Supply which was purchased commercially. These circuits could be used to identify three types of particles using the two counters: particles that activated the guard counter only, particles that activated the sample counter only, and particles that activated both counters essentially simultaneously (Fig. AI-l). Particles that activated only the guard counter are primarily M — mesons produced from cosmic rays. Particles that activate the sample counter only (anticoincident) were assumed to be from the sample. Particles that activated both counters (coincident) were also assumed to be jul— mesons. The anti- coincident counting rate is the sum of sample plus back- 186 Sample Detector Guard Detector Figure Al—1. Types of particles identifiable using the two detectors: 1. Particles that activate guard detector only. 2. Particles that activate sample detector only. 3. Particles that activate both detectors simultaneously. 187 ground rates, which includes counter contaminants and electronic noise. The sample and guard counters were operated inside a Pb shield consisting of a minimum of 4 in. of Pb in any direction. A sliding tray on the top of the shield was fabricated of 3/4 in* thick Al onto which were loaded 4 in. of Pb bricks. As noted by Lai e_t al. (i960) a very good indicator of the cleanliness of the counter is the ratio of co incident to anticlincident counting rates, P. The former is mainly a function of the counter size, while the latter represents primarily the sum of the counting rates due to contamination of the counter materials, to electrons produces in the wall and gas, and to gamma—flux inside the Pb shield. These authors considered the counters to be "clean" if R was 21^5 or larger. By this criteria all counters used in this study were satisfactory. All the lucite parts of the counters were cleaned in Alconox and water, soaked in dilute HC1, and then rinsed thoroughly in deionized water, as suggested by Rajagopalen (1973)# Plots of counter characteristics are given in Fig. Al—2, Table AI-1 summarized these characteristics. The counters were fairly stable and over long periods of time maintained relatively constant background rates. The counting plateaus did shift somewhat and the operating voltages were increased twice over a two year period of operation. 210 210 The counter efficiencies, monitored with a Pb (Bi ) 188 Figure AI-2. Plots of* counter operating characteristics: relative count ing rate versus applied bias voltage. 3 0 0 0 2 9 0 0 2 8 0 0 2 7 0 0 2 6 0 0 2 5 0 0 2 4 0 0 £ ^ 2 3 0 0 O 2200 2100 2000 1900 1800 1700 1600 2 8 0 0 2 7 0 0 2 6 0 0 2 5 0 0 2 4 0 0 £ 2 3 0 0 0 2200 2100 1000 1100 1200 ,1400 1300 1200 1100 § 1000 Q. O 9 00 8 0 0 7 0 0 6 0 0 5 0 0 4 0 0 GUARD 175 VOLTS 12% SLOPE Operating Voltage Arc 125 VOLTS 5% SLOPE Operating Voltage DETECTOR 1000 1100 1200 150 VOLTS |< * 1 - 3% SLOPE t Arc *t Operating Voltage DETECTOR 2 1100 1200 1300 Volts 190 Table AX—1* Counter characteristics Counter 1 Counter 2 Guard Plateau slope y/o 3 7o 12°/o Threshold voltage (volts) 975 950 1075 Operating voltage (volts) 1025 1025 1150 Coincidence rate (cpm) 1.289 1.311 - Background anticoincidence rate (cpm) .056 .061 — R factor (cleanliness) 23.0 21.5 — 191 source, showed changes by a maximum of 10 percent. On a few occasions the electronic noise increased drastically so that the counters were unusable for low activity samples. This occurred a total of about 6 days over a 2 year period and may be related to a temperature instability of the electronics (Applequist, 1975)• 928 For the Ac measurements it is necessary to know 234 the absolute efficiency of the counters. For the Th measurements this is not necessary; one need only know the relative counting efficiency of this low level beta counter versus the alpha spectrometer. To determine the absolute counting efficiency two methods were used. It should be mentioned that all measurements (Ac and Th) were made using a thin-source mount on stainless steel plates. The plates used were of a diameter exactly equal to the diameter of the counter (2.2 cm). The evaporating TTA—benzene phase was kept centered in a small area on the plate, hence the counting geometry is essentially 2 . Counting efficiency for 228 Ac (beta E =1.1 Mev) was determined as 5^ percent max 7 ^ (- 10 percent max variation over 2 y) by repeated extrac tions from an old Th(NO^)^ solution (supplied by ¥. S. Moore) which has Th22^/Th2^2 = 1.0. The absolute count ing efficiency could also be determined by counting the 23Z 1 Pa (beta E =2.2 Mev) on a uranium mount made from a max 7 Black Hills uraninite solution (U^'^/U^'"^ = 1.0). About 6 192 months after preparation of such a mount the plate should 228 22^ 224 have equal activities of U , Th , Pa , and The 2 o2t 2 38 U + U alpha activity was determined precisely on a gas-flow proportional counter of known efficiency. This activity is then compared with the beta activity of the same plate counted on the low level beta counter, using an 238 23b Al adsorber to cut off activities due to U , T J , and 23b Th . This resulted in an efficiency of 58 percent for the 2 3b 228 Pa betas, which is higher than that determined for Ac • The difference could be due to the more energetic beta of 2 r>h Pa (and hence greater back-scattering) or due to a 23b / small percentage of the relatively weak Th betas (beta \ 23b Emax =0.2 Mev) being counted along with Pa . The relative counting efficiencies (beta/alpha) of the low level beta counter and the proportional counter versus the alpha spectrometer were both calibrated in the same way. The uranium mount made from the Black Hills p q Q uraninite was counted for U alphas on the alpha 234 spectrometer and then counted for Pa betas on both the proportional and the low level beta counter. The ratio of these two raw count rates (cpm) after background correction gives the relative counting efficiencies (beta/alpha). For the proportional counter versus alpha spectrometer this factor was 1.53# For the low level beta counter versus alpha spectrometer, this factor was 1.74. 193 Fiber Calculations _ 228 . ^ 228 Ra via Th. After removal of Th from the fiber Ra solution (second section), the solution was stored. During this 228 228 time Th grew towards equilibrium with the Ra in the 228 solution (Fig. AX-3)* The ratio of Th present at time, 228 t, of isotopic analysis, Th^_ , to the original activity of Ra^^^ present after Th removal, RaQ22^, can be calculated from the Bateman Equation: J where; A-t- QnJ /I* are the decay constants of Th22^ and A p p Q Ra^^°, respectively. From the alpha spectrometrie analysis for Th of the 228 2 30 stored Ra solution, one gets the ratio Th /Th . Figure AX—k is an example of a Th spectrum resulting from an analysis of a stored fiber Ra solution. Multiplication 2 20 poQ of this ratio by the Th spike activity gives the Th ? 