The Natural Radioactive Isotopes of Beryllium in the Environment

4t 5 I

THE NATURAL RADIOACTIVE ISOTOPES OF BERYLLIUM IN THE ENVIRONMENT

CONVEN'ERS* KARL K. TUREKIAM ('YALE) H,ERBm \FOLJCHOK ('E.Mt.ME ) # - .. DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof. DISCLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document. 4

The following pages are an exact representation of what is in the original document folder. THE NATURAL RADIOACTIVE ISOTOPES OF BERYLLImI IN THE ENVIRONFENT

Abstracts and a report of a meeting sponsored by Yale University and the U.S. Department of Energy

Conveners: Karl K. Turekian (Yale) and Herbert L. Volchok (Ern, DOE)

A considerable amount of work has been done to date on the environmental pathways of the two cosmogenic radioactive isotopes of beryllium, 7~e(53 day half-life) and ''Be (1.5 x lo6 year half-life). 7~ehas pro'ven useful as an index of aerosol scavenging over the oceans and a tool for the measurement of bioturbation rates in lake and estuarine sediments. ''Be has been of value in determining rates of accumulation of deep-sea sediments and manganese nodules. Both nuclides are produced in the stratosphere and troposphere by cosmic rays. The rate of production of each nuclide, the rate and locus of its entry into the troposphere, and the rate and locus of its delivery to the earth's surface can be determined for each nuclide. This is of value in helping us understand the ultimate disposition of other nuclides injected into both the troposphere and the stratosphere. In addition, determining the mechanism of removal of beryllium isotopes from the marine system provides one more nuclide useful as a model of the behavior of reactive nuclides introduced into the sea. '

The usual methods of counting 7~eand ''Be have been by tedious radiochemical separations followed by either gamma (7~e)or beta (l 'Be) counting. Recently two methods have greatly improved the possibility of increasing the quality and quantity of data for both these nuclides. 7~eis determinable, in manv cases, by.high resolution, high efficiency lithium germanium detectors - made even more useful with new computer spectrum analysis techniques. ' 'Be can now be measured by accelerators (cyclotron or Van de Graaff) thereby reduc- ing background uncertainties as well as requiring smaller amounts of sample for analvsis.

We felt that a conference on 7~eand 1°Be in the environment was appro- priate at this time for the reasons implicit in the above, namely:

1. A large amount of 7~edata in air, sea water and sediments already exists. A review of the significance of these data from a number of laboratories is desirable in planning for future work.

2. The number of high quality l0Be data obtained by low level beta counting has increased dramatically in the past few years and has challenged our earlier ideas about both the atmospheric production rate and the methods of transport to marine repositories. Recently the potential of the accelerator in determining ''Be quickly and accurately has been demonstrated. A discussion of the types of geochemical and geophvsical problems to be addressed with both techniques, as well as the need for intercalibration amont the various laboratories is critical at this time. PRO GRAM - - - ______- - . -- .- THE NATURAL RADIOACTIVE ISOTOPES OF BERYLLIUM IN THE ENVIRONMENT (Sponsored by Yale University and the,U.S. Department of Energy) I Room 101 Kline Geology Laboratory, Yale University, New Haven, Conn. Monday, 0ctober 1, 1979 I 8 : 30 AM Introductory remarks t 8:40 O'Brien: Cosmogenic production of 7~eand ''Be in the atmosphere 9 : 25 Peirson: Fifteen years of atmospheric 7~emeasurements around the world 1O:lO Coffee 7 10: 30 Sanak : Be in the atmosphere 7 I 11:15 Feely : Be in surface air 12:OO LIJXCH : Dining Hall, Kline Biology Tower 1:00 PM Naidu: Sir years of 7~emeasurements in precipitation at Brookhaven 7 I I 1:45 Hussein: Determination of Belozone ratios in the stratosphere

I 2: 30 Silker and Young: 7~ein the oceans 3: 15 Coffee :3 : 35 Benninger: 7~ein the study of particle mixing rates in nearshore and lake sediments 7 4 :20' Cutshall: Be in river and estuarine sediments -5:05 Dion : 7~ein Sargasso Sea and Long Island Sound waters 5:20 Session ends

Tuesday, October 2, 1979 8: 30 AN Introductory remarks 8:40 Tanaka : 'OB~ stratigraphies in Pacific sediment cores back to four million years I 9: 25 . Raisbeck: ''Be in air and water mass stdie's 10:lO Coffee 10: 30 Ku : ''Be in the marine environment I 11:15 Parker and Cochran: . The use of a Van de Graaff accelerator in determining 1°Be in marine materials 12: 00 LUNCH 1:30 PM Finkel: 'OB~ in glacial studies 7 2:15 Volchok: Summary of Be section and discussion

3:00 ' Coffee 10 3: 20 Turekian: Summary of Be. section and discussion PARTICIPANTS

Joseph Bennett University of Washington Dept . of Oceanography, KB-10 Seattle, L'N 98195

Larry K. Benninger Dept. of Geology University of North Carolina Chapel Hill, NC 27514

J. Kirk Cochran Dept. of Geology and Geophysics Yale University New Haven, CT 06520

Worman H. Cutshall Bldg. 3504 Oak Ridge National Laboratory Oak Ridge, TX 37830

Eric P. Dion Dept. of Geology and Geophysics Yale University New Haven, CT 06520

Vincent A. Dutkiewicz X.Y. State Dept. of Health Div. of Laboratories and Research Empire State Plaza D535 Albany, NY 12201

Environmental Measurements Laboratory U.S. Department of Energy 376 Hudson St. New York, fu'Y 10014

Robert Finkel Scripps Institution of Oceanography S-002 Universitv of California, San Diego La Jolla, CA 92093

David X. Edgington Center for Great Lakes Studies University of Wisconsin P .O. Box 413 Xlwaukee , IJis . 53201

Liaqnat Husain N.Y. State Dept. of Health Div. of Laboratories and Research Empire State Plaza Albany, NY 12201

I<. J . Jenkins Woods Hole Oceanographic Institution (617) 548-1400 IJoods Hole, Mass. 02543 ext. 2554

Michael D. Krom Dept. of Geology and Geophysics Yale University New Haven, CT 06520

Dept. of Geological Sciences University of Southern California Los Angeles, CA 90007 Augusto Mangini Dept . of Geology and Geophysics (203) 436-0372 Yale University New Haven, CT 06520

J.R. Naidu Safety and Environmental Protection (516) 345-4263 Bldg. 535A Brookhaven National ~aborator~ Upron, NY 11973

Keran O'Brien Environmental Measurements Laboratory (212) 620-36332 U.S. Department of Energy 376 Hudson St. New York, NY lOOlh

Charles Osterberg Office of Health and Environment U.S. Department of Energy Vashington, D.C. 20SLt-5

Peter D. Parker Wright Nuclear Structure Laboratory (203) 436-2320 Yale University New Haven, CT 06520

Douglas H. Peirson Environmental and Medical Science Division Abingdon (0235) Harwell, Oxon 24141 Ext. 4160 United Kingdom

Alex Pszenny Box 3, Graduate School of Oceanography (401) 792-6256 university of Rhode Island Kingston, RI 02881

Grant M. Raisbeck Laboratoire Ren4 Bernas Centre de SpectromBtrie Nuclgaire et de Spectrom6trie de Masse 91406 Orsay, France

John A. Robbins Great Lakes Research Division University of Michigan Ann Arbor, Michigan 48109

Ali Rusheed Environment Health Center Div. of Laboratories and Research N.Y. State Department of Health Empire State Plaza Albany, NY 12201

Joseph Sanak Centre des Faibles Radioactivites C.N.R.S. 91190 Gif-sur-Yvette, France

Wyatt B. Silker Battelle-Pacific Northwest Laboratory Box 999 Richland, IJN 99352

J.R. Southon Tandem Laboratory (416) 525-911 GSB 105, McMaster University Ext. 4046 Hamilton, Ontario, Canada Shigeo Tanaka Institute for Nuclear Study University of Tokyo Tanashi, Tokyo 188, Japan

Karl K; Turekian Dept . of Geology and Geophysics Yale University New Haven, CT 06520

Herbert L. Volchok Environmental Measurements Laboratory U.S. Department of Energy

' 376 Hudson St. ,, New York, h9' 10014

James A. Young Battelle-Pacific Northwest Laboratory 329 Bldg., P .O. Box 999 Richland, I\% 99352 ABSTRACTS

These abstracts and summaries are for information only and this report should not be cited in referencing. Permission to quote data that are not yet in print should be obtained from the author of the abstract and should then be cited either as "Personal communication" or "j.n press" (in a particular journal or citable technical report series). Keran 0 'Brien Cosmogenic production of 7~eand ''Be in the atmosphere D.H. Peirson and R.S. Cambrav Beryllium-7 in surface air Joseph Sanak and Ggrard Lambert 7~eand 32~in the Antarctic atmosphere

Herbert LI. Feely, Herbert L. Volchok and Colin Sanderson Seasonal variations in the concentration of Be-7 in surface air J.R. Kaidu, A.F.. ?loorthy and A.P. Hull '~elevels in precipitation and air as measured at BhT during the vears 1971-1973 Liaquat Husain, Ali Rusheed and Vincent Dutkiewicz Determination of 7~e/ozoneratios in the stratosphere J.A. Young end F7.B. Silker Aerosol deposition velocities on the Pacific and Atlantic 9:sans 7 calculated from Be measurements S. Krishnaswami, L.K. Benninger, R.C. Aller, and K.L. Pon Damm Be-7 in the study of particle mixing rates in nearshore and lake sediments S.H. Cutshall and I.L. Larsen 7 Be in estuarine sediments Erik Aaboe. Eric Dion and Karl K. Turekian The behavior of 7~ein Sargasso Sea and Long Island Sound waters Shigeo Tanaka 10 Be stratigraphies in the Pacific sediment cores back to 4 million years G.PI. Raisbeck and F. Yiou ''Be measurements with a cyclotron and their application to the study of air and water masses in the environment Teh-Lung Ku ''Be in the marine environment K.K. Turebian, J.K. Cochran, S. Krishnaswami, K.A. Lanford, P.D. Parker and K.A. Bauer The measurement of ''Be in manganese nodules using a tandem Van de Graaff accelerator

P.D. Parker, W.A. Lanford, K. Bauer, J.K. Cochran, K.K. Turekian & S. Krishnaswami Measurements of ''Be distributions using a Tandem Van de Graaf accelerator R. Finkel ''Be in polar glaciers: A tree ring analog for the entire Pleistocene 7 COSMOGENIC PRODUCTION OF Be AND ''Be IN THE ATMOSPHERE ?? rg,.

Keran O'Brien

Environmental Measurements Laboratory U. S. Department of Energy New York, New York 10014

ABSTRACT '

The high-energy particles we call cosmic rays fill the galaxy. They are believed to originate in the explosion 'of supernova and as a by product of the activity of pulsars. These particles are mostly protons and E-particles with a small admixture of heavier elements. Cosmic rays are contained in the galaxy by its magnetic fields. Since the mean containment time of these particles is in the hundreds of millions of years, their concentration in the neighborhood of the solar system is essentially constant and their angular distribution isotropic - barring a local event such as a nearly supernova explosion as suggested by Higdon and Lingenfelter [1973].

Cosmic-ray intensities in the solar system are not constant however. Inter- planetary space is filled with the solar wind, a.plasma emitted by the sun. The electric and magnetic fields embedded in the field repel the charged cosmic rays

incident on the solar system,. The effect is as much as though there were an electric potential centered on the sun with a magnitude of several hundred million volts at a distance of one astronomical unit. The effectiveness of this repellent force is proportional to the activity of the sun as measured by, say,

the Zurich sunspot number. The magnitude of the heliocentric potential at the

.. earth's orbit ranges from about 100 MV at solar minimum to about 1000 MV during

.a.large solar maximum - such as 1958's.

Before ths cosmic rays penetrating the earth's orbit can interact with the

earth's atmosphere, they must penetrate the geomagnetic field. Particles I1 travelling in a direction vertical to the earth's surface must have momentum to I1 charge ratio (known as the rigidity) greater than 15 cos';. GV, where ), is the geomagnetic la titude to reach the atmosphere

v = 15 cosC?. C

is the vertical cutoff rigidity.