2 8 activity of the solution at time = t; i.e., Th ~ . u o p Q p p Q Dividing this number by the above ratio, Th^ /Ra0 » 228 228 gives RaQ , which is the initial Ra activity of the fiber Ra solution at the time of fiber processing. 228 This number may then be corrected for the Ra decay between sample collection and fiber processing by 19k LOG ACTIVITY (dpm) 100 80 60 \ 40 20 600 1200 3600 1800 2400 3000 4200 228 Figure AI-3. Growth of Th towards equilibrium with Ra22^ in a Ra solution initial ly free of Th. 195 Figure AX-4, Sample of Th alpha spectrum (plotted as raw counting data) resulting from an analysis of Th contained in a stored Mn-fiber Ra solution after ingrowth of Th^28 anci spiking with T h ? 3 ° . 196 COUNTS GROW-IN" Th(Ra) GULF of CALIFORNIA At = 1749 m in . 226 ^.,230 Tn spike 200 150 100 224 00 40 60 80 120 140 100 160 200 180 220 240 C H A N N E L dividing by the ratio: 228 " ~ X Ra0 = e A *r 228 acollect. •where t = the time between collection and processing, 228 We now have the Ra activity of the fiber when it 226 . . was collected. Dividing this by the Ra activity of the 223 226 fiber Ra solution gives the Ra /Ra activity ratio of the fiber at collection, which is equal to this ratio in the seawater sampled. Multiplication of this ratio by the Ra^ content of the seawater (measured in 201 samples) 228 gives the Ra of the seawater at collection. These are the data listed in Table II—2, etc. Errors are based on counting statistics 1 sigma). Absolute Th Calculation As described in the second section, this involves both alpha and beta counting. The fiber Th mount is first 234 beta counted to determine the Th activity. This number 2 34 is then corrected for Th decay between sample collection and fiber chemical processing using the equation: Th23^ = 23^ C Th J o The Th mount is then counted in the alpha spectrometer to determine the relative Th232, Th23°, and Th22^ activities, 228 Of these only Th need be corrected for decay between 198 collection and analysis by using the above equation sub— 2 2 8 stituting the decay constant for Th . Dividing these 2l4 numbers by Th then gives the fiber activity ratios at the time of collection: Th232/ ^ 23^, Th23°/Th23^, Th228/Th23Zf These fiber ratios are then related to the actual concentra tion in seawater by the following equation: 232 230 228 /, \ ^232 _ 2 30 228 Th , Th , Th (dpm/1000 1) = Th . Th . Th x Th23^ x 2°°° (dpm/l03l) 2460 j&/ 1 where: — 1.53 or 1.74 depending on which beta-counter was used 2000 dpm/lOOO 1. = Th23^ activity of surface seawater 2 0/1 2460 dpm/lOOO 1. = Th activity of bottom seawater These are the final Th values tabulated. Errors are based on counting statistics (- 1 sigma). Plankton Calculations „ 210 / . 210\ Pb____ (via Po___ I As described in the fourth section, the Po isotopic analysis on the alpha spectrometer gives the activity ratio 210 0 OQ Po /Po^ (Fig. AX—5)• Multiplying this ratio by the 209 Po spike activity and dividing by the dry sample weight gives Po210 = Pb2' 1 ' 0 in dpm/g: 199 Figure AI-5# Sample Po alpha spectrum of plankton sample plotted as raw counting data. 200 spike PLANKTON n | , Po At® 1521 min. — J 160 180 Po210 X Po209 (dpm) X — = Po210£» Pb210 (dpm/g) Po209 210 These are the final Pb values tabulated. Errors are based on counting statistics (- 1 sigma). U As described in the fourth section, the U isotopic analysis on the alpha spectrometer gives the activity ratios: and (Fig. AX—6) . Multiplying 2 32 223 this ratio by the U (Th —free) spike activity and dividing by the dry sample weight gives: u238. U234 x U232-Th228 free (dpm) x ^ = U238, U232 U234 (dpm/g) These are the final U values tabulated. Errors are based on counting statistics (- 1 sigma). Th As described in the second section, this analysis requires both alpha and beta counting. The sample Th 2 32 2 30 isotopes of interest are alpha emitters (Th , Th , and 228\ 23b Th ) but the spike, Th , is a beta-emitter. The cal culation is completely analogous to that previously des cribed in this appendix for calculating absolute fiber Th numbers. The equation in this case is: 202 Figure AX-6. Sample U alpha spectrum of plankton sample plotted as raw counting data. 203 zoh 300 250 c / ) _ _ l~ 200 z D O o 150 100 00 U 238 20 40 60 PLANKTON ttl6 U At * 5535 m in . 8 0 100 120 140 160 180 CHANNEL Th232. Th23°« Tft228 x ^ x Th232f (dpm) x = Th23 ~ Th232, Th230, Th22® (dpm/g) p p Z i . where: Th (dpm) = spike activity added to the sample as of the time of alpha counting. These are the final Th values tabulated. Errors are based on counting statistics (- 1 sigma). Ra ^ The Ra content of the plankton Ra solutions was determined by the emanation method described in detail 226 previously in this appendix. This Ra activity, when corrected for the 6 percent loss during the Pe(OH)^ precipi tation (Applequist, 1975)» gives the final Ra numbers tabu lated. The calculation of error was described previously. 205 APPENDIX XX A 228 - 0 228 Ac Counting for Ra 206 APPENDIX II Ac228 COUNTING FOR Ra228 Calculations As proposed by Moore (1969)* ^ possible to 228 measure Ra indirectly by determining tbe activity of its 228 daughter Ac • The second section described in detail the chemical procedure involved in isolating Ac from the fiber Ra solution and mounting it on a stainless steel plate. This plate was then covered with clear mylar (6 ja thick) and counted in the low—level ^ -counter described in Ap pendix I. 228 To calculate the initial Ac activity on the 223 plate (and hence the Ra activity of the fiber Ra solu tion) , an integrated count rate procedure was used. The time intervals that must be considered are: *0 _____________ tl _____________ *2 _____________ t3 (TTA extraction) (start Ac count) (end Ac count) (end BKGD) (start BKGD) Since two TTA extractions were performed, tQ was taken as the time between the two extractions. Customarily, the time interval varieci from about 60 to 120 minutes. 207 The time interval was approximately l400 minutes (or 228 1 day) and during this time the majority of* the Ac has decayed away. The time interval t^-t^ was usually 200 to 300 minutes. The count rate recorded between times t^ and t^ was used as the background count rate (A gj^Qp) • This included the detector background as well as any long-lived con taminants. The number of counts recorded between t^ and t^ was used as the gross counts recorded, C . By multiplying gro s s the background count rate, times the time interval, t^-t^j the background counts, was calculated. This number ( ) was then subtracted from C to give v BKGD* gro s s 228 ^ e t j i.e., the total number of Ac decays that were detected in the time interval t^ to t^,. 228 To determine the total number of Ac decays that took place between t^ and t^, C, we must divide by the absolute counting efficiency of the low-level beta counter: C net = C This value, C, must then be related to N , the 228 number of Ac atoms present at the time of TTA extrac tions. There are two time corrections that must be made. First, during the time interval (elapsed time 1^00 to 1500 minutes) only 95 percent of the Ac decays have taken place. To calculate the total number that would have taken 208 place, C is divided by tbe factor: I - e~x This then gives the total number of Ac atoms, (N^), that were present on the plate at time t^. Second, we must ac count for the Ac which has decayed between t and t^. This is calculated from: A/. = N , c * “ •-*■) 228 This value, Nq , is the total number of Ac atoms in the fiber Ra solution at the time of extraction, which is equal 228 to the Ra concentration because they are in equilibrium. Problems with Ac Chemistry 228 Although the technique of determining Ra via OpQ Ac is often cited as a useful one (Moore, 1969* 1973» etc.), there are several problems with it. The most important problem deals with the TTA extrac tion of Ac. An excellent review of this problem was made by Kirby (l95^) and X shall draw heavily from this work. Quot ing the author: The best—known, most widely cited, and perhaps most highly overrated extractant for Ac is TTA . . . This reviewer has been unable to uncover any published or unpublished evidence that TTA extraction has ever been successfully applied to the quantitative and reproducible determination of Ac, although its use is recommended in every review of either Ac or TTA . . . Unfortunately, this work performed some 20 years later would not change Kirby*s view. 209 According to Hagemann (l950)» Ac was essentially completely extracted from aqueous solution by an equal volume of 0.2 5M TTA-bensene at pH ^>5.5. Unfortunately, details such as ionic strength, time between pH adjustment and extraction, contact time, etc. were not described. These are critical because quantitative extraction of Ac is severely hampered by hydrolysis at pH>;5.5. Polymeriza- +3 tion of Ac limits the Ac available for chelation. Also, the formation of a TTA chelate releases one II+ for each positive charge of the cation, so the pH can change as a result of the extraction, thus effecting reproducibility. To counteract these two problems (polymerization and pH changes) the present author deviated from the procedure of Moore (1969) and instead of adjusting the pH by drop- wise addition of NaOH, a sodium acetate-acetic acid buffer solution was used (pH = 5.7) to dissolve the BaCl2 crystals for TTA extraction. This results in improved reproduc ibility because the extraction is accomplished more rapid ly (thus limiting the fraction of Ac present in polymeric form) and because the buffering action prevents any pH changes. In addition, the high salt concentration en hances the distribution coefficient between the organic and aqueous phases (a salting out effect). However, even with this modification, on the average, only a precision of - 15 percent for the Ac measurements was achieved. Salcanoue ejt al. (1970) developed a technique utiliz — 210 ing a solvent extraction at pH 1.3 (di-2—ethyl hexyl phosphoric acid in toluene) followed by cation exchange. This more involved procedure is not quantitative, so they used Gd^"^ as a tracer for Ac (ti = 93 y» - 3»l8 Mev) . 2 Extraction of Ac at low pH is preferable (less polymeriza tion) ; however, stability of rare earth complexes in creases with a decrease in ionic radius and the ionic radii (for +3 valence in 8—fold coordination) of Gd and Ac differ by 20 percent. As shown by Kirby (l95^)» the stability constants for the D2EHP complexes of the two nuclides could easily differ by an order of magnitude or more. Xt is indeed unfortunate that Ac possesses no suitable —emitting isotopes to use as a tracer. Potential Counting Problems The fiber Ra solution contains an appreciable amount of Ra2^^ whose daughter products Pb^"^ and Pb^"^ could 228 interfere with the counting process for Ac because they are strong B—emitters and they would be extracted with TTA. 210 The Pb presents less of a problem because it is removed in a cleanup TTA extraction before the first Ac count and can grow back toward equilibrium only very slowly (22 y P 1 I x half-life). The short-lived 26.8 m Pb , however, must be removed from solution each time before Ac counting. This is accomplished by heating the solution and taking it 222 214 slowly to dryness to remove Rn . Any Pb in the solu— 211 tion becomes unsupported and decays away in a few hours. Since at least 2 hours would pass routinely between final evaporation of the Ra fraction and the start of the Ac 2l4 count, this should be sufficient to remove all Pb . How ever, on several occasions the initial count rate observed during the Ac measurement was significantly greater than the integrated count rate, indicating contamination with a short-lived isotope. On all Ac counts recordings were made periodically over the 2k hour period to determine the apparent decay half-life being recorded and if any signifi- 228 cant departure from a 6.1 hour half-life of Ac was ob served, the count was discarded. . 228 _ 228 _ Ac versus Th Results Realizing the limitations posed by this method, an attempt was made to compare it with the more reliable, but 228 more time-consuming, method of Th counting after ap proximately 1 year or more of in—growth. For this purpose 64 samples were analyzed by both methods. In order to improve precision of the Ac^^^ analysis several means were employed. An acetate buffer was used for the TTA extraction. All chemical processes were timed, and all samples were run under identical time limits. All Ac counts x^ere repeated 2 or 3 times and the valid results were averaged. All counts were screened for possible problems such as (l) high initial count rate, (2) apparent 212 22 8 decay rate different from Ac ~ , and (3) high level of long-lived activity* Number 1 above could be corrected by disregarding that portion of the counting interval. Number 2 indicated that the count was not usable and so was re jected. Number 3 was corrected by projecting the back ground observed from t^,—t^ over the interval Also, 2 2 8 for a given sample if one Ac~ calculation was very much lower than the other two (indicating bad TTA extraction), it was rejected. Using the above somewhat arbitrary criteria, the fiber Ra^^^/Ra^^^ activity ratios calculated using Ac^^^ 2 P 8 measurements for Ra ~ are compared to those determined 1 228 year or more later by Th measurements (Table AIX-1 and 22 8 2). Most of the Ac results (>\50 percent) are within 1 228 — 15 percent of the Th results. They are neither con- 228 sistently higher nor lower than the Th results. Con sidering the efforts taken to maximize reproducibility, this is probably a fair estimate of the routine precision of the method (- 15 percent). It is concluded that for 22 8 22 8 most samples the delayed Th counting for Ra produces a precision better than — 10 percent and is to be preferred 228 over the Ac method. 