The field strength has not been constant over recorded time. Archaeo-

magnetic data indicate that the earth's magnetic field has varied sinusoidally

over the past 9000 years from 0.5 times the current value about 6000 years ago

to about 1.5 times the present value about 2000 years ago [COX, 19691. There is

evidence from excavations in the Lake Mungo area of southeastern Australia that

between 25,000 and 35,000 years ago the geomagnetic field may have reached 3 times

its current value during a geomagnetic excursion [~arbetti,19761.

Collisions of cosmic-ray particles with nuclei of the air result in the

production of energetic secondary nucleons and mesons which in turn collide with

other air nuclei, generating a cascade of hadrons through the atmosphere. After its collision with an energetic hadron, the nucleus will be in one or more

fragments.

fl Each of these collisions is called a "star" from its a,ppearance in a

photographic plate. About 1.4% of the time, one of the fragments associated

with the star will be a Belo nucleus and about 3% of the time,one of the

7 fragments will be Be .

Assuming a reasonable energy distribution of protons and a-particles in the

galaxy incident on the solar system, the energy spectra incident on the earth

was calculated in the electric field approximation described above. These . '

Ehmert potentials were .varied from zero to 1000 MV to encompass the maximum range

' of solar activity.

The geomagnetic field strength was allowed to range from 0.5 to 5 times the

current value to encompass both the "normal" range of variations as exemplified

by the last 9000 years and excursions of the Lake Mungo type.

7 10 The production rates of Be and Be were obtained by calculating the star

density as a function of altitude and geomagnetic latitude in the atmosphere.

The method of calculation was a high-energy radiation transport code designed for

the purpose [~'~rien,19791.

The combined effect of solar modulation and geomagnetic field variation on

world-wide star production is shown in Figure 1. The results are in good agreement at each modulation level with the prediction of Elsasser, Ney and Winckler [19,6]

that isotope production should go as

where M is the geomagnetic field strength given as multiples of the current value.

7 10 Be and Be production rates were obtained from these data using the

theoretical cross sections of Silberberg and Tsao [1973].

The inventories of these isotopes can be obtained by averaging out the

fluctuations produced in the production rate by variations in the solar activity 10 level and the geomagnetic field intensity. This yields 3.5 MCi of Be and 7.7 . . 7 7 FlCi of Be . Due to its short half-life, a stratospheric inventory for Be can

also be calculated. It is 4.6 MCi. REFERENCES

Barbetti, M., "The Lake Mungo geomagnetic excursion," Phil. Trans. Roy. Soc .

London Ser. A., 281, 515, 1976'.

COX, A., "Geomagnetic reversals ," Science, by237 ,' 1969.

Elsasser , W., E. P. Ney and J . R. Winckler, "Cosmic-ray intensity and '

geomagnetism, Nature, 178, 1226, 1956.

Higdon, J. C. and R. E. Lingenfelter, "Sea sediments,,cosmic rays, and pulsars,"

Nature, 246, 403, 1973.

O'Brien, K., "Secular variations in the production of cosmogenic isotopes in the

earth's atmosphere," J. Geophys . Reg., 84, 423, 1979.

Silberberg, R. and C. Tsao, "partial cross sections in high-energy nuclear

reactions and astrophysical applications, I. Targets with Z 5 28, Astrophys . J.

Suppl. Ser., 5, 315,-1973. FIGURE CAPTION

Figure 1. Worldwide star production rate as a function of M, the geomagnetic

field strength in units of the curren.t value, and the dependence

predicted by Elsasser et al. 10 I /

- ELSASSER. NEY, WINCKLER FORMULA '

1.0 -

0.1 -- I I I I I 0.I 1 .o 10 M- GEOMAGNETIC FIELD INTENSITY IN .MUUIPLES OF CURRENT VALUE Conference on "Natural radioactive isotopes of beryllium in the envirohmentfl i Yale University: 1-2 October 1979

- : NOT FOR PUBLICATIOJS

Berylliuin-7 in surface air k D H Peirson and R S Cambray AERE Harwell , UK

(Extended abstract )

The concentration of Be-7 in surface,air has been measured at.14 stations

0 0 located between 70 h' and 65 S, in some cases since 1964.

Sampling stations

Station

I ( Tromso, Korway ! Lem'ick, Shetland Eskdalernuir,. Dunfriesshire I / Orfordness, Suf folk i Milford Haven, Pembrokeshire I Chilton, Oxon Gibra1 t ar Hong Kong Singapore Darwin, Australia Pretoria, South Africa Aspendale, Australia Ohakea, New Zealand ! Argentine Ia. Antarctica

Airborne dust is sampled continuously at one metre above ground through

filters currently of polystyrene fib,re, previously of esparto grass. The flow

rate corresponds to about 2 tonne/day and the filters are changed weekly. The

be.ryllium-7 content is measured by gamma-ray spectrometry. The results

reported in this note are derived from the measurements of fission products from

'nuclear weapon fallout(1), in which the contribution of Be-7 was dete,rmined

incidentally. The data have been swarised and presented graphically to illustrate:

(a) seasonal vari ation

(b) latitude effects, and

(c) variation with time ~1 Seasonal variation (average 1964-78) Chi lton: on average peaks in May (as for Cs-137): peak about 40%

I above trough* which implies a reduction of 2 or 3 times

compared with the production rates in the stratosphere

relative to the troposphere as calculated by La1 (2).

spring pe,&s are variable eg negligible in 1973, prominent

in 1974.

Gibraltar: peak in July, larger amplitude than Chilton.

Hong Kong: pronounced trough during southerly wind regime in mid-year

when sun is overhead.

Pretoria: peak in September* when sun is at equator: possible fine structure

of subsidiary peaks.

Aspendale: peak in January: sub$idiar;\- peak in ~e~ternber-0ctober

(cf Hicks (5))

Latitude effect (annual means 1978)

Results are available in 1978 for all stations. %in peaks at

about 35O~orthand South, of amplitude about twice the level

at equator: slightly lower peak in sduthern hemisphere consistent

with transfer of less air from stratosphere to troposphere.

Variation with time (annual means 1964-78)

UK stations: factor of two difference between 51' and 60'~: similar patterns

at Eskdalernuir and Lerwick.

Overseas stat ions: no discernible pattern.

September 1979 I

References .

1. H S Cambray et al, Radioactive fallout in air and rain: resu1.t~to end of 1978. AERE Report No. AERErR9441 (1979).

2. D La1 and H E Suess, The radioactivity of the atmosphere and hydrosphere, Ann.Rev. of Nucl.Sci. 18, 407, (1968).

3. B B Hicks and H S Goodman, Seasonal and latitudinal variations of

, atmospheric radioactivity along Australia's east coast (150'~ longitude). Tel lus -29, 182 (1977).

SEASONAL , VARIATION Be-? in surface air at Chilton (57"N): Average monfhly concentra fions 7964 -78.

J M M J S N Month

SEASONAL VARIATION Be-7 in surface air at Hung kong (22'N): Average monthly concenfra f ions 796 4 -78.

I I I I4 I r?l I J I Is 1 r;l Month

SEASONAL + VARIATION L A TI TUDE VARIATION' WITH TIME Be-7 in surface air at United Kingdom station3: Annual mean concentrations r time.

7~eand 32~IN THE ANTARCTIC ATMOSPHERE

Joseph SANAK and Gerard LAMBERT Centre des FaibZes Radioactivitbs Laboratoire mixte CNRS-CEA Domaine du CNRS 91190 - Gif-sup-Yvette (France)

Since March 1977 the. atmospheric concentration of 7~eaerosols has been monitored daily at Dumont-Durvil le station (Terre Adel ie - 66'40 South, 140' East). These data have been completed after January 1978 by 32~measurements. These two nuclides appear to be in good correlation (r = 0.9) . 3 Figure 1 shows mean monthly 7~econcentrations of about 0.4 dpmlm and therefore 1.5 time higher than at the South Pole (W. Feely. L.E. Tonkel , and R.J. Larsen (EML), W.M. Maenhaut, W.H. Zoller and D.G. Coles, 1979). This result could be explained by the proximity of Dumont Durville to the geomagnetic pole. A small seasonal variation can be observed in the same figure with maxima during the Austral Summer. This last effect is much more marked in the case of 32~(figure 2) whose mean annual concentration is about 6 x 10-3 dpm/m3. 32 More interesting conclisions can be drawn from the 7~e/ P activity ratios. In fact this figure is approximately equal to the ratio of the production rates in the stratosphere, i.e. about 100 (La1 and Peters, 1957). On the opposite in the troposphere, the scavenging of the aerosols does not enable these nuclides to reach the radioa tive equilibrium : a tropospheric air mass is therefore characterized by a ?Be/32P activity ratio comprised between 25 and 40 (Fig. 3).

Figure 4 shows that the mean monthly ratio is always greater.than 40 at Dumont Durville. This result can only be explained by a stratospheric ori~in of 7~eand 32P, measured at sea level, in this station.

Theoretical considerations show that, in the case of stratospheric air masses injected into the troposphere, the 7~e/32~activity ratio should be greater than the stratospheric limit of 100, and reach even about 239 (Fig.5). This last value has been actually observed a few times during our 2 year records.

Good negative correlations observed between 7~econcentrations and both 222~nconcentrations and air humidity show that stratosphere to troposphere injections of spa1 lation products happen at Polar latitudes . Thi s resul t wo\!l d mean that most of the stratospheric material s detected in Antarctica are not transported at low altitude after an injection into the troposphere at mid-latitudes, but rather directly exchanged over the Antarctic continent.

FEELY W.. TONKEL L.E. and LARSEN R.J., E.?I.L. 353, April 1, 1979.

LAL D. and PETERS B., Cosmic ray produced radioactivity on the Earth. In En~~clopaedi'aof Physics, vol. 46/2. Ed. K. Sitte (Springer Verlag, New York) 551-612, 1967.

MAENHAUT W., ZOLLER W.H. and COLES D.G., Radionuclides in the South Pole atmosphere, J. Geophys. Res., 84, 3131-3138, 1979.

r P FIGURE 1

Mean monthly 7 Be activity at Dumont Durville FIGURE 2

Mean monthly 32~activity at Durnont Durville

DRYS AFTER CLEANING

3 2 7~e/P activity ratio in an air mass init.ially free of spallation products :

A - Cli thout aerosols scavenging B - For aerosols half-life of 30 days C - For aerosols half-1 ife of 5 days.

Seasonal Variations in the Concentration of Be-7 in Surface Air

Herbert W. Feely, Herbert L. Volchok and Colin Sanderson U. S. Dept. of Energy Environmental Measurements Laboratory, 376 Hudson Street, New York, N. Y. 10014

7 We have measured concentrations of Be in monthly filter samples of surface

air at a number) of sites in the Western Hemisphere since 1970. The seasonal

variations in concentration at these sites show a number of different patterns which appear to reflect regional tropospheric meteorology.

7 In general, higher concentrations of Be are found at mid-latitude sites than

at high latitude or low latitude sites. Concentrations are generalry higher at I1 high altitude sites than at sites at lower elevations, and at continental sites than at sites under a marine influence. Mean annual concentrations range from 3 3 205 fCi/m at Mauna Loa Observatory, Hawaii (3401 m. elevation) to 35' fCi/m 0 at" Punta Arenas, Chile (53 S). . .

The Arctic sites, Nord, Thule and Kap Tobin, Greenland and Barrow, Alaska show minimum concentrations during the summer quarter, June to August. Perhaps 7 Be is transported from the stratosphere or upper troposphere to the lower I troposphere most effectively at mid-latitudes, and spreads poleward and equator- ward within the lower troposphere from there. Observations of the Arctic haze layer have suggested that transport to the Arctic from mid-latitudes occurs most

efficiently during the winter and least efficiently during the summer. It 7 appears likely that the variations in Be concentrations in the Arctic result

from these same variations in tropospheric circulation. In the Antarctic, at the South Pole Station, the highest concentra- 7.. tions of Be occur during the summer quarter, December to February, and the lowest occur during the winter. This contrast with the pattern of seasonal variations in the Arctic may be related to the high elevation (2800 m.) of 7 the South Pole Station. These variations in Be concentrations are the opposite of reported variations in ozone concentration, which reaches its highest concentrations at South Pole Station during the winter. his difference indicates that the source of the winter peak in. ozone is not,. as has at times

' been suggested, direct transport of air from the polar stratosphere to the earth's surface during the southern polar winter.

Most of our sites at temperate latitudes in North America sh0w.a broad maximum extending through the spring and summer. , In part this probably results from increased transport of air from the stratosphere into the troposphere during the spring, and in part from a minimum in the stability of the troposphere during the warm months. This pattern is not followed at Miami, Florida or at sites on the west coast of the United States.