213 Table AII-1. Fiber Ra-228/Ra-226 Sample via Ac—228 via Th—228 Sample via Ac-228 via Th—2 2 8 Geopac 322 .010 .015 Bartlett 13 .107 .082 320 .011 .015 15 .106 .091 324 .015 .017 17 and 18 .099 .083 326 .020 .030 19 . 112 .092 331 .030 .032 20 . 12 8 .095 33^ .099 .055 B-18 ♦ 130 .120 343 . 100 ♦ 105 B-19 .111 .081 345 .106 .078 B—22 .079 .090 346 .090 . 086 B-39 .084 .062 347 .064 .074 b-4i .087 .065 Bartlett 1 .151 .152 b-42 .097 .080 2 .037 .037 B-43 .244 .099 3 .098 .060 H 1 O .290 .298 4 .072 • 066 0 1 ro .117 .100 5 .092 .075 C- 3 . l4l . 110 7 .15^ .082 0 1 .060 .038 9 .076 .059 C- 5 .115 . Ill 10 .137 . 104 H 1 O .089 .071 11 . 122 .049 C-18 .169 .103 12 .062 .032 C-19 .179 .149 21^ Table AII-2. Fiber Ra-228/Ra-226 via via Sample_______________________ Ac-228____________________Th-228 Gulf .545 .73^ PS 1- 0 .245 .262 1-20 .100 O - 3 - 1 H .153 .177 1 O .098 .095 1 - “ ' J o .081 .078 2-120 . 068 .074 1 O .092 .096 3-80 .084 .089 3-150 .063 .076 0 1 .069 .095 4-60 .051 .076 4-150 .036 .059 PD 2- 0 .072 . 116 2-60 .077 .127 2-150 .072 .105 1 o .078 .119 1 o .070 .092 3-75 .055 .076 3-150 .038 .059 0 1 .051 .097 4-90 .138 .071 4-145 .048 .048 5- 0 .102 .096 5-75 .059 5-150 .044 .069 215 Seawater APPENDIX III and Activity Ratio APPENDIX III SEAWATER U AND U ^ V u 238 ACTIVITY RATIOS Seawater U Analyses In order to determine the Th isotope concentrations 2 34 in surface and bottom waters the Th content of the water was used as a natural tracer. This was inferred from the 2 o 2 38 previously measured values for Th /U for surface (Matsumoto, 1975) and bottom (Amin et al.. 197^0 waters O oQ and the U content of the ocean. The U content and isotopic ratio of seawater was recently reviewed by Ku et al. (1976) and data from that work pertinent to this study is presented in Table AIII-1. In the Pacific Ocean the U concentration and isotopic ratio appear to be depth independent and average 3*3^ M s/1 and 1.13> respectively. As a part of an ill-fated attempt at simultaneously 234 determining absolute Th in Pacific bottom water, it was necessary to determine U concentrations in the Fe(OH)^ 234 / precipitations and this allowed calculation of the U / U238 ra- j -^0 seawater. The average U2"^/U2^8 value of 1.12 for the 32 analyses for Pacific bottom waters confirms that derived from the three bottom water samples listed in 217 Table AIII-1. Uranium data from the Pacific Ocean (samples collected July-August 1972, 8th Cruise of R/V Dmitry Mendeleev) Depth M ... Temperature (°c) Salinity (0/00) Uranium lug/l) Activity Ratio 234u/^38u Station D.M. 556 (07°50'S, 99°55,Ws PDR depth: k 251 ra) 0 24.89 35.173 3.23 + .05 1.13 i .02 250 11.21 34.872 3.3^ T .04 1.13 i .01 450 8.24 34.672 3.38 * r .05 i.!3 i .02 1000 4.41 34.590 3.18 T .05 i.i3 i .02 4200 1# 75 34.682 3.55 + .07 1.14 ± .02 Station D.M. 564 (09°08'S, 110°33*W; PDR depth: .30.73. ra) 3025 1.75* 34.68* 3.3^ + .05 1.14 ± .02 Station D.M. 567 (11°12.1'S. 106°221W; PDR depth: 3750 m) 3500 1.75* 34.68* 3.39 + .14 1.12 i .05 * Data read from Figure l4 (temperature) and Figure 45 (salinity) of Muromtsev (1963). ro 00 AIII-1. 2o/i The value of 3*34 A*g/l u may be used to estimate 2 24 2 24 / 2 28 the Th content of seawater if the ratio Th ^ /U is known. Prom a careful analysis of 84 seawater samples from the upper 200 m of the Pacific Ocean, Matsumoto (1975) + 2 38 found a value of 0.8 - 0.1 for this ratio. Using my U value and his Th^'^/U^'^ value, I calculate a value of 2.0 i 0.3 dpra/l for Th^"^. This is the value I used in all Mn-fiber Th calculations for surface (200 m) samples. Similarly, Amin et al. (1974) analyzed 9 Pacific 224 / 228 Ocean bottom water samples and found Th /U =1.0. 228 Using a value of 3*34 ;u.g/1 U , this results in a value of 2.4 dpm/l Th^^^. This is the value I used in all Mn-fiber Th calculations for bottom samples. The technique used for U analysis is described in Ku e_t al. (1976). Pore Water U and Ra On the eighth cruise of the Soviet RV Mendeleev, the author had an opportunity to collect large volumes (>500 ml) of porewater. These interstitial waters had been extracted at in situ temperatures using a modified Presley- Kaplan squeezer from freshly collected deep-sea sediment cores from the East Pacific. The samples were squeezed from the top sections of cores through 0.45 M Millipore filters, acidified, spiked, and co-precipitated with 219 Fe(OH)^ on board (Ku e_t Etl. , 1976). The results (DM-H and -i) are shown in Table AIII-2. In addition four samples of "core-top” water were collected from the tops of gravity cores by siphoning off from the sediment surface while the plastic core-liner was kept in the vertical position. These samples were frozen until analysis. At that time the thawed samples were passed immediately through 0.4 p. Millipore filters and processed exactly as the seawater samples above. These samples should be some combination of bottom water and pore water extruded from the sediment by the process of gravity coring. Their U data (Table AIII-2) are not significantly different from those of bottom water, in agreement with that observed by Somayajulu ejt al. (1973)> however, they are not in accord with some reported data showing large variability and departure from normal seawater in U con centration (Baturin, 1971? Dysart et al., 1975) and in u234y/lj238 activity ratios (Dysart et al. , 1975) • o o Ra in the "core—top" water of Bartlett station 17 was also analyzed and show a concentration of 52.8 dpm/ 100 kg (see Table AIII-3)* This was determined by the emanation method on the supernatant from the ferric hydrox ide precipitation used to concentrate uranium. A 6 percent 226 correction was applied for the Ra carry-down by the precipitation. According to Chung e_t al. (1972) the Ra content of bottom water in this region is about 37 dpm/ 220 Table AIII-2. Uranium in East Pacific sediment pore water and core top water• Sample Sample Location Sediment U 23kv/238v Number Size (ml) Lat. Lon#. _ . Type (ug/l) _ D.M.—H 800 11°12'S 106°221¥ Metalliferous ooze 3.61 - .11 1.14 i .Ok D.M.-I 500 7°50'S 99°55*¥ Metalliferous ooze 3.68 — .16 1.11 t • 06 Bartlett 16 610 7°^0'N 9^°30f¥ Calcareous ooze 3.51 - .29 1.29 - .12 Bartlett 17 3360 9°^3'N 9^°32*¥ Calcareous ooze 3.02 i .05 1.12 i .02 CSII-6#1 kn 29°33'N 118°26 * ¥ Light-brown clay 3.35 - .15 1.17 - .06 CSII-6#2 930 29°33'N 118°26,¥ Light-brown clay 3.37 - .1^ 1.16 i .05 N ro H 100 kg. The Mcore—top” water value Is about 40 percent higher, indicating that the water may contain a significant amount of Ra-enriched pore water. Xt is, however, a factor of k lower than the pore water value of Ra^^^ ob served by Somayajulu e_t al. (l973) in a South Pacific sedi ment sample. 