0 . Guayaquil, Ecuador, at 2 S latitude has low concentrations throughout the year. Most sites along the west coast of South America show concentration mzsima during the warm months and minima during the cold months. Our data for

Antofagasta, Chile do not fit this pattern. At Chacaltaya, Bolivia, in the

Andes, concentrations are high during the dry winter season, when westerly I winds predominate, and low during the wet summer season, when southerly winds predominate.

The Pacific Island sites at Tutuila, American Samoa and at Easter Island have relatively low concentrations throughout the year. In contrast, the high elevation site at'~%unaLoa Observatory, Hawaii has high concentrations throughout the year. It appears that air from the upper troposphere is sampled at Plauna Loa

throughout the year. At Chacaltaya, on the other hand, during the wet summer

season at least, the air reaching the site appears to have had a substantial

previous history within the lower troposphere. This is true in spite of the

greater elevation of Chacaltaya, and probably is attributable to its location within a long, high mountain range which has a major impact on the circulation of the tropospheric air.

. .

J FMAM J J AS 0 N D.

0 SOUTH POLE STATION 250 - @ THULE - -

;;;' F 5 2 200- - \ 2 0 i= '? 150 2 lus u0

I 100 8

50 - -

JFMAMJJASOND-

RICHMOND - 250 - 0 BALBOA $ A CUAYAQUIL 5 5 c3 - 2 200 - 2 0 i= 9 k 150 cc3 ? 0 b I00 4

0 ANTOFAGASTA SANTIAGO

JFMAMJJASOND

0 MAUNA LOA @ SAMOA 250 - A ISLA DE PASCUA -

DOE Conference on "Natural Radioactivity Isotopes of Beryllium in the Environment"

~ctober1-2, 1979, Kline Geology Laboratory, Yale University, New Haven, CT.

7 BNL Paper: Be levels in precipitation and air as measured at BNL during

the years 1971-1978.

J. R. Naidu, A. R. Moorthy and A. P. Hull

Met hod : Collection and Analysis of Samples

A. Mr Samples:

High volume (500 1 min'l) positive displacement air pumps ,

(Cast 3040) were operated at a monitoring station southeast

of the Solid Waste Management areas (Fig. 1, S-6), and at

the northeast and southwest perimeter-stations (P-9 and

P-4). The air sampling media consisted of a 7.6 cm diameter

air particulate filter (Gelman type G) followed by a 7.6 cm

x 2.5 cm bed of petroleum-basec charcoal (Columbia Grade LC

12/28 x mesh) for collection of radiohalogens. Analyses for

w gamma emitting nuclides were pe~formedon a monthly com?o-

site of all individual air'particulate samples. Additional

gamma analyses were also scheduled at six month and one year

post-collection to facilitate the resolution of short and long

lived nuclides with full energy peaks too close to be resolved

by the Nal detectioil system employed. B. Precipitation:

Two pot-type rain co.llectors each with a surface area of

0.33 m2, are situated adjacent to the Sewage Treatment plant

(see Fig. 1). Two routine collections were made fron these, ... . ,7 ,*-J,, V31V lN3W33QNVW 31SVM vs d313WIU3d 3N 6-6 1313WIU3d 3s L'd d313WIU3d IHS Vd 8313W113d MN Z'd SNOIIvls 3NUOUNOW IV 13VlS 3110 LM YDVlS 4-iVV13 30 NVA 106 13VlS U84H SOL 13VlS 1NVld WV3S 019 X3VlS AdlSIW3H3 SSS )IDVlS 11W 161 13v~3uw w 5038 131113 ONVS OSZ NOUVN9IS3Q

lil . *"

IW1 *IY.m % rCI&.a ;( :wmmYO.

. ,.. /

' - - .. one whenever precipitation was observed during a previous 24 hour

(or weekend) period, and the other once a week whether or not

precipitation occurred by washing down the rain collector with 8 a known volume of water. Part of each collection was evaporated

for gross beta counting, a small fraction composited for montllly

tritium analysis, and the balance put through ion exhange columns

for subsequent cluarterly 89~r,90~1- and gamma analyses.

C. Counting Systems:

i. NaI System: It consists of two 4" NaI(TI) detectors, one of

which is stationary and the other can be moved up and down to

accomodate samples of varying geometry.

ii. GeLi System: It consists of a 145 cc GeLi detector giving a

7% efficiency for 7~e(477 KeV) .

iii. The Quality Control and the Minimum Detection Limits for the

two systems are given below. The error in 7~edetermination averages

12% and ranges from 10 to 25%.

D. Results:

Table 1 gives the 7~econcentration in air samples for the years

1971-1978. Figure 2 is a plot of the same.

Table 2 gives the 7~econcentration in precipitation and rainfall data

for the years 1971-1978. Figure 3 is a plot of the same.

Table 3 gives the 7Be concentration in air samples asdetermined by DOE-

EML for the period 1971-1978.

Figure 4 is a plot of the average values for rainfall, 'Be concentration

in air samples and precipitation for the years 1971-1978. QUALITY CONTROL

Alpha (a), .])eta (d) and Gamma .(y).i Certified radioactive standards from the National Bureau of

Standards, U.S. Uel~artlncrit of Conunerce, are used to standardize

radiation measurcme~~tinstruments. These Standards are certified to

. be at least within 5'% of stated values. In some cases/certified

standards were also obtained from An~ershamlSearle. Daily checks of

performances are ulclde using the Standards as well as backgrounds. In

. . additiot~,some samples art. counted both in NaI system and Ge(Li) system.

Ge(L1) system were calibrated using a new multi-gamma NBS Standard ob-

tained in October 1977. The results from NaI and Ge(L1) systems agree within 5%. For tritium measurements a number of Standards and blanks

are included with each run of a liqllid scintillator counter which has

a programmed automatic sample changer.

The Analytical Laboratory of the Safety and Environmental Protection

Division is a participant in the inter-lab comparisons.of radioactivity

in gamples of different matrices of water, air filters, soil, vegetation

and bone. These samples are distributed by the Department of Energy (DOE)

through the Environme~ltal Measurements Laboratory (EML), New York,

formerly known as the Health and Safety Laboratory (HASL), on a quarterly

baais. The redl onuclidcs assayed were 3 H, 'OS~, pl~tonfunifjotopes

(following wet chemistry) and a number of y emitting nuclides. hr results

agree within 10% for wattar samples and witl~in 15% for other eample matrices. Sotlie of tllc values ill gamlra scans by NaI detector are not indicated in tlle toh1t.s as clrese values were at or below MDL. 'L'tle MDL values arc a t'u~lctioi>of Matrix (efficiency), Count Time (background), etc. Typical tahlcs ior NaI and Ge-Li systems are given below:

units: IU-6 pci Dcct!cLor: 'ho 4" Nal crystals Gcun~cLry: Plancl~etand air j~artictrlates

Count Time (sec)

Count Tir~rt- (sec) - 4,000 8,000 40,000 60,000

Units: wci - i)etector: 145 Ge-Li Decector Geometry : Filter paper

Count Tin~e (sec)

Count Tirnc:

(sec) ' 4,000 50,000

Count Ti~tlc: (sec) 4 ,OOb SO. 000 Discussion:

The following Table indicates the variability of the particulates concentration as observed at BNL*. It confirms the spring maximum observed in the concent;ation of 7~ein air samples (Tgble 1 and Figure 2). . . Year Season ~onc.(pg/m3 Mean p*

1975 Winter 23.7

Spring 24.6

Summe r , 22.0

Fa11 .23.0

Winter 27.3

Spring 47.4

Summer 29.5

Fall 21.5

(*S. SethuRaman and R. M. Brown-BNL-22998-1976). (**Mean of >8500 continous readings). The correspondence of 7Be concentration in precipitation to rainfall is also demonstrated in Figures'3 and 4.

\ We would like to ;elate, in the near future, the 7~econcentration in precipitation to that in the Peconic River. Other sources of 7~eon-site, such as the AGS,,BLIP, are insignificant , and if any are limited to confined spaces within the facility and do not contribute to environmental levels as observed.

I Acknowledgement:

A special note of thanks to E. Hartmann, J. Nobile, L. Olmer, A. Wallner and J., Steimers whowere responsible for collectingand analysis of the samples. Table 1

. , qi~~;i~~~i~~;j~+n>;t:p;.!~:~:.~~i~g~~t!;!~.i"~ -st 3 Air ,Partj.culate Filters - Be (p~i/m)

Feb 0.103 0.153 . 0.116

Mar 0.243 0.156 0.108

&?r 0.374" 0.180 0.120

May 0.476* 0.230 0.089

.Jm 0.498* 0.262, 0.130

Jul 0.435* 0.228 0.150

A117 0.222 . 0.196 0.110

S12p 0.167 0.152 0.160

Oct 0.084 0.119 0.147

Nov C.077 0.086 0.118

I:.* ;.a in errcxr excluded from plot.

- --.-. . .-- . ------.-.. . - 1971 t972 1973 1974 197s 1976 19n 1978 PL.f.11 7L IWnlall 7m illinfall 7- illiafsll 7- illinfall 7- illinfall 7 if7 Ui.f.11 7h mth em OCUm ell. lrollm cu nCi/m CM I cu nCi/m eu nCilm CM IICll. cu ecil. Y. 5.00 5.2 5.21 3 .O 9.65 13.85 13.14 1.38 27.2 ?.L 11.73 14.4 7.24 3.4 :::: I 10.2 4.86 18.36 7.89 14.98 7.01 9.86 3.98 13.6 6.6 31.0 1 - 6.02 9.6 10.66 7.0 9.01 10.58 7.55 3.54 11.16 8.5 I 1 4.32 4.6 9.W 18.3 13.59 9.45 9.31 4.55 9.00 6.1 7.92 26.6 11.89 125 14.05 13.3 7.23 66.27 5.91 10.67 8.64 2.07 3.85 12.8 16.6 35.1 k 19.0 4.47 I 6.41 11.98 6.74 9.03 2.1 nr 4.60 3.1 19.96 2.11 7.48 8.04 3.20 llh &uwr 14.1~ 6.7 7.47 4.42 0.32 23.06 11.74 17.95 17.8 I 2.9 I 1.9 3'00 13.3 50.5 -Pt 5.03 1.4 9.19 8.94 16.09 10.70 4.94 9.68 10.8 I Q= 7.90 5.9 19.71 5.49 7.33 11.13 2.26 10.68 10.6 IA 8.76 1.7 7.80 I 9.2 ::: 1 16.2 3.91 10.35 10.35 18.57 2.32 6.02 23.6 7.1 51.3 D.c 8.20 4.1 13.W 18.29 13.21 12.19 5.68 0.32 14.20. 15.5 -Z 89.47 91.2 114.51 78.4 123.09 53.6 91.99 102.45 108.96 48.54 1W.79 26.25 1W.13 67.8 135.8 167.9 1 7.46 7.6 9.54 6.53 10.26 4.47 7.67 8.54 9.08 - 4.65 8.40 2.19 8.34 5.65 ' 11.32 13.99

Table 3

3 7~eConcentration ,(pCi/cm ) in Surface Air for New York,* New York and BNL for the Years 1971-1978 - Year/ ' BNL EML BWL EML BNL EML BNL EML BNL EML BNL EML BNL EML BNL EML Month . 1971- 1972 1973 1974 1975 1976 1977 1978 . F I I

Jan ,094 .I11 .086 .158~ .113 .I14 .I52 ,103 .I15 .086 .093 .I14 .120 ,102 , .I40 .091

Peb .I03 .lo2 -153 .I11 .116 .I20 .158 .I15 .lo4 .089 , ,105 .125 ,210 .113 ,113 .072 Mar -243 .I72 .156 .122 .lo8 .117' .207 .140 .I75 ,149 .138 .136 .180 .121 .270 .I25

April .3746 .I18 .I80 . .188 .I20 .124 .201 .131 .I46 .148 .I27 .I40 .I90 .I60 .097 .lo5

May -4766 -136 -230 ' .134 -089 -106 -168 ,114 -118 .I33 ,102 .I26 ,180 15 .086 -069 June .48% .131 .262 . .lo8 .I30 .I40 ,131 ,110 ,107 .132 .lo8 .I56 .I40 ,208 .180 .I15 July .43* -139 -228 -091 .I50 .168 .I38 .130 .I23 .I43 .118 .137 .I20 .128 ,133 .110 hg -222 .I70 .I96 .I11 .110 .146 .I29 .I22 .I10 .I40 .I17 .142 .I20 .I19 .I26 .082 sePt -167 ND .I52 .I23 .I60 .I24 .I27 ,123 .074 .092 ,122 ,141 .lo0 .082 .I80 ND

Oc t .084 .I42 .119. .lo1 .I47 .110 .093 -097 .078 .096 .I15 .lo1 -090 .111 .127 ND

Nov .077 -093 .086 .087 .118 .085 .I14 .094 .I16 .128 .I15 .084 ,170 -079 .lo0 .I16 Dec .096 .I12 .070 .086 .I30 .lo0 .099 .086 .098 .I15 .17& .094 .I90 ,073 .085 ND

I: 1.086 1.426 1.918.'1.42 1.491 1.454 1.717 1.365 1.364 1.451 1.438 1.496 1.810 1.451 1.637 .885 - X 0.136 .I30 ,160 ,118 .124 .121 .143 .114 .114 .121 .I20 .I25 .151 .I21 .I36 .098 a: Error between 20-100% ND: No data *Determined by DOE-WI,, New York. New York %ata in error, thus total end average represent data for 8 months only.

"NATURAL RADIOACTIVE ISOTOPES OF BERYLLIUM IN THE ENVIRONMENTv Octaber 1-2, 1979 Kline Geology Laboratory, Yale University, New Haven, CT

Determination of 7*Be/Ozone ratios in the Stratosphere Liaquat Husain, Ali Rusheed, Vincent Dutkiewicz New York State Department of Health Division of Laboratories and Research Empire State Plaza Albany,NY. 12201

Most of the terrestrial ozone is produced in the stratosphere, the concentration peaking at about 25 km and monotonically decreasing to the tropopause. Transfer from the stratosphere to the troposphere occurs when the tropopause becomes vertical in the core of the jet stream, and then folds beneath the jet core. Reed and Danielsen (1959) showed that the folded structure could be identified by its large values of potential vorticity and used the term "tropopause foldingn to describe the process. Danielsen and Mohnen (1977) have experimentally demonstrated that ozone rich stratospheric air is transported into the troposphere with each major cyclonic development. Interactions of cosmic .rays with oxygen, nitrogen, and argon in the stratosphere (and to a much smaller degree in the troposphere) produce a variety of radionuclides. During the tropopause folding, along with ozone, these radionuclides are transported to the troposphere. Therefore, these cosmogenic radionuclides can be used as tracers of the stratospheric air. Data obtained in this laboratory (~usainet al., 1977 elsewhere (Reiter et al., 1975, 1977; L dwick et al., 1976 demonstrated that correlations between 3Be and ozone exist, sug- gesting that at least at times ground level ozone originated in the stratosphere. However, a considerable amount of ozone is also produced by photochemical reactions at ground level. The resolution of stratospheric 03 from that produced in the troposphere poses a difficult task. The critical information essential in quantifying the stratospheric ozone at ground level are (1) a knowledge of 7~e/0~mixing ratios in the upper troposphere and the stratosphere, (2) variations of 7~e/03mixing ratios with latitude and seasons, and (3) differing residence times of 03 and 7~ebearing aerosols. In ddition to obtaining the essential data base for correlation of tBe, correlations with 0 , we have also made the measurements of 7~e/0~mixing ratios at ?0-12 km and 210-560N. The upper atmospheric samples (10-12 km) were collected on IPC 1478 filters using an automated sampling system mounted in the nose of two commercial B747 jetliners by NASA as part of their Global Atmospheric Sampling Program (GASP). The on-board instruments also continuously monitored outside ozone using a W ab orption photometer. The whole filters were later analyzed for 4.Be with a 45-cc Ge(Li) detector co ected to a 4096 channel pulse height analyzer. The absolute %e concentration was determined by comparing the area under the 478 Kev peak in the sample with that of a 7Be standard counted in the same geometry. Except in cases where low air volumes were samples, statistical uncertainties in the absolute 7~econcentration were (2 10%. The measured 7 Be and 03 concentrations, and the 7~e/03ratios for the lower stratosphere samples..aregiven in Table 1. Information regarding the exact location of the tropopause at the time of the sample collection was not available at this writing. Therefore, we have tentatively assigned samples to the stratosphere and the troposphere on the basis of the observed 03 concentrations. When the observed 03 concentrations exceeded 200 ppb 7 , we assumed the sample was collected in the stratosphere. The Be concentrations varied from 2,300 to 6,400 fci/m3, with an average of 4,179 fCi/m) (Table 1). The 7Be/0j ratio varied from 7.1 to 17 with most data points close to the average ?Be/O ratio of 11.5. No evidence of deasonal variation is evident alt2 ough it must be admitted that the data presented in Table 1 are limited in number. -Determination of Ground level Stratospheric Ozone The stratospheric 7~e/03ratio can be combined with the ground . level 7Be concentrations to yield ground level ozone of stratospheric . origin. Figure 1 shows daily 7~eand ozone concentrations measured at Wh'teface Mountain,New York during June 15 to July 14,1977. The peak +Be concentrations are approximately -1096 of the stratospheric values, and in some cases higher than the 5 Be concentrations in the upper troposphere. The 7Be peaks are accompanied by increases in ozone concentrations Both ozone and 7Be peaked on June 24-25 and -' 28 and July 1. The 3Be peaks of June 15-16 and July 3 are also accompanied byhigh ozone, but ozone concentrations peaked a day later. Husain et (1977) have explained this one day delay as increased tropospheric contribution due to the existing stagnant conditions. The 7~e/0~r tio measured at Whiteface Mountain on June 15 and 16 is 9.9 fCl/m3 /ppbv, approaching the annual average ratio of 11.5 fCi/m /ppbv observed in the lower stratosphere. Thus, it appears that at least at times,.astratospheric air mas may reach ground level without signinicant fractionation of the 7Be and ozone. From the observed 7Be concentrations on June 15-16, 24-25, and 28 and July 3 we calculate the respective stratospheric ozone contributions of 43, 17, 30 and 27 ppbv. The stratospheric ozone contributions are quite substantial, ranging between 25 and 8696 of the observed 24-hr average ozone. Thus, at a remote location such as Whiteface Mountain both' tropospheric and stratospheric ozone sources play importain roles. Stratospheric-Tropospheric Exchange Since 7Be appears to accurately trace stratospheric air massess, global 7Be measurements can provide a basis for tudying stratospheric- tropospheric exchange. Continous monitoring of 7Be concentrations in ground level air has been carried out by the Environmental Measure- ments Laboratory (1978) at stations in both the northern (Mi) and southern (SH) hemispheres. The measurements be een 1970 and 1977 provide the means to deduce the annual average %e concentrations in ground level air at a number of latitudes in both hemispheres, Fig. 2. Since each point in Fig. 2 re resents the average,of measurements over an eight-year period, the 9Be concentrations should be representative of the annual average 7Be concentrat ons in ground level . air. The solid lines in Fig. P represents average 'Be concentrations for 0'-300, 30°-60°, and 60°-go0 latitude bands in the NH. In the SH, insufficient data were available for the entire latitudeoband, 68'-90'; hence, average 7Be concentrations are only deduced for 0 -30°, 30 -600, and 60°-70°. The 7~edata (Fig. 2), suggests significant hemispheric asymmetries in the stratospheric-tropospheric exchange. The ratio of 7Be concentrationsin the NH and SH for sites a comparable latitudes vary between 2 and 3. Since no stratospheric Be hemispheric asymmetries are expected or observed, the greater Be concentrations observed in the NH can be directly attributed to a greater stratospheric flux in the NH. Global Tropospheric Ozone Using the 7Be/03 ratios and average 7~econcentrations shown in ig. 2(corrected for tropospheric 7Be and the mean residence times of 7Be and 03), we calculate the ground level ozone contributed from the stratosphere. The calculated latitudinal ozone profiles for both hemis heres are shown as the solid lines in Fig. 3. Fishman and Crutzen (19787 have calculated a value of 25 ppbv (dashed line in Fig. 3) for the total tropospheri6 ozone concentration at 30°-60' N. In Fig. 3, we also include recent measurements carried out continously .on a yearly basis at various sites. The average of these individual ozone measurements yield a value of 282 7 ppbv for total tropospheric ozone, in substantial agreement with the estimates of Fishman and Crutzen(l978). We calculate the concentration of stratospheric ozone in the lower troposphere to be 20 ppbv at 30°-60° N. Therefore, most of the observed ozone in the troposphere of the mid-northern latitudes must originate in the stratosphere. Conclusions The use of 7~econcentrations measured at ground level with stratospheric 7Be/03 ratios provides a technique for determining the stratospheric ozone contributions to the troposphere. Our calculations suggest that the stratosphere is the primary source of tropospheric ozone in both hemisphere. Furthermore, the stratospheric -tropospheric exchange rate in the NH is two to three times that in the SH.

Acknowledgments The authors are grateful to NASA Lewis Research Center and Porter Perkins, Jr., Erwin Lezberg, and Francis Humenik for their excellent cooperation in supplying the filters and the ozone data. The Whiteface Mountain Observatory is operated by the Atmospheric Sciences Research Center, State University of New York, Albany, New York . This work was part1 supported by U.S. Department of Energy Contract No. EE-77-S-02-1 501. References Danielsen,E.F. and V.A. Mohnen, Project Dustorm Report: Ozone Transport in situ Measurements, and Meteorological Anal sis of Tropopause Folding,-J . Geophys . Res . 82, 5867-5977 (1 9777. Dutkiewicz, V.A. and L. Husa'n,Determination of Stratospheric Ozone at Ground Level Using +Be/Ozone Ratios, Geo~hys.Res. Lett. -6,171-173 (1979). Fishman, J. and P.J. Crutzen, The Origin of Ozone in the Troposphere, Nature 274 855-858 (1978). Husain, L., P.E. Coffey, R.E. Meyers and R.T. Cederwall, Ozone Transport from Stratosphere $0 Troposphere, Geophys. --Res.Lett.2, 363- 365, 1977. Ludwick, J.D., R.D. Fox, and L.L. Wendell, Ozone and Radionuclide Correlations in Air of Marine Trajectory at Quillayute, Washington, J. Air Poll. Contr. Assoc. 26, 565-569, 1976. - -7 - Reed, R.J. and E.F. Danielsen, Fronts in the Vicinity of the Tropopause, -Arch. Meteor. Geoph. Biokl. Ser.A, B11, 1, 1959.

Reiter, R., H.J. Kanter, R. Sladkovic and K. Potzl, Measurement of Airborne Radioactivity and its Meteorological Application, Part VI, ERDA Document No. NYO-3425-14, 1977. Reiter, R., R. Sladkovic, H.J. Kanter and W. Carnuth, Measurement of Airborne Radioactivity and its Meteorological Application-Part IV, AEC Document No. NYO-3425-10, January 1975.