222 Table AIII—3* "Core top" water Ra 226 Bartlett Station 17 52.8 dpm/100 kg BIBLIOGRAPHY BIBLIOGRAPHY Aller, R. and Cochran, J. (1976), Th /u disequilibria in nearshore sediment: particle reworking and diagenetic time scales. Earth Plant, Sci, Lett. 29, 37-50. Amin, B. S., Krishnaswami, S. and Somayajulu, B. L. K. (197*0 * Th23*+/u238 activity ratios in Pacific Ocean bottom waters. Earth Plant. Sci. Lett. 21, 342-3*+**. Applequist, M. D. (l975)» Lead—210 in the deep sea: Pacific Ocean investigations. M.S. thesis, U.C. San Diego, 127 p. Bagnall, K. ¥. (l972)» The Actinide Elements. Elsevier Publishing Co. Bathen, K. H. (l97l)» Spatial and temporal variations in the depth of the mixed layer in the North Pacific Ocean. Look Lab/Hawaii 2, 3—23* Baturin, G. N. (1970), Uranium in oceanic ooze solutions of the Southeastern Atlantic. Doklady Akad. Nauk SSSR, 198, 224-226. Baturin, G. M. and Kochenov, A. V. (1973), Uranium in interstitial waters of marine and oceanic sediments. Geochem. International 10, 1529—1536. Bell, T. H. (l97*+), Vertical mixing in the deep ocean. Nature, 2 51, *+3—44. ________ (l975a), Lee waves in stratified flows with simple hormonic time dependence. J. Fluid Mech. 67, 705-722. _______ (1975b), Topographically generated internal waves in the deep ocean. J. Geophys. Res. 80, 320-327. Bender, M. L. and Kaufman, A. (1969), Th^'^/Th^'^ and Th232 concentrations of Southeast Pacific Ocean water. Annual Report to A.E.C., Contract AT (30—l), 3139. 225 Benninger, L. K* (1976), The uranium-series radionuclides as tracers of geochemical processes in Long Island Sound. Ph.D. dissertation, Yale University, 151 P* Berger, ¥., Adelseck, C., and Mayer, A. (1976), Distribu tion of carbonate in surface sediments of the Pacific Ocean. J. Geophys. Res. 81, 2617-2627. Bernat, M. and Goldberg, E. D. (1969), Thorium isotopes in the marine environment. Earth Plant. Sci. Lett. 5, 308-312. Bernat, M., Bieri, R. H., Koide, M., Griffin, J. J., and Goldberg, E. D. (l970)> U, Th, K, and Ar in marine phillipsites. Geochim. Cosmochim. Acta 3^ 9 1053-1071* Bhandari, N. (1969)* A selective and versatile low level beta-, x-, and gamma-ray detector assembly. Nuclear Xnstru. Methods 679 251-256. Bhandari, N., Bhat, S. G., Krishnaswami, and Lai, D. (l97l)9 A rapid beta-gamma coincidence technique for determination of natural radionuclides in marine deposits. Earth Planet. Sci. Lett. 11, 121-126. Bhat, S. G., Krishnaswami, S., Lai, D., Rama, and Moore, ¥. S. (1969a), Th-234/U-238 ratios in the ocean. Earth Planet. Sci. Lett. 5» 483-^91* Bhat, S. G. and Krishnaswami, S. (1969b), Isotopes of uranium and radium in Indian rivers. Proc. of Indian Acad. Sci. LXXX, Sec. A, 1-17* Bostrom, K., Joensuu, 0., and Brohm, I. (l97^)> Plankton: its chemical composition and significance as a source of pelagic sediments. Chem. Geol. l4, 255—271* Bowden, K. F. (196^), Turbulence. In Oceanography and Marine Biology Ann. Review (editor,H.Barnes) , v. 2, p. 11-30, Publishers George Allen and Unwin, Ltd. Bowen, V. T. and Roether, W. (1973)9 Vertical distribu tions of Sr-90, Cs-137* and tritium near 45°N in the Atlantic. J. Geophys. Res. 78, 6277-6285* Brewer, P. and Spencer, D. (1975)> Minor element models in coastal waters. In Amer. Chem. Soc. Symposium 18 (editor, T. Church), p. 8O-96. 226 Broecker, ¥. S. (1963)* Radioisotopes and large-scale oceanic mixing. Xn The Sea (editor, M. N. Iiill), v. 2, p. 88-108, Xnterscience• ________ (1963), An application of natural radon to problems in oceanic circulation. In Symposium on Diffusion in Oceans and Fresh Waters, p. 116-1^5, L.D.G.O., Palisades, N. Y. ______(1966), Radioisotopes and the rate of mixing across the main thermocline of the ocean. J. Geophys. Res. 71, 3827-3836. Broecker, ¥. S., Li, Y.~H., and Cromwell, J. (1967)* Radium—226 and radon-222 concentrations in the Atlantic and Pacific Oceans. Science 138* 1307-1310. Broecker, ¥. S., Cromwell, J. and Li, Y.—H. (1968), Rates of vertical eddy diffusion near the ocean floor based on measurements of excess Rn-222. Earth Planet. Sci. Lett. 3, 101-103. Broecker, ¥. S., Kaufman, A., Ku, T.—L., Chung, Y.-C., and Craig, H. (l970a), Ra- 226 measurements from 1969 North Pacific GEOSECS Station. J. Geophys. Res. 75, 7682-7685. Broecker, ¥. S. and Kaufman, A. (197Gb), Near surface and near bottom radon results from the 1969 North Pacific GEOSECS Station. J. Geophys. Res. 75, 7679-7681. Broecker, ¥. S. (l97Gc), Interchange of water between the major oceans. J. Geophys. Res. 75, 35^5-3552. Broecker, ¥. S. and Peng, T.-H. (l97l), The vertical distribution of radon in the BOMEX area. Earth Planet. Sci. Lett. 11, 99-108. Broecker, 17. S., Kaufman, A., and Trier, R. M. (1973), The residence time of thorium in surface seawater and its implications regarding the fate of reactive pollutants. Earth Planet. Sci. Lett. 20, 35-^. Bruland, K. ¥. , Koide, M. , and Goldberg, E. D. (197^0, The comparative marine geochemistries of Pb—210 and Ra—226. J. Geophys. Res. 79, 3083—3086. Cairns, J. L. and Nelson, K. ¥. (l970), A description of the seasonal thermocline cycle in shallow coastal water. J. Geophys. Res. 75, 1127-1131. 227 Chaiken, P. L., Goluber, N. U., and Zherekhov, V. G. 41962), Determination of* MsThI by MsThll jB-particles in the presence of Ra-226. Soviet Radiochem. 4, 88-92. Cherry, R. D., Gericke, I. H., and Shannon, L. V. (1969)* Th-228 in marine plankton and seawater. Earth Planet. Sci. Lett. 6, 451-456. Cherry, R. D. and Shannon, L. V. (l97^)» The -radio activity of marine organisms. Atomic Energy Review 12, 3-^5. Cherry, R. D., Fowler, S. W., Beasley, T. M., and Heyrand, M. (1975)» Po—210: its vertical oceanic transport by zooplankton and metabolic activity. Marine Chem. 3> 105-110. Chung, Y.-C. (l97l)> Pacific deep and bottom water studies based on temperature, radium, and excess radon measure ments. Ph.D. dissertation, U.C. San Diego. Chung, Y.—C, and Craig, II. (1972), Excess radon and temperature profiles from the eastern equatorial Pacific. Earth Planet. Sci. Lett. 14, 55—64. , (l973a) 9 Radium—226 in the eastern equatorial Pacific. Earth Planet. Sci. Lett. 17> 306—318. Chung, Y.-C. (l973h), Excess radon in the Santa Barbara Basin. Earth Planet. Sci. Lett. 179 319-323# _______ (1974a), Transient excess radon profiles in Pacific bottom water. Earth Planet. Sci. Lett, 21, 295-300. _______ (I97^b), Ra—226 and Ra—Ba relationships in Antarctic and Pacific waters. Earth Planet. Sci. Lett. 23> 125- 135# Chung, Y.-C., Craig, H., Ku, T.-L., Goddard, J., and Broecker, ¥. S. (l97^c), Ra—226 measurements from three GEOSECS intercalibration stations. Earth Planet. Sci. Lett. 23, 116-124. Church, T. M, and Bernat, M. (1972), Thorium and uranium in marine barite. Earth Planet. Sci. Lett. 14, 139- 144. 228 Craig, H. (1969)* Abyssal carbon and radiocarbon in the Pacific. J. Geophys. Res. jk9 5^91-5506. Craig, H. , Chung, Y.