...... ,. U .S . ~epartmentof Ener , Environmental Measurements Laboratory Report, EL342 c1978y Table 1. otermination of 7Be.and 0 Concentrations and-=+ Be/O) Ratio in the stratasphere -- Latitude Altitude 7~e O3 '~el~~ Date (ON) (km) (fci/m3) (ppbv) ( fCi/m 3 /ppbv)

January 9 February 17 April 18

' April 23 .. May 24 . May,27 June 17 ;December 29 'U December 29 8 r 1979 January 9 January 18 February 27

AVERAGE ' 78- ' 79 34'-42'~ 11 4,179 ' 364 11.5 WHITEFACE MT. 1977

I I I ozone I

JUNE JULY

Figure 1. Daily 7Be and 24-hr average ozone concentrations for June and July 1977 at the 1.5 km summit of Whiteface Mountain,New York. Selected days are indicated in the figur to aid in intercomparisons . pCi/KSCM= fCi/m 5 . (after ktkiewicz and Husain,G.R.L. 6,171- 173,1979.) 9-9 9

1 60'- - - - (Fishman and Crutzen, 1978) - . . - : 8 Project TROZ (Fabian + - 50 - Pruchniewicz, 1977) - : A (Singh et a!., 1978) - - 0 Whiteface Mountain, NY 40 - A - (Mohnen et 01.. 1977) - A - A - • A : 0 Aspendale, ~ustralia(~ittock, I977 ) 30 - 8 i Auckland, New Zeoland (farkas,1978) - - 8. ------C'---'--. -9-- - - - J- - -l 8b8 8 ---- - I r8 20-- -- - 8- --- 8 f I - 8 m- - - - 10 - I - I - - - I I I I I I 1 I I I I I I 90" 70" 5 0" 30" 1 0" 10" 30" 50" 70' 90". NORTH LATITUDE SOUTH AEROSOL DEPOSITION VELOCITIES ON THE PACIFIC AND ATLANTIC OCEANS CALCULATED FROM 7~eMEASUREMENTS

3. A. Young W. B. Silker

September, 1979

Prepared for the U.S. Department of Energy Under Contract EY-76-C-06-1830

Pacific Northwest Laboratory Richland, Washington 99352 AEROSOL DEPOSITION VELOCITIES ON THE PACIFIC AND ATLANTIC OCEANS CALCULATED FROM '~eMEASUREMENTS

3. A. Young and W. B. Silker

7 The concentrations of Be have been measured in Pacific and Atlantic ocean water for the past several years to determine the deposition velocity of aerosol particles on the ocean surface. Beryl1 ium-7 is produced at a relatively constant rate in the atmosphere by spallation reactions of cosmic rays with amos- .pheric nitrogen and oxygen. Immediately after its formation 7Be becomes attached to aerosol particles, and therefore can serve as tracers of the subsequent behavior of these particles . Isopl eths of 7Be surface water concentration, 7~e' inventory in the ocean, and deposition velocity have been prepared for the Pacific Ocean from 30"s to 60°N and' for the Atlantic Ocean from 10°N to 55ON. The concentrations, inventories and deposi tion velocities tended to be. higher in regions where precipitation was high, and generally increased with latitude. The average 7 flux of Be across the ocean surface was calculated to be 0.027 atoms sec-' which is probably not significantly greater than the worldwide average 'Be flux across land and ocean surfaces of 0.022 atoms cm-2 sec-' calculated by La1 and Peters. The average deposition velocity was calculated to be 0.80 cm sec-l . This value may be 10 to 30%,too low, since it.was calculated using atmospheric 7~econcentrations which were measured at continental stations. Measurements of atmospheric 7~econcentrations at ocean stations suggest that the concentrations at the continental stations averaged 10 to 30% higher than the concentrations over the ocean. INTRODUCTION

Beryllium-7 is produced at a fairly constant rate in the atmosphere by spallation reactions of cosmic rays with atmospheric nitrogen and oxygen. Its production rate increases by about three orders of magnitude between ground 'level and the lower stratosphere (La1 et al., 1958; Young et.al., 1970). Imme- 7 diately after its formation Be becomes attached to atmospheric aerosols and thus can serve as a tracer of the behavior of these particles. Knowledge of the rate of deposition of various atmospheric particulate contaminates upon . . the ocean surface has been desired for many purposes, but is very difficult to obtain. Therefore, hundreds of measurements of 7Be concentrations were carried out mostly in the top 100 meters of the Atlantic and Pacific Oceans (mainly in 7 the northern hemisphere) to determine the average flux of Be and its associated 7 particulates across the ocean surface. The deposition velocity, Vd, of Be on the ocean surface was then calculated using the equation - 1 FLUX ACROSS THE OCEAN SURFACE (atoms -cm-' sec- ' ) 'd (Cm = Atmospheric Concentration (Atoms crn-3)

Using this equation the flux across the ocean surface of any material having a deposition velocity similar to that of 7~ecan be calculated from its atmospheric concentration and the 7Be deposi ti on veloci ty.

The deposition velocity was calculated using the steady state assumption ~,. that .the rate of decay of 7~ein the ocean is equal its rate of deposition on the ocean surface by wet and dry processes. This assumption is not strictly true, since atmospheric concentrations show seasonal variations, and because there are short-term variations in the rate of deposition, especially deposition in precipitation. However, 7Be has a 53-day half-1 ife, so its decay rate in the ocean is determined by the average rate of deposition over several weeks, which should minimize the effects of short-tern fluctuations in the rate of deposition. The effects of seasonal variations should also cancel out when. average deposition velocities over long periods of time are calculated.

Beryllium-7 is an excellent tracer for the determination of the deposition velocities of particulate materials on the ocean surface for several reasons. First, its production rate in the atmosphere remains fairly constant. There- fore, deposition velocities can be calculated fran oceanic and atmospheric con- @ centration measurements made during different years, and measurements can be averaged over several years to obtain more accurate estimates of average deposition velocities. The deposition velocity of 7 Be on the ocean surface is also relatively eas- to determine. Its concentrations are fairly easy to measure in the atmosphere and in ocean water. Our measurements of vertical profiles of 7Be concentration 7 in the ocean have shown that because of its relatively short half life Be is confined almost entirely to the top 100' meters of the ocean. Therefore, inven- 7 tories of Be in the ocean can be calculated from measurements down to this rather modest depth. The shoi-t half 1ife also insures that 7 Be measured in the ocean was deposited within a few months prior to measurement, and minimizes, but obviously does not entirely el iminate, the effect of horizontal transport upon the measured concentrations. In regions of strong surface currents a sizable fraction of the 7Be measured at one location could have been deposited a con- siderabl e distance away. A one-knot current would transport water 1300 nautical ,miles in 53 days. A tracer with a much shorter half life would have to be used to entirely el iminate the effects of horizontal transport. 7 A third reason that Be is an excellent tracer for the determination of the deposition velocities of particulate materials is that its size distribution in the atmosphere is similar to that of many other particulate contaminants of interest, and therefore it would be expected to be deposited on the ocean at a similar rate by wet and dry deposition- processes as these other contaminants. Beryllium-7 is attached primarily to submicron sized particles in the atmosphere. At Richland, Washington 88%of the 7 Be was found to be present on particles smaller than 1.1 microns in diameter, and less than 1%was on particles larger than 7 microns in diameter (Young, 1974). Many anthropogenic pol lutants are a1 so present in the atmosphere primarily on submicron sized particles. For example, in St. Louis it was found that 67% of the lead, 75% of the bromine, 62% of the copper, 69% of the arsenic, and 81%of the sulfates were present on particles smaller than 0.3 microns in diameter (Young, 1975). In the period immediately following a nuclear test the debris may contain a fairly large fraction of large particles, especially in the case of a ground level test which ingests considerable crustal material into the fireball. However, the larger particles settle 7 out fairly rapidly, causing the size distribution to approach that of Be. Shl-eien et a1 . (1966) reported that 400 days after the last nuclear test, 88% of the debris was present on particl es small er than 1.7 microns. Different submicron size particulate materials tend to approach the same equilibrium size distribution in the atmosphere after a few days as a result of processes such as gas to particle conversion, coagulation, and the settling out of the few particles which grow to large size (Junge, 1963). These particle growth processes also tend to lessen the physical and chemical differences between different particles by producing particles of mixed composition. There- fore 7 Be should be deposited on the ocean surface at a rate similar to that of other submicron particulate contaminants. It is also possible that the deposition velocity of contaminants whose size distribution contains an appreciable large particle fraction could be calculated by adding the large particle grav.i- tational settl ing velocity to the 7 Be deposition velocity. 7 The Be deposition velocities which are reported in this paper were calcu- 7 lated from atmospheric I Be concentrations measured at ground 1eve1 . However, a 7 considerable fraction of the Be deposition results from precipitation. Gener- al ly, the precipitation scavenging of submicron particles resul ts primarily from 7 in-cloud rather than below-cloud processes, so the rate of Be deposition by precipitation should be proportional to the concentration of the air feeding the clouds, rather than the ground level concentration. Therefore, the fluxes of other atmospheric contaminants across the sea surface can be calculated accurately from their measured atmospheric concentrations and the 7 Be deposi tion velocity only if the variation in their concentrations with a1 titude up to cloud level is ,. .< . not too different from that of 7 Be. The rates of increase in the 7Be concentrations with a1 titude measured by Battelle Northwest up to 19 lan at 12 to 18" N, 35" N, and 46" N were generally very similar to those of radionuclides which 1 had been introduced into the abnosphere by nuclear tests (Young et a1 ., 1969; Young and Sil ker, 1971; Young and Wogman, 1972). Therefore, it should be possible 7 to use the Be deposition velocity to calculate the fluxes of these radionucl ides across the sea surface. However, the concentrations of most other anthropogenic atmospheric contaminants decrease with a1 ti tude, at least near their source, so the fluxes of these contaminants across the sea surface calculated from the 7 Be deposition velocity might be somewhat too high. However, these contaminants become mixed through the boundary layer fairly rapidly, and the moist air feeding clouds generally originates at low altitudes, so the error introduced by using sea level concentrations may not be very great. In any event the relative variation in the deposition velocities of these contaminants with time and location should be similar to that of 7 Be. The ground level atmospheric concentrations 7 and rates of wet and dry deposi tion of Be and other atmospheric contaminants are now being measured by Battel le-Northwest at a station on the coast of Washington State to determine how accurately the rates of deposition of other 7 atmospheric contaminants can be calculated using Be deposition velocities.

EXPERIMENTAL

All of the seawater samples for this study were processed with a Battelle large volume water sampler (Sil ker et al., 1971). a unit which first directs the water through a para1 lel bank of filters to remove particulates, then through a 7 bed of aluminum oxide which has been determined to retain 60% of the Be which has passed through the filter at the operational flow rate of 35 1i ters/min. The absorption efficiency of the bed was determined in the laboratory by passing 7 7 sea water spiked with Be through the bed and measuring the Be absorbed by the bed and the 7 Be remaining in the sea water. The absorption efficiency was also 7 7 determined at sea comparing the Be absorbed from seawater samples with the Be scavenged by hydrous ferric and manganese oxides from 600-1 iter seawater samples collected at the same time (Sil ker, 1975). Laboratory tests have demonstrated ' 7 that successive beds remove the same fraction of the remaining Be in the sea water. The absorption efficiency of the aluminum oxide beds has therefore been checked periodically at sea by calculating the efficiency from the relative amounts of 7Be absorbed from sea water'by two beds arranged in series (Silker, - 1975). The three methods for determining the absorption efficiency showed good agreement. Surface seawater samples were obtained either from the salt water intake line of the research vessel or by using a float to suspend the end of a hose 20 to 30 an below the ocean surface. Samples of sea water obtained simultaneously 7 from these two sources have shown essentially identical Be concentrations. The equipnent used to collect water samples from depth consisted of a large motorized reel which was used to lower and retrieve a bundle of five 2.54 cm diameter hoses, each terminating at a different depth. Water was pumped from each depth through the Battelle sampler using a deck-mounted centrifugal pump. The depth of each sample was determined from the gas pressure in small-bore plastic tubes filled with compressed gas attached to each sampl ing line and terminating at each sampling depth (Silker, 1975). After processing a 2,000 to 4,000-1 iter sample, the Battel le large volume water sampler was disassembled and the filters and alumina removed and reserved for.nondestructive radiochemical analysis employing a multidimensional gamma ray spectrometer (Wogman et al., 1967). This instrument consists of two princi- pal NaI(T1) detectors surrounded by an anticoincidence shield, all housed in a lead cave with 10-cm thick walls. When a radionuclide emits a single photon which interacts with one of the two detectors, the event is stored in a 4096 channel computer memory at a position determined by the energy of the photon. However, ifa radionuclide emits two photons simultaneously, and one photon is detected by one detector and the other photon is detected by the other detector, the event is stored at a position in the memory uniquely determined by the individual energies of the two photons. In this manner, two radionuclides which emit photons of similar energy can be distinguished easily ifat least one of the radionuclides emits a second photon simultaneously. 7 Although several hundred surface water samples have been analyzed for Be, only about 30 vertical profiles have been measured because of the difficulty and expense involved in obtaining samples from depth (several more profi1 es were measured east of Barbados, British West Indies during the Bomex experiment in the summer of 1969, but they have not been used in this paper because the concentrations have been shown to have been influenced by Amazon River water [Young and Silker, 19741). A large percentage of the surface water samples were collected by technicians aboard NOAA-operated research ships while the ships were moving. In order to obtain depth samples, it was necessary to place the rather bul ky-depth sampling reel aboard the ship, to have two Battelle people aboard ship to operate ., I the equipment, and to sample for an hour or so while the ship was held stationary, a fairly difficult task at sea.