-C., and Fiadiero, M. (1972), A benthic front in the south Pacific. Earth Planet. Sci. Lett. 16, 50-65. Craig, H. , Krishnaswami, S., and Somayajulu, B. L. K. (1973), Pb—210 -Ra-226: radioactive disequilibrium in the deep sea. Earth Planet. Sci. Lett. 17* 295-305* Craig, H. (197^0* A scavenging model for trace elements in the deep sea. Earth Planet. Sci. Lett. 23, 1^9-159* Crank, J. (1956), Mathematics of Diffusion. Oxford Univer sity Press. Crowe, F. J. and Schwartzlose, R. A. (1972), Release and recovery records of drift bottles in the California Current region, 1955—1971* CalCOFI Atlas No. l6, 1^0 p. Dawson, R. and Duursma, E, (197M 9 Distribution of radio isotopes between phytoplankton, sediment, and seawater in a dialysis compartment system. Netherlands J. Sea Res. 8, 339-353* Duursma, E. (1972), Geochemical aspects and applications of radionuclides in the sea. Xn Oceanogr. Mar. Biol. Ann. Rev, (editor, H. Barnes), v. 10, p. 137-223* Edgington, D. N., Gordon, S. A., Thommes, M. M., and Almodovar, L. R. (1970), The concentration of Ra, Th, and U by tropical marine algae. Limnol. Oceanogr. 15* 9^5-955. Edmond, J. M. (1970), Comments on the paper by Ku et al., Ra in the Indian-Antarctic Ocean south of Australia. J. Geophys. Res. 75* 5286. _______ (l97^)> On the dissolution of carbonate and silicate in the deep ocean. Deep Sea Res. 21, ^55—^80. Eittrem, S., Biscaye, P., and Amos, A. (l975)> Benthic nepheloid layers and the Ekman thermal pump. J. Geophys. Res. 80, 506l—5067- Emery, Iv. 0. (i960), The Sea off Southern California: a modern habitat of petroleum. J. Wiley and Sons, Inc. 229 Feely, H. ¥. , Trier, R. M. , and Kent, C. A. (1973) , First progress report on the analysis of* GEOSECS samples for Ra—228 and Th-228. Contract no. NSF-Gx-28l64-c. Fenyves, B. and Haiman, 0. (1969), The Physical Principles of Nuclear Radiation Measurements. Academic Press. Folsom, T, R. and Beasley, T. M, (1972), Contributions from the alpha—emitter, Po-210, to the natural radiation environment of the marine organisms. In Radioactive Contamination of the Mar. Environ., IAEA Symposium, 625-631, Vienna. Friedlander, G,, Kennedey, J., and Miller, J. (1964), Nuclear and Radiochemistry. John Kiley and Sons, Inc. Garrett, C. and Munk, ¥. (1972), Oc eanic mixing by break ing internal waves. Deep Sea Res. 19, 823-832. Griggs, G. B. (1974), Nearshore current patterns along the central California coast. Estuar. Coast. Mar. Sci. 2, 395-405. Heye, D. (1969), U, Th, and Ra in ocean water and deep sea sediments. Earth Planet. Sci. Lett, 6, 112-116. Higashi, S. (l959a), On the Th transfer from lithosphere to hydrosphere in connection with the occurrence of Th in seawater. Oceanogr. Soc. Japan J. 15, 69-75. ________ (1959b), Estimation of microgram amount of Th in seawater. Oceanogr. Soc. Japan J. 15, 65-68. Hodge, V., Hoffman, F., Foreman, R., and Folsom, T. (1974), Simple recovery of Pu, Am, U, and Po from large volumes of seawater. Anal. Chem. 46, 1334-1336. Hunkins, K. L. (1974), Subsurface eddies in the Arctic Ocean. Deep Sea Res. 21, 1017-1033. IAEA (1967), Radioactive Dating and Methods of Low-Level Counting. Proc. of Symposium at Monaco, Vienna, _______ (1970), Reference Methods for Marine Radioactivity Studies. Technical Reports Series no. 118, Vianna. _______ (1973a), Environmental Behavior of Radionuclides Released in the Nuclear Industry. Proc. of Symp. at Aix-en-Provence, Fr., Vienna. _______ (1973L), Radioactive Contamination of the Marine 230 Environment. Proc, of Symp. at Seattle, Vienna, Imai, T# and Sakanove, M. (1973), Content of Pu, Th, and Pa in seawater and recent coral in the north Pacific, Oceanogr. Soc. Japan J. 29, 76—82. Jackson, ¥. M. , Gleason, G. X,, and Hammons, P. J. (1970), Distribution Studies between TTA in MIBK and aqueous solutions using radiotracers of Group II and III elements. Anal. Chem. 42, 1242-47* Jackson, ¥. M. and Gleason, G. I. (1973), Distribution studies of Ra and other metallic elements between TTA in MIBK and aqueous solutions. Anal. Chem. 45, 2125- 2129. Jones, J. II. (1973), Vertical mixing in the equatorial undercurrent. J. Phys. Oceanogr. 3, 286-296. Joshi, L. U. and Ganguly, A. K. (1972), Anamolous Th-228/ Th-232 and Th-230/Th-232 activity ratios in backwater sediments along the west coast of India. In Radio active Contamination of the Marine Environment IAEA Symp. at Seattle, 747-755, Vienna. Kaufman, A. (1969a), The Th—232 concentration of surface ocean water. Geochim. Cosmochim. Acta 33, 717-724. ______________ (1969b), The distributions of Ra-228 in the oceans. Annual Report Contract AT (30—l), 3139* Kaufman, A., Trier, R. M., Broecker, ¥. S., and Feely, H. ¥. (1973), The distribution of Ra—228 in the world ocean. J. Geophys. Res. 78, 8827—8848. Kharkar, D. P., Thomson, J., Turekian, K. K., and Forster, ¥. 0. (1974), Natural radioactivity in plankton samples from the Caribbean. Preprint submitted to J. Geophys. Res. Kirby, II. ¥. (1954a), Decay and growth tables for the naturally occurring radio-active series. Anal. Chem. 26, 1063-1071. _______ (1954b), The analytical chemistry of actinium. In Progress in Nuclear Energy (editor, S. Elion), Series IX, vol. 8, pt. 1, p. 97-139* Kitano, Y. and Oomori, T. (l97l), Coprecipitation of U with CaCO^. Oceanogr. Soc. Japan J. 27, 34-42. 231 Knauss, K. G. , Ku, T.-L., and Moore, W. S. (1975), Radium and thorium isotopes in the surface water of the eastern equatorial Pacific. Trans. Am. Geophy. Un. 56, 1001. Knauss, K. G. and Ku, T.-L. (1976), Radium and thorium in coastal waters of the east Pacific. Trans. Am. Geophy. Un. 57, 255. Koczy, F., Picciotto, E., Poulaert, G., and ¥ilgain, S. (1957), Mesure des isotopes du Th dans l'eau de mer. Geochim. Cosmochim. Acta 11, 103-129. Koide, M., Griffin, J. J., and Goldberg, E. D. (1975), Records of Pu fallout in marine and terrestrial samples. J. Geophys. Res. 80, 4l53-4l62. Kraus, K. A. and Holmberg, R. ¥. (1954), Hydrolytic behavior of metal ions. Ill Hydrolysis of thorium (iv ). J. Phys. Chem. 58, 325-330. Krestov, G. A. (1963), Thermodynamic characteristics of atoms and ions of rare earths and actinide elements. Soviet Radio-chem. 5, 229-239# Krishnaswami, S., Lai, D., and Somayajulu, B. L. K. (1972a), Trace elements in marine environment: ex traction techniques and results. J. Mar. Biol. Ass. India l4, 372-383# Krishnaswami, S., Lai, D., Somayajulu, B. L. K., Dixon, F. S., Stonecipher, S. A., and Craig, H. (1972b), Si, Ra, Th, and Pb in seawater: in situ extraction by synthetic fiber. Earth Planet. Sci. Lett. 16, 84—90. Ku, T.-L. (1966), Uranium series disequilibria in deep sea sediments. Ph.D. dissertation, Columbia U. Ku, T.-L., Li, Y.-H., Mathieu, G. G., and Wong, H. K. (1970), Radium in the Indian-Antaretic Ocean south of Australia. J. Geophys. Res. 75, 5286-5292. Ku, T.-L., Knauss, K. G., and Mathieu, G. G. (1974), Uranium in open ocean: concentration and isotopic composition. Trans. Am. Geophy. Un. 55, 3l4. Kupferman, S. L. (l97l), Cesium-137 iri the north Atlantic measured by selective adsorption in situ. J. Mar. Res. 29, 11-18. 232 Kuznetsov, Y. , Legin, V., Lisitsyn, A., and Siraonyak, Z. (1964), Radioactivity of ocean suspensions. Radiokhimiya 6, 242—254. Kuznetsov, Y., Elizarova, A., and Frenlikh, M. (1966a), Investigation of mechanism of precipitation of Pa-231 and Th—230 from ocean waters. Radiokhimiya 8, 459— 468. Kuznetsov, Y., Simonyak, Z., Elizarova, A., and Lisitsyn, A. (1966b), Content of Pa and Th isotopes in ocean water. Radiokhimiya 8, 455-458. Kuznetsov, Y., Elizarova, A., Lisitsyn, A., and Faizullin, F. Z. (1973), The role of ion exchange in the removal of Ra from ocean water into sediments. Soviet Radio— chem. 14, 441-446. Lai, D. and Schink, D. R. (i960), Low—background thin— wall flow counters for measuring beta-activity of solids. Rev. Sci. Instru. 31, 395-398. Lai, D., Rajogopalen, G., and Rama (1967)9 Sensitive and descript J3 and ^ / counting assemblies. In Radio active Dating and Methods of Low-Level Counting IAEA Symp. at Monaco, 615-627, Vienna. Li, Y.-H, Ku, T.-L., Mathieu, G. G., and Wolgemuth, K. (1973), Barium in the Antarctic Ocean and implications regarding the marine geochemistry of Ba and Ra-226. Earth Planet. Sci. Lett. 19, 352-358. Li, Y.-H., and Gregory, S. (1974), Diffusion of ions in seawater and in deep sea sediments. Geochim. Cosmochim. Acta 38, 703-714. Lietzke, T. A. and Lerman, A. (1975), Effects of bottom relief in two-dimensional oceanic eddy diffusion models. Earth Planet. Sci. Lett. 24, 337-344. Loganthan, P. and Burau, R. G. (1973), Sorption of heavy metal ions by a hydrous manganese oxide. Geochim. Cosmochim. Acta 37, 1277-1293. Lonsdale, P. (1976), Abyssal circulation of the southeast Pacific and some geologic implications. J. Geophys. Res. 81, H 63- H 76. Mantyla, A. W. (l975), On the potential temperature in the abyssal Pacific Ocean. J. Mar. Res. 33, 341-354. 233 Martin, J. H# and Knauer, G. A. (l973)> The elemental composition of* plankton. Geochim. Cosmochim. Acta 379 1639-1653. Matsumoto, E. (1975)» Th-234/U-238 radioactive dis equilibrium in the surface layer of the ocean. Geochim. Cosmochim. Acta 39> 205-212. Mauchline, J. and Templeton, W. L. (1964), Artificial and natural radioisotopes in the marine environment. In Oceano^r. Mar. Biol. Ann. Rev, (editor, H. Barnes), vol. 2, p. 229-279. Miyake, Y., Saruhashi, K., Katsuragi, Y., Kanazawa, T., and Sugimura, Y. (1964), U, Ra, Th, Io , Sr-90, and Cs-137 in coastal waters of Japan. In Recent Research in the Field of Hydrosphere, Atmosphere, and Nuclear Geochemistry^p.. 128-l4l,Maruzen. Miyake, Y., Saruhashi, K., and Sugimura, Y. (1968), Bio- geochemical balance of natural radioactive elements in the oceans. Rec. Oceanogr. Wks• Japan 9* 179-187. Miyake, Y., Sugimura, Y., and Mayeda, M. (l970a), The U content and activity ratio U—234/U-238 in marine organisms and seawater in the Western North Pacific. Oceanogr. Soc. Japan J. 26, 123—129. Miyake, Y., Sugimura, Y., and Yasujima, T. (l970b), Th concentrations and the activity ratios Th—230/Th—232 and Th-228/Th-232 in seawater in Western North Pacific. Oceanogr. Soc. Japan J. 26, 130—136. Miyake, Y., Saruhashi, Y., and Sugimura, Y. (1973a), The isotopic ratio U-234/U-238 in seawater and its bearing on the isotopic ratio in river waters. Rec. Oceanogr. Wes. Japan 12, 23-2 5. ________ (1973b), The excess of Th-228 over Th—232 in ocean water. Rec. Oceanogr. Wks. Japan 12, 27-30. Miyake, Y. and Sugimura, Y. (1975a), Pu content in Pacific Ocean waters. Presented at Trans—Uranium IAEA meeting in San Francisco, IAEA—SM—199/22. Miyake, Y., Shimada, T., Sugimura, Y., Shigehara, K., and Saruhashi, K. (l975b), Distribution of tritium in the Pacific Ocean. Rec. Oceanogr. Wks. Japan 13» 17-32. 234 Monsecour, M., DeRegge, P., and Demidlt, A. (1973), Preparation of highly purified Ac-228 tracer solutions, Radiochem, Radioanal. Lett, l4, 365—371* Moore, ¥. S, and Sackett, W. M, (1964), U and Th in- equilibrium in seawater, J, Geophys, Res, 69, 5^01- 5405. Moore, ¥. S, (1967), Amazon and Mississippi River concentra tions of U, Th, and Ra isotopes. Earth Planet, Sci, Lett. 2, 231-233. (1969a), o ceanic concentrations of Ra—228. Earth Planet. Sci. Lett, 6, 437—446. (1969b), Measurement of Ra—228 and Th-228 in sea water. J. Geophys. Res. 74, 694-704. (1969c), 0 ceanic concentrations of Ra—228 and a model for its supply. Ph.D. dissertation, SUNY at Stonybrook. (l972a), Ra-228: application to thermocline mix ing studies. Earth Planet. Sci. Lett. 16, 421-422. (l972b), Oceanic mixing as determined by radio isotopes. J. Mar. Biol. Ass. India 14, 15-25. Moore, W. S. and Reid, D. P. (1973), Extraction of Ra from natural waters using MnO impregnated acrylic fibers. J. Geophys. Res. 78, 8880—8886. Moore, ¥. S. and Cook, L. M. (1975a), Ra removal from drinking water. Nature 253, 262-263* Moore, ¥. S. (l975b), Sampling Ra-228 in the deep ocean. Submitted to Deep Sea Res. Munk, ¥. H. (1966), Abyssal recipes. Deep Sea Res. 13, 707-730. Murray, J. ¥. (1975), Interaction of metal ions at MnO^- solution interface. Geochim. Cosmochim. Acta 39, 505-519. Neumann, G. (1968), Ocean Currents. Elsevier. Nikolayev, P., Lazarev, K. and Graschenko, S. (1962), Concentration of Th isotopes in the water of the Sea of Azov. Akad. Nauk SSSR Dokl. 138, 674-677. 235 Nozaki, Y. and Tsunogai, S. (l973a), Pb-210 in the North Pacific and the transport of terrestrial material through the atmosphere. Earth Planet. Sci. Lett. 20, 88-92. Nozaki, Y., Tsunogai, S., and Nishimura, M. (l973b), Pb-210 in the Japan Sea. Oceanogr. Soc. Japan J. 29* 231-236. Okubo, A. (l97l)* Oceanic diffusion diagrams. Deep Sea Res. 18, 789-802. Oldnall, R. J. (1975), Possible sources of metals in pelagic sediments: with special reference to the Bauer Depression. M.S. thesis, U. of Hawaii. Osterberg, C., Carey, A., and Curl, H. (1963), Acceleration of sinking rates of radionuclides in the ocean. Nature 200, 1276-1277. Patin, S. A. and Tkachenko, V. N. (1972), Biogeochernistry of Th isotopes in the ocean. Doklady ESS 217, 697“ 701. Peng, T-H, Takahashi, T. , and Broecker, W. S. (197^0, Surface radon measurements in the North Pacific Ocean station PAPA. J. Geophys. Res. 79, 1772-1780. Percival, D. R. and Martin, D. G. (197^), Sequential determination of Ra-226, Ra-228, Ac-227, and Th isotopes in environmental and process waste samples. Anal. Chem. 46, 1742-1749. Picer, M. and Strohal, P. (1968), The determination of Th and U in biological material by neutron activation analysis. Anal. Chim. Acta 40, 131-136. Rajagopalen, G. (1969), High sensitivity radiation count ing and its application in nuclear physics. Ph.D. dissertation, U. of Bombay. Rama, Koide, M., and Goldberg, E. D. (1961), Pb-210 in natural waters. Science 134, 98-99* Reid, J. L. and Schwartzlose, R. A. (1962), Direct measure ment of Davidson Current off central California. J. Geophys. Res. 67, 2491-2497* Reid, J. L. and Arthur, R. S. (1975), Interpretation of maps of geopotential anomaly for the deep Pacific Ocean. J. Mar. Res. Supplement 33, 37-52. 236 Reid, D. F. (l976), On the half-life of Ra-228 and the errors in oceanographic data due to the use of the 6.7 year (wrong) value. Geophys. Res. Lett. 3, 253-256. Rooth, C. G. and Ostlund, H. G. (1972), Penetration of tritium into the Atlantic thermocline. Deep Sea Res. 19, 481-492. Sackett, ¥. M., Potratz, H. A., and Goldberg, E. D. (1938), Thorium content of ocean water. Science 128, 204-205. Sakanoue, M., Nomura, T., Imai, T., Uzuyama, H. (1970), Radiochemical studies on Ra-228 in seawater and Pa-231 in sediments. In Hydrogeochemistry (editor, Sarmiento, J., Feely, H., Moore, ¥. , and Broecker, ¥. S. (1976a), Relationship between vertical eddy diffusion and density gradient. In GEOSECS proposal for 1976, LDGO, Appendix III. Sarmiento, J., Hammond, D., and Broecker, ¥. (1976b), The calculation of the statistical counting error for Rn-222 scintillation counting. Submitted to Earth Planet. Sci. Lett. Southern California Coastal ¥ater Research Project (1973), The ecology of the Southern California Bight; im plications for water quality management. SCC¥RP TR 104. Schwartzlose, R. and Reid, J. (l972), Nearshore circulation in the California Current. In Calif. Mar. Res. Comm., CALCOFI Rpt. 16, 57-65. Shannon, L. V., Cherry, R. D. (1970), Po-210 and Pb-210 in the hydrosphere. In Hydrogeochemistry (editor, Shannon, L. V., Cherry, R. D., and Orren, M. j. (1970), Po-210 and Pb-210 in the marine environment. Geochim. Cosmochim. Acta 34, 701-711. Shannon, L. V. and Cherry, R. D. (l97l), Ra-226 in marine plankton. Earth Planet. Sci. Lett. 11, 339-343* Sharpe, J. (1964), Nuclear Radiation Detectors. J. ¥iley. Sholkovitz, E. and Soutar, A. (1973), Changes in composition of bottom water of Santa Barbara Basin: effect of turbidity currents. Deep Sea Res. 22, 13-21. 237 Silker, ¥. B. (1972), Horizontal and vertical distributions of radionuclides in the North Pacific Ocean. J. Geophys. Res. 77, 1061-1070. Sill, C. ¥. and Olsen, D. G. (l970), Sources and prevention of recoil contamination of solid-state alpha detectors. Anal. Chem. k2, 1596-1607. Smayda, T. S. (1970), The suspension and sinking of phyto plankton in the sea. Xn Oceanogr. Mar. Biol. Ann. Rev. (editor, H. Barnes), vol. 8, 353-^1^. Smith, R. L. (1968), Upwelling in Oceanogr. Mar. Biol. Ann. Rev. (editor, II. Barnes), vol. 6, 11—^6. Somayajulu, B. L. K. and Goldberg, E. D. (1966), Thorium and uranium isotopes in seawater and sediments. Earth Planet. Sci. Lett. 1, 102-106. Stommei, H. (l9^9), Horizontal diffusion due to oceanic turbulence. J. Mar. Res. 8, 199-223. Stommei, H. and Arons, A. B. (l960a), On the abyssal circulation of the world ocean - I; stationary planetary flow patterns on a sphere. Deep Sea Res. 6, 140-154. _______ (1960b), On the abyssal circulation of the world ocean — II; an idealized model of the circulation pattern and amplitude in oceanic basins. Deep Sea Res. 6, 217-233. Strickland, J. D. H. (1965), Production of organic matter in the primary stages of the marine food chain. In Chemical Oceanography (editors, J. Riley and G. Skirrow), vol. 1, Academic Press. Szabo, B. J. (1967), Radium content in plankton and sea water in the Bahamas. Geochim. Cosmochim. Acta 31, 1321-1331. Taft, B. A. and Jones, J. H. (1973), Measurements of Equatorial and Undercurrent in the East Pacific. In Progress in Oceanography (editor, B. Varren), v. 6, 47-110. Thomson, J. and ¥alton, A. (l97l), Natural radioactive decay series elements in the oceans and sediments. Proc. Roy. Soc. Edin. 72, 167-182. 238 Thomson, J#, Turkeian, K#, and McCaffney, R. (1974)* Ac cumulation of metals in and release from sediments of Long Island Sound. In K. Turekian, Ann. Rpt. 1974, Grant AT (ll-l) - 3573# Trier, R. M., Broecker, ¥. S. and Feeley, H. W. (1972), Ra-228 profile at the second GEOSSCS Intercalibration station, 1970, in the North Atlantic. Earth Planet. Sci. Lett. 16, l4l-l45. Tsuchiya, M. (1975)* Subsurface countercurrents in the East Equatorial Pacific Ocean. J. Mar. Res. Supplement 33* 145-173. Turekian, K. and Chan, L. (l97l)> The marine geochemistry of Th-230 and Pa-231. In Activation Analysis in Geo chemistry and Cosmochemistry (editors, Brunfelt and Steinmesj, p. 311-320. Turekian, K., Katz, A., and Chan, L. (l973n), Trace element trapping in pteropod tests. Limnol. Oceanogr. 18, 240-249. Turekian, K., Kharkar, D. P., and Thomson, J. (l973b), Uranium and thorium decay series nuclide abundances in marine plankton. In Final Rpt., ARPA Order no. 1973* Prog. Code No. 3F10, Contr. No. N00014-67-A-0097-0022. Watt, D. and Ramsden, D. (1964), High Sensitivity Counting Techniques. Macmillan Comp. Whitfield, M. (1975)* Extension of chemical models for seawater to include trace components at 25°C and 1 atmosphere pressure. Geochim. Cosmochim. Acta 39* 1545-1557. Wickham, J. (1975)* Observations of California Counter- current. J. Mar. Res. 33* 325—3^0. Wblberg, J. (1967)* Prediction Analysis. D. Van Nostrand. Wong, C. S. (1972), Deep zonal water masses in the Equa torial Pacific Ocean inferred from anomalous ocean ographic properties. J. Geophy. Res. 77* 7196-7202. Wong, K., Brown, G., and Noshkin, V. (1976), A rapid procedure for Pu separation in large volumes of fresh and saline water by Mn0o precipitation. Preprint submitted to J. Radioanal. Chem. 239 Wooster, W. and Reid, J. (1963), Eastern boundary cur rents, In The Sea (editor, M. Hill), vol. 2, John Wiley and Sons. Wooster, W. and Guillen, 0. (1974), Characteristics of* El Nino in 1972. J. Mar. Res. 32, 387-404. Wyrtki, K. (1966), Oceanography of the eastern equatorial Pacific Ocean. In Oceanogr. Mar. Biol. Ann. Rev. (editor, H. Barnes)" vol. 4, 33-68. _______ (1973), Tel econnections in the equatorial Pacific Ocean. Science 180, 66-68. Wyrtki, K., Magaard, L., and Hager, J. (1976), Eddy energy in the oceans. J. Geophys. Res. 81, 2641-2646. Yashushi, K. (l97l)> Coprecipitation of U with CaCO . J. Oceanogr. Soc. Jap. 279 34-42. Young, J. and Silker, W. (1974), Determination of air-sea gas exchange and oceanic mixing rates using Be-7 dur ing Bomex Experiment. J. Geophys. Res. 799 4481— 4489. Zaneveld, J., Pak, H., and Plank, W. (1973), Optical and hydrographic observations of the Cromwell Current between 92°W and the Galapagos Islands. J. Geophys. Res. 78, 2708-2714. 240

Asset Metadata

Creator Knauss, Kevin Gibbons (author)

Core Title Natural decay series isotopes in surface waters, bottom waters, and plankton from the East Pacific

Contributor Digitized by ProQuest (provenance)

Degree Doctor of Philosophy

Degree Program Geological Sciences

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

Tag geochemistry,oai:digitallibrary.usc.edu:usctheses,OAI-PMH Harvest,physical oceanography

Language English

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c29-343867

Unique identifier UC11221028

Identifier DP28545.pdf (filename),usctheses-c29-343867 (legacy record id)

Legacy Identifier DP28545.pdf

Dmrecord 343867

Document Type Dissertation

Rights Knauss, Kevin Gibbons

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