RESULTS AND DISCUSSION

7 Surface Water Be Concentrations 7 Isopleths of Be surface water concentrations for the Pacific and Atlantic Oceans are shown in Figure 1. The contours were drawn by computer. Sampl ing locations are shown by points in the figure. Too many samples were collected in the Pacific Ocean to show all of the locations, so each point represents the

, average location of generally two samples collected in close proximity to one another at about the same time. Fewer samples have been collected in the Atlantic, so each sampling location is shown by a point, except that a few points were omitted in areas where many samples were collected. Many of the measured concentrations have been reported in earlier publications (Sil ker et al., 1968; Silker, 1972a; Silker, 1972b; Silker et al., 1973; and Young and Silker, 1974). The surface water concentrations generally increased with latitude, a1 though there were regions of concentration minimums and maximums which in most cases corresponded fairly well to regions of minimums and maximums in precipitation. This is in accord' with the generally accepted belief that deposition'occurs primarily in precipitation in areas where precipitation is light.

Average annual rainfall for the Pacific and Atlantic Oceans is shown in Figure 2 (Drozdov, 1953). It can be seen that the concentration minimums off the west coasts of California, South America and Africa occur in regions of minimum precipitation, and that the highs in concentration in the Gulf of Mexico and off the west coast of Central America occur in a region of high precipitation associated with the intertropical convergence zone. Other areas of high or low concentration also appear to be associated with areas of high or low precipitation, respectively, a1 though the correlations are not as striking.

The surface water concentrations generally did not change very greatly with 7 time of year. The Be concentrations measured from May through October and from November through April are plotted in Figures 3 and 4, respectively. The concentrations at high latitudes in the eastern Pacific were higher from May through October than from November through April, presumably because of the higher atmos- 7 pheric Be concentrations. and lower rates of vertical diffusion in the ocean from 7 May through October at these locations. No Be concentrations were measured from November through April north of 30" N in the western Pacific or north of 42" N in the Atlantic, so it was not possible to compare the concentrations for the two times of year at these locations. There are two concentration maximums which appear in Figure 3 showing concentrations for May through October which are absent in the November through April data. The concentrations measured close to the west coast of Central America in June of 1968 and October of 1974 were more than double the concentrations measured at about the same location in March of 1973 and February of 1974. Also, the concentrations measured off the east coast of the United States at around 30" N in July of 1971 were more than double the concentrations measured in March of 1973 and 1974. The other concentrations maximums and minimums were generally at about the same locations for the two times of year.

'~eInventories More than 30 vertical profiles have been measured at different locations i the Pacific and Atlantic Oceans during the past several years. Eighteen of the profiles are shown in Figure 5. The average monthly mixed layer depth (MLD) given by. Bathen (1972) is shown for each Pacific Ocean Profile. We have not yet discovered published average monthly mixed layer depths for the Atlantic Ocean. The 7 Be concentrations generally decreased rapidly from a maximum at the surface to essentially zero at a depth of around 100 meters. At times when the mixed layer depth was shallower than 100 meters the concentrations appeared to show a more rapid decrease below the base of the mixed layer. However, in most cases where the mixed layer was deeper the 7 Be concentration decreased to very low values before the base of the mixed layer was reached. In general, the correlation between the mixed layer depth and the shapes of the vertical profiles was not very s tri king . The 7 Be inventory was calculated for each profile assuming that the concentrations varied linearly between depths at which concentrations were measured. The concentrations were extrapolated to zero below the deepest depth measured. There were two profiles measured in the. mid-Pacific at 4' N and 6' S which showed

4%a low, nearly unifom concentrations down to the deepest depth measured, 102 meters. - ,.:.. The mixed layer depths were 120 and 100 meters, respectively, at these two loca- :I' tions. In calculating the inventories at these two locations it was assumed that the concentration decreased 1 inearly to zero between 102 and 150 meters. The number of surface 7 Be concentrations that have been measured is considerably greater than the number of 7 Be inventories that have been calculated from vertical profiles. Since the 7 Be inventory' is necessary for the calculation of , ,. 7 the Be deposition velocity, it would be desirable to be able to estimate the 7~einventories from the measured surface concentrations. heref fore, the ratio of the surface concentration to the inventory for the vertical profiles was plotted versus latitude to determine whether the ratio varied regularly wi th latitude (Figure 6). It was found that the ratio was fairly constant for a given latitude and time of year. However, the ratio was significantly higher for August through November than during the rest of the year, indicating that vertical mixing was appreciably slower during these three months, and probably also during the preceeding month or so. The ratio increased significantly with latitude during these three months, but showed 1 i ttle change with latitude during the rest of the year. The ratios in three profiles measured in the southern hemisphere corresponded fairly we1 1 to the ratios measured at similar latitudes in the northern hemisphere, except of course for the six-month offset in the seasons. The ratios of surface concentration to inventory in the Pacific Ocean were also plotted versus mixed layer depth to determine whether 'the value of the ra ' could be predicted better from the mixed layer depth than from the latitude. 1 L was found that the ratio decreased by about 50% when the mixed layer depth increased from 20 to 180 meters, indicating increased vertical transport (Figure 6A). However, the standard deviation of the ratios from the best fit line of ratio versus depth was 20%, somewhat higher than the 15% standard deviation of the ratios from the best fit lines of ratio versus latitude in Figure 6. 7 The Be inventory was calculated for each measured surface water concentration using both the plot of the surface concentration to inventory ratio versus latitude and the plot of the ratio versus mixed layer depth. It was found that inventories calculated using the two methods were generally very similar . This was true because the mixed layer depth generally varied fairly regularly with latitude in the open ocean. During August through September when the plot of the ratio versus latitude gave high values for the ratio, the mixed layer was shallow, so the plot of the ratio versus mixed layer depth also gave high values for the ratio. During these months the mixed layer depth decreased with latitude, so the plot of the ratio versus mixed layer depth gave values which increased with latitude, just as did the plot of the ratio versus latitude. The main differences occurred near continental boundaries, where the mixed layer depth sometimes is considerably different from that in the open ocean at the same latitude. For example, the mixed layer is very deep in the ~ulf'of Alaska in the spring, so the plot of the ratio versus mixed layer depth gave higher values for the inventory. However, the mixed layer is very shallow west of Central America in the Sprf ng, so the plot of ratio versus mixed layer gave lower values for the inventory. It is 1i kely that no method of estimating inventories from measured surface concentrations wi11 be very satisfactory at conti nental boundaries where the sinking and upwell ing of water or the mixing with river water occurs. Since the two methods of estimating the inventories from the surface concentration generally gave about the same results, and since we did not have values for the mi'xed layer depth in the Atlantic. Ocean, the plot of the ratio of the surface concentration to inventory versus latitude has been used to calculate the inventories used in this paper. The isopleths of 7 Be inventory calculated using this method are plotted in Figure 7. The locations of the measured tnventories are indicated in the figure. The isopleths are very similar to t 2 7 of Be surface concentration, with the highs and lows at the same locations. The inventory increase with latitude is st ightly less than the surface concentration increase, because the surface concentration to inventory ratio increases with latitude. The inventories shown in Figure 6 were integrated over 10 degree latitude 9 bands to obtain average values of the '~eflux across the ocean surface (Table I). The 7~kflux increased from 0.01 8 atom5 cm-2sec-1 at 0 to 1O0 S to 0.034 atoms at 40 to 50' N. The average flux in the northern hemisphere was calculated to

, be 0.027 atoms cm-2sec'1. This number is probably not significantly greater than the average worldwide rate of deposition on land and water surfaces of 0.022 atoms cm-2sec-1 estimated by La1 and Peters (1967). There has been considerable debate in the past whether rates of deposition are greater on the ocean than on land. Unfortunately, it is probably not possible to answer this question by comparing 7 the fluxes of Be across the ocean surface with the worldwide average for land and ocean surfaces because there is so much more ocean than land that the flux

-I: across land surfaces has relatively little effect on the worldwide average. If ... the flux across land surfaces was zero, then the flux across ocean surfaces would only have to be 0.029 atoms cm'2sec-1 to produce an average flux of 0.022 atoms -2 -1 cm sec for land plus ocean surfaces.

Atmospheric Be Concentrations 7 7 Atmospheric Be concentrations are necessary for the calculation of the Be deposition velocity on the ocean surface. Unfortunately, atmospheric concentrati ons have not been measured at oceanic locations over the extended. periods of time necessary to determine seasonal variations and to average out the short- term fluctuations resulting from meteorological processes such as precipitation scavenging and vertical mixing . However, several investigators have measured 7~e concentrations continuously at continental locations. Health and Safety Laboratory 7 (HASL) has reported Be concentrations from 1970 through 1975 at 28 stations from 82' N to 90' S on its 80th meridian network (HASL-302). Battelle-Northwest has 7 measured Be at Richland, Washington (46' N) and Point Barrow, Alaska (71' N) 7 since 1962 (Thomas, 1973). Two-month averages of the Be concentrations reported by HASL and Battelle-Northwest were plotted as a function of latitude. The concentrations for May-June are shown in Figure 8. The viariation with latitude is fairly similar for other months. .Part of the asymnetry between the two hemispheres in Figure 8 results from the fact May and June are spring months in the northern hemisphere and fa11 months in the southern hemisphere. Beryl 1ium-7 concentrations reach thei r maximum in the spring at mi.d-latftudes (Figure 9). However, the measured 7Be concentrations at some stations in the southern hemisphere were lower throughout the year than those at similar latitudes in the northern hemisphere. It is interesting that the concentrations at 90' S in Figure 8 are higher than those at somewhat lower lati t'udes. In the northern hemisphere concentrations also increased wi th latitude north of 65' N during the period from November through April'.' The measured 7Be concentrati-ons varied considerably between stations at nearby latitudes, presumably because of 1ocal differences between meteorological processes such as preci pi tation scavenging and vertical mixi ng . For exampl e, the 7Be concentrations measured by Battel le Northwest at Quillayute on the west coast of Washington State have averaged about one-ha1 f those measured at Richland, Washington on the east side of the Cascade Mountains. The yearly rainfall is approximately 80 inches at Quillayute, much higher than the 9 inches at Richland, so the 7Be concentrations are often depleted by 'rainfall at Quillayute. However, the concentration difference between Richland and Quillayute was greatest in the summer, when precipitation is very light at both Richland and Qui 1layute. Probably the higher concentrations at Richl and are ca'used by vertical mixing which occurs when air from the coast crosses over the Cascade Mountains. Beryl 1ium-7 concentrations measured by aircraft north of Richl and increased by about a factor of two for every 2 km increase in altitude in the troposphere, so vertical mixi ng would rai se ground 1eve1 concentrations considerably. 7 The high variability of the Be concentrations at continental stations which has resul ted from local meteor01 ogical condi tions suggests that concentrations measured at conti nental stations may not correspond very 'we1 1 to concentrations over the ocean, so it is instructive to compare the eoncentrations that have been measured at sea with the continental concentrations. Beryl 1ium-7 concentrations were measured by Battell e-Northwes t i n several hundred ai r samples col 1ec ted on a daily basis from May through July aboard ships stationed from 7' N to 18' N

I east of Barbados, British West Indies, during the BOMEX experiment in 1969. The 7~econcentrations averaged 90% of the concentrations expected from the continental 7~emeasurements. During the GEOSECS experiment, 45 air samples were collected at sea on a cruise which extended from 52' N, 177' W to 66O S, 173O E from August 23, 1970 to June 9, 1971. Each sample was collected over a period of 4 to 7 days. The concentrations in these samples averaged 70% of the concentrati predicted from the continental measurements. The 7Be concentrations measured in air filter samples collected from 47' N, 129' W to lo0 S, 170' W on another cruise in March of 1971 also averaged 70% of the concentrations predicted from the continental measurements. The 7 Be concentrations measured in air filter samples collected from 47' N, 129' W to 100 S, .170° W on another cruise in March of 1971 also averaged 70% of the concentrations predicted from the conti nental measurements. These results suggest that the 7Be concentrations measured at the continental stations averaged 10 to 30% higher than the average concentrations over the ocean. ow ever, the air concentrations would have to be measured for a much longer period of time at sea to confirm this suggestion, since 7 Be concentrations fluctuate widely on a short-term basis .as a result of fluctuations in precipitation and vertical mix.ing.

Beryl 1ium-7 Deposition Velocity 7 7 The Be deposition velocity was determined for each calculated oceanic -Be inventory from the atmospheric concentration at that latitude obtained from the plots of atmospheric 7Be concentrations versus' latitude and time of year derived from the measurements of 7Be at continental stations reported by HASL and Battelle- Northwest. The average atmospheric concentration for the month preceding and the month during which each seawater concentration was measured was used to calculate each deposition velocity. Isopleths of 7Be deposition velocity for the Pacific and Atlantic Oceans are shown in Figure 10. The contours are much the same as the contours of 7Be inventory and surface concentrations. The highs and lows are 7 at the sane locations. The high atmospheric '~econcentrations at mid-latitudes tend to depress the calculated mid-lati tude deposition velocities relative to the high and low latitude deposi tion velocities, The deposition velocity varied from *. a low of 0.2 cm sec-I off the coast of Africa to a high of 2 cm sec-' off the west coast of South America. The average deposition velocities for 10-degree latitude intervals is given in Table I. The average deposition velocity for the northern hemisphere was calculated to be 0.80 cm sec- 1 If the atmospheric 7Be concentrations measured at the continental stations are 10 to 302 higher than the concentrations over the ocean, as measurements of atmospheric 7Be concentrations at sea suggest, then these reported values for the deposition velocity are 10 to 30% too low. The deposition velocity did not appear to change very greatly with time of year. The deposition velocities for the Pacific Ocean for May through October and for November through April are shown in Figures 11 and 12. 'There are some differences in the contours, a1 though they may be due as much to differences in 0 the sampl ing locations as to actual differences in the deposition velocities. As might be expected, the maximum deposition velocity of 1.4 cm sec-' for the northeast Pacific in November through April was somewhat higher than the maximun of 1 cm sec-' for August through October.

SUMMARY AND CONCLUSIONS

The deposition velocity of aerosol.,^ on the ocean surface has been calculated 7 for much of the Pacific and North Atlantic Oceans from measurements of Be con-: centrations in the ocean and measurements that have been made of atmospheric '~econcentrations at continental sampl ing stations. Since the deposition velocity is defined as the flux across the ocean surface divided by the abnos- pheric concentration, it should be possi ble to use these deposition velocities to calculate the fluxes across the ocean surface of other particulate contaminants whose atrnospheri c concentrations are known. However, the cal cul a -ted deposition velocities may be 10 to 30% too low because the few atmospheric '~econcentrations that have been measured at sea have averaged 10 to 30% lower than the concentrations measured at continental sampl ing stations. Additional measurements of atmospheric '8e concentrations need to be made at sea to improve the accuracy of the calculated deposition veloci ti es . Additional measurements of 7 oceanic Be inventories would also improve the accuracy of the calculated deposition velocities. REFERENCES

Bathen, K. H., On the s,easonal changes in the depth of the mixed layer in the North Pacific Ocean, Journal of Geophysical Research, -77, pp. 7138-7150, 1972. Drozdov, 0. A,, Annual amounts of precipitation, Russian Norsko, Atlas -2, Chart 48b, 1953. (Obtained from Hill., M. N., "The Sea", vol. 1, 131, 1962.)

. . Lal, D., P. K. Malhotra, and B. Peters, On the .production of radioisotopes in the atmosphere by cosmic radiation and their application to meteorology Journal of Atmospheric and Terrestrial' Physics, -12 306, 1958. Lal, D., and B. Peters, Cosmic ray produced radioactivity on the earth. Handbuck der Physik, 46, No. 2, ed. K. Sitte, Springer Verlag, Heidelberg, 551, 1967. Health and Safety Laboratory environmental quarterly, ~ppendix,HASL-302, B23, 1976. .., Junge, C. E., Air Chemistry and Radioactivi ty, International Geophysics series, 4, 119, 1963. Shleien, N. A. Gaeta; and A. G. Friend, Determination of particle size characteristics of old and fresh airborne fallout by graded filtration Health Physics, 12, 633, 1966.

Silker, W. B. ,. Beryl1 ium-7 and fission products in the GEOSECS I1 water column and applications of their oceanic distributions, Earth and Planetary Science Letters, -16, 131, 1972a. Silker, W. B., Horizontal and vertical distributions of radionuclides in the north Pacific Ocean, Journal of Geophysical Research, -77, 1061, 1972b. Silker, W. B., Collection and analysis of radionuclides in seawater, Advances in Chemistry Series, -147, Analytical Methods in Oceanography, 139, 1975. Silker, W. B., R. W. Perkins and H. G. Rieck, A sampler for concentrating radionuclides from large volume samples, Ocean Eng. -2, 49, 1971. Sil ker, W. B., D. E. Robertson, H. G. Rieck, Jr., and R. W. Perkins, Beryl 1ium-7 in ocean water, Science, -161 , 879, 1968. Silker, W. B:, J. A. Young, and M. R. Petersen, Oceanic distributions and relationships of 7~eand fission products, Radioactive contamination of the marine environment, IAEA-SM-158146, 687 ,. 1973. Thomas, C. W., and J. A. Young, Atmospheric radionucl ide 'concentrat'ions at Richland, Was hi ngton and at Point Barrow, A1 as ka, Pacific Northwest Labo- ratory Annual Report for 1972 to the USAEC Division of Biomedical and Environmental Research, Vol. 11: Physical Sciences, Part 1, Atmospheric 8 . . ' Sciences, 120-122, 1973. Wogman, N. A., D. E. Robertson, and R. W. Perkins, A large detector, anticoincidence shielded mu1 tidimensional gamma-ray spectrometer, Nucl . Instr. Methods, 50, 1, 1967. Young, 3. A., N. A. Wogman, and C. W. Thomas, Radionuclide concentrations between 5,000 and 63,000 feet in 1967 and 1968, Pacific Northwest Laboratory Annual Report for 1968 to the USAEC... Division of Biology and Medicine, Vol. 2, Part 2, pp. 92-100, '1969. '.

Young, J. A., and W. B. Silker, Project BOMEX studies of atmospheric and oceanic mixing and air-sea exchange using radioactive tracers, Pacific Northwest Annual Report for 1970 to the USAEC Division of Biology and Medicine, Vol . 2, Part 2, pp. 42-46, 1971. . . Young, J. A., and N. A. ~ogman, Vertical profiles of radionucl ide concentrations in the atmosphere, Pacific Northwest Laboratory Annual Report for 1971 to the USAEC Division of Biology and Medicine, Vol. 2, Part 1, pp. 107-109, 1972.

Young, . J. A., The particle size distribution of manrnade and natural radionucl ider, Pacific Northwest Laboratory Annual Report for 1973 to the USAEC Division of. Biomedical and Environmental Research, Part 3, Atmospheric Sciences, 16, 1974.

Young, J. A., T. M. Tanner, C. W. Thomas, N. A. Wogman and M. R. Petersen, Concentrations and rates of removal of contaminants from the atmosphere in and downwind of St. Louis, Pacific Northwest Laboratory Annual Report for 1974 W the USAEC Division of Biomedical and Environmental Research, Part 3, Atmospheric Sciences, 70, 1975.

Young, 3. A., C. W.' Thomas, N. A. Wogman and R. W. Perkins, Cosmogenic radionuclide production rates in the atmosphere, Journal of Geophysical Research, -75, 2385, 1970. Young, J. A., and W. B. Silker, The determination of air-sea exchange and oceanic mix1ng rates using 7~e,during the BOMEX experiment, Journal of Geophysical Research, -79, 4481, 1974. TABLE I

Average 7~eFluxes and Deposition Velocities Across the Pacific and Atlantic Ocean Surfaces

Deposition Latitude Flux in vc?locity Interval Atoms cm-*set-1 in crn sec-l

* Pacific. Ocean only

Average 713e flux across the ocean surface in the liorthern I-lemisphere. = 0.027 atoins ~irr-~sec-l.

Average deposition velocity in the Northern Hemisphere = 0.80 cm sec-l. FIGURE 1 SURFACE WATER '~eCONCENTRATION IN HUNOREDS OF DISINTEGRATIONS PER MINUTE PER CUSIC METER MEAN ANNUAL PREC 1 P ITATI ON (cm~)

FIGURE 2 MEAN ANNUAL DISTRIBUTION OF PRECIPITATION OVER THE OCEANS IN cm PER YEAR. (AFTER DROZDOV, 1953) 7 FIGURE 3 SURFACE WATER Be CONCENTRATION IN HUNDREDS OF DISINTEGRATIONS PER MINUTE PER CUBIC METER FOR MAY THROUGH OCTOBER

------MID I I I V V / 0 d I d 0 ----MLD t- \ - /" - Lo/O ,/ 149Ow :;;w 11/5/68 - 0 3/7/71 I --PALD CONCENTRATION = 0 AT 1WM I I I I . 200 400 0 XO 400 600 'i3e CONCENTRATION IN DPhl M-~ 7 . Figure 5. Vertical Profiles of Be Concentration in the Ocean 7~eCONCENTRATION IN DPM M-~

9 Figure 5. Vertical Profiles of '~eConcentration in the Ocean

Q 025 - 0

0 ..- - ooG- LJ a OZ0 I -' AUG- ocf /Q a .I A n A.

. n A/

0 A&-A , . I n ,I1 0.010 - nn 7 n NOV- JULY I O AJG - SEP - 9CT Q005 - biOV T6.ROUGI-I JULY . . 0 SGiJTkERN HEMI SPHERt

I 1 I I I A

0 16 23 30 40 50 60 . 70

7 Fir,ure 6. !?~ti'oof , Be siirfact. concentrati'on to total - .. -,.. . FIGURE 6A. The ratio of ~c surfacf waterconcentration to total inventory as a function of mixed layer ds;jti~.

LAT I TU DE 7 FIGURE 8 A:Y!QSPUERIT; BeC@NCENTRAIIO>!S FOR MAY AND JUKE

FIGURE 10. -1 DEPOSITION VELOCITY IN cm sec

FIGURE 12 DEPOSITION VELOCITY IN crn sei' FOR NOYfMBER THROUGH APRR Be-7 in the Study of Particle Mixing Rates in Nearshore and Lake Sediments

S. Krishnaswami, L.K. Benninger, R.C. Aller, and K.L. Von Damm Department of Geology and Geophysics

Yale University '

The vertical distributions of paxticle-associated radionuclides in the sediment column are governed by sediment accumulation and sediment-mixing processes. In general a measured depth profile reflects both modes of particle transport, but some combinations of transport rates and decay half- lives permit separation of transport effects. Aller and Cochran (1976) showed that excess Th-234 (half-life 24 days), produced in sea water by decay of U-238, can be used to estimate short-term particle-mixing rates in the estuarine sediments of Long Island Sound. Their approach is not feasible in most lakes because uranium concentrations in lake waters are typically too low. We have 'investigated the use of cosmogenic Be-7 (half-life 53.3 days) as a tracer for short-term particle-mixing in nearshore marine and lake sediments. Our study included three parts: 1) measurement of atmospheric deposition of Be-7 at New Haven, Connecticut (locations in Figure 1); 2) comparison of short-term mixing rates as deduced from distributions of excess Th-234 and Be-7 in sediments of Long Island Sound; 3) determination of short-term mixing rates in sediments of Lake Whitney, New Haven, Connecticut. Average daily atmospheric deposition of Be-7 in precipitation + dry fallout in monthly samples collected during March-November 1977 at New Haven 2 varied betwecr. 0 -03 - 0.10 dpm/cm -day. The reasons for the variability are incompletely understood; variable precipitation betwein sampling periods is a partial explanation. (deposition data in Figure 2) Sediment Be-7 distributions were measured in diver-collected box cores from stations 3 (15 m water depth) and 4 (13 m) in Long Island Sound (Figure 1). Both stitions are in a mud (silt + clay) facies, but average grain size ,is smaller at station 3. X-radiographs showed surficial sediments to be well- mixed at both stations. Be-7 was present in measurable concentrations to a depth of 3 cm at station 3 and to 2 cm at station 4 (Figure 3). Diffusion- analog particle-mixing coefficients calculated from the Be-7 data are about 2 a 3 x cm /s (station 3) and 1 x cm2/s (station 4). These coefficients are lower by about a factor of'j than coefficients calculated in the same way from distributions of excess Th-234 measured in the same sets of samples from the same stations. Samples from Lake Whitney were collected by a diver from the middle of a small basin in about 8 m water depth; the sediment was fine-grained and highly flocculent near the sediment-water interface. Again Be-? was present in measurable concentration to -a depth of'.3 cm in the sediment colqn (~igure4).

In this case, however, the surface (0-1 cm) concentration was so high relative I to those below that the profile could have been a sampling artifact. Alter- natively, mixing might have been accomplished by passive infilling of tubes opened by methane ebullition. Bioturbation by macrobenthos was probably not an important mixing mechanism because of oxygen depletion of the hypolimnion in the months preceding sampling due to seasonal stratification. 2 The average atmospheric deposition of Be-7 in New Haven (0.07 dpm/cm -day) 2 would support a steady-state inventory of 5.4 dpm/cm in a perfect collector. Sediment cores from Long Island Sound contain about half this inventory, consistent with either a mean residence time for Be-7 in the water column of about one half-life, or post-depositional loss of Be-7 from Long Island Sound sediments. The Lake Whitney cores contained about 5 dpm/cnL, much nearer the atmospheric delivery. A higher inventory of Be-7 in fresh water, as compared to marine, sediments could be due either to a shorter mean residence time for Be-7 in fresh water or to lateral transport processes in the lake or its catchment. High inventories of excess Pb-210 and Pu-239,240 in Lake Whitney sediments demonstrate the importance of lateral transport on longer time I Figure 1. Sampling locations in Long Island Sound (stations 3 and .4) and in Lake Whitney - reservoir, New Haven (inset). I

t $ 10- ++ - 9- -

8- -

7- -

6- - 5- ++ - 4- -

3- -

2- -

I- -

1 I 1 1 I 1 1 I - - 1 MAR APR MAY JUN JUL AUG SEP OCT NO\/ DEC

Figure.2. Daily average atmospheric deposition flux of Be-7 at New Haven, March through. November, 1977. Vertical lines show -+ 1 lr couritiig crrors in Be-7 determinations,

. horizontal lines show actllal collection iatervals (range 27 - 37 days). Be-7 CONCENTRATION, DPM/cM'~

7 Be in Estuarine Sediments N.H. cutshall and I .L. Larsen, Oak Ridge National Laboratory

7 Be is measured directly in surficial sediments in a low-level gamma-ray spectrometer. Nominal 200-gram samples are placed in a 550-crnJ Marinelli beaker

I on a Ge(Li) detector for 200-minute counting times. Using a detector with i 22% relative photopeak efficiency and 1.8 Kev FWHE1 resolution yields one count I per 630 disintegrations and a detection limit of approximately 60 pCi.

Surficial sediments taken from the Janes and Susquehanna Rivers and from

Chesapeake Bay have been analyzed. Concentrations of 78e faynd range from not detectable to 2.6 pCi/g corresponding to areal concentrations as high 2 as 1.7 pCi/cm . '~eis more frequently found in sediments taken from zones of high 'deposition rate than in low deposition areas. Highest concentrations are found in the high deposition rate area associated with the turbidity 7 maximum in the James. Whether or not salinity directly affects Be concentra-. 7 tions is not apparent in the present data. Be is present in sediments taken from fresh and brackish areas with salinities as high as 15 to 20Z,.

Higher salinity areas 'have not yet been sampled. Box CORES (0-2 CM) The behavior of 7~ein Sargasso Sea and Long Island Sound waters

Erik Aaboe, Eric Dion and Karl K. Turekian Dept. of Geology and Geophysics, Yale University, New Haven, CT.

7 Our initial interest in Be (54 day half life) was in its potent'ial

use as a tracer for the determination of bioturbation rates in estuarine

and lake sediments. The results of the study exploring this possibility

are presented in the previous section in the paper by Krishnaswami et a1

(1979). As part of that study it was necessary to have an idea of the 7~e

atmospheric flux in the New Haven area as a function of time. It has been

known for some time that at specific times of the year there is an increase 7 in the Be concentration in the troposphere due to stratospheric folding.

Krishnaswami et a1 (1979) however, showed that the expected tropospheric

increase at 40°N resulting from the Spring stratospheric folding event was not replicated by the precipitation flux in New Haven. This has since been confirmed in a second year cycle for the SEAREX qxperiment. The

closest correlation of 7~eflux for both years was with the precipitation

flux at the New Haven site.

In addition to the continuationof the 7~eprogram at New Haven as part of the SEAREX program,we have also monitored precipitation at Bermuda.

(This program was undertaken in part as an aid to modelling the 210~b flux

to the oceans) .

Krishnaswami et a1 (1979) showed that the standing crop of 7~ein

Lake Whitney sediments (Table 1) effectively was that predicted by the measured atmospheric flu whereas the three Long Island Sound cores were 7 markedly deficient in 7~e.The extended Be time series precipitation measurements at New Haven confitm this disparity. The difference in 7~estanding crop between the fresh water and marine

sediments could be ascribed to the greater efficiency of adsorption of reac-

tive species on particles in low ionic strength waters observed for other nuclides such as 137~s. If this indeed were the case then the remainder of 7 the Be expected from the precipitation should be found in the Long Island

Sound water column. It was' this expectation that activated our program for 7 determining Be in seawater samples.

Our procedure involved collecting a cubic meter (one ton) of surface

seawater with a pump into a large polyethylene "beaker". To this was added

BeSO carrier followed by coprecipitation with iron hydroxide. The precipi- 4 tate was further purified and prepared for counting in the laboratory as

described by Krishnaswami et a1 (1979).

In order to compare the Long Island Sound results with normal seawater

a survey of surface water of the western Sargasso Sea was made at the same

time (Fall 1978) aboard the Endeavor. The sampling locations are shown in

Figure 1.

Y The Long.Island Sound sample was collected several miles south of New

Haven in October 1978 using a chartered lobster boat.

Be-7 inventories in the North Atlantic: Silker (1972) has measured

7~eprofiles in the North Atlantic. The results indicate that 7~eis homo-

geneously distributed in the mixed layer, and decreases exponentially with

depth in the thermocline. If this decrease is treated as due to diffusion,

then,in the thermocline:

where C is the concentration of 7~eat depth z in the thermocline (z = 0 Z 7 at the base of the mixed layer), C is the Be concentration in the mixed m 3.

7 layer, X the Be decay constant and K the vertical diffusion coefficient. z 2 Silker obtained avalue for Kz of about 1.3 cm lsec for the North ~tlantic'

profiles. He was then able to determine the standing crop of 7~ein the

water column by summing the mixed laykr contribution, which was dominant, I and the thermocline contribution. 7 We have followed the same procedure using surface ocean Be concentra-'

tions, 'the depth of the mixed layer as determined by XBTS and Silker 's

K for estimating the standing crop in the thermocline. These results are z I shown in Table 2. The summary in Table 3 shows that the open ocean seawater standing I crops are compatible with those expected from precipitation data. Thus 'it is possible to make a reasonable e'stimation of the 'Be flux at any site in 7 the ocean by measuring the Be concentration in surface seawater samples

and using mixed layer depth data from XBTs. Whatever the meaning of K is I 2 7 , it provides a way of estimating the Be standing crop below the mixed layer.

As this represents about one third of the total standing crop considerable

error is allowed in the estimate without seriously affecting the total

standing crop estimate in most parts of the ocean.

I Be budget of Long Island Sound: The 'Be concentration of a Long Island I. sound water sample was found to be less than 240 dpm/1000 1 (Table 2), which corresponds to a standing crop of less than 0.5 dpm/cm2 for a mean depth of

20 meters. As reactive. nuclides such as thorium are known to have a shorter I residence time relative to removal to bottom deposits in Long Island Sound I than waters of the open shelf (Aller, Benninger and Cochran, 1979; Kaufman, I LS and Turekian, 1979) then this low standing crop of 'Be indicates that I. scavenging of this nuclide also is strongest in Long Island Sound. Indeed if anything the flux to the sediments would be expected to be greater than the flux estimated from atmospheric precipitation due to transport of addi- 7 tional Be from the shelf.

It has already been noted that the standing crop in the sediment cores analyzed for 7~eindicates a marked deficiency relative to that predicted from the atmospheric flux. We now see that the water column cannot satisfy 2 this deficien~y~addingonly 0.5 dpm/cm to the sediment standing crop of about 2 2 2 dpm/cm , compared to 4 to 5 dpm/cm predicted from the atmospheric precipitation data (Table 3).

Since, as noted above, Long Island Sound is more likely to be a sink for reactive nuclides than a source, we do not expect that the apparent

7~edeficiency is due to export from the system. Therefore two other possibilities exist. Either the three cores analyzed by Krishnaswami et a1

(1979) are not representative of Long Island Sound sediments generally, or there is another sink for 7~ein the Long Island Sound system.

Although we cannot categorically exclude the first possibility, the fact that none of the cores comes close to the expected value whereas the

234~hstanding crops for the cores analyzed by Aller, Benninger and Cochran

(1979) generally do, indicates that this is not the best hypothesis with which to proceed. The possibility of another repository exists at the margirsof the Sound. Here salt marshes abound and can act as specific nuclide scavengers. We have no direct information on the importance of thin sink for 7~eyet but it appears to be the most fruitful hypothesis to fdllow in the case of the missing 7~e. References

Aller , R.C ., L .K. Benninger and J .K. Cochran, Tracking particle associated

prbcesses in nearshore environments by use of 234~h/238~disequilibrium,

Earth planet. Sci. Letters, in press, 1979. 228T, Kaufman, A,, Y .-H. Li and K.K. Turekian, The rekoval rates of 234~hand

from waters of the New York Bight, Earth Planet. Sci. Letters, to be

submitted 1979.

Krishnaswami, S., L.K.. Benninger, R.C. Aller and K.L. Von Dam, Cosmic-ray . . 7 produced Be: its application to the study of particle mixing rates in

near-shore and lake, sediments, Earth Planet. Sci .. Letters, submitted

1979.

Silker, W.B., Beryllium-7 and fission products in the GEOSECS I1 wat'er column

and applications of their.oceanic distributions, Earth Planet. Sci.

Letters, 16, 131-137, 1972. Table 1 : Standing crops of 7~ein sedimenrs from Lake ~hitneyand Long

Island Sound (from Krishnaswami, et al, 1979)

2 Station 7~estanding crbp (dprnlcm )

Lake Whitney 5.25

Long Island Sound 3 4

New Haven precipitation March 1977-November 1977 Table 2: Standing.cr0p.s of 7~ein North Atlantic and Long Island Sound

waters

7 2 Be standing crop (dpm/cm ) -l Depth of I Be conc. Mixed Be.low c. Station Mixed Layer (dpm/lo3 1) layer ' mixed layer Tot a1 (m>

T- 8 245 ,524 12.8. (bottom)

Long island 20 < 240 . ' < 0.5 Sound

2 * Determined using K = 1.3'cm Isec. z Table 3: 7~esummary balance sheet

7 2 Be standing crop (dpm/cm ) A. Atmospheric precipitation

New Haven (March-November 1977 New Haven* (September 1977-May 1978) Bermuda* (September 1977-May 1978)

B. Sargasso Sea (see Figure 1)

Average

C. Long Island Sound system

Sediments (average of three cores) Water column Total Deficiency

* SEAREX data

,- - $8.

------* - - - - * - SC *

I I I I I LIIM----J~J'~

0 5 10 15 22 25 RAINFALL ( CM )

BE-7 F.LUX AT BERMUDA V) - (3 - 0 - x - I- - ZLP- Od - - x5 U - \=J - - C, \ .-- a2 - Q? - .4 - L - x 3 - 4 - 4ai h - I - a hl m 0-: , I , I I ,-L~~~~~~~~ 8' , 0 5 1J 15 28 25 RAINFALL C CM ) . 'OB~ stratigraphies in the Pacific sediment cores .

back to 4 million years ,

by Shigeo Tanaka

Institute for Nuclear Study University of Tokyo Tanashi, Tokyo 188 Japan ''Be ''Be concentration variations have been investigated in right Pacific sediment cores. Those sampling locations are listed in Table 1. Beryllium was chemically separated5) from the sediment

samples (8-35 g), and thebeta activity of ''Be was measured with

low-level needle counters 697) (background count rates : 3-15 counts per day). The results are illustrated in Figs. 1, 2 and 3. First purpose of this work was to establish the ''Be chronology during the past millions of years. This has been done with the assumption of constant ''Be precipitation, P, onto a sediment surface and of varying rate of sediment accumulation with time during the past 2.5 million years 13293). The absolute age, T(x), is uniquely determined for all values of depth x by the forinula of

where N(x) is a ''Be concentration at depth x and A is the decay constant of l0~e. The value of P in different locatioqs are mainly controlled by the sedimentation process in the oceans; this fact has been proved by various experimental evidences 1,223). The values of P in Table 1 do not reflect the fall-out pattern of the cosmogenic I nuclides in the atmosphere. The particle to particle scavenging or the sediment focussing.in the oceans would control P 1 > . The global average of P values seems to be between the estimated atmospheric production rate of 1.8 x 10-2 10Be ato:ns. -1 cm-2 .sec by Amin et a1.8) and that of 4.2 x atoms.crn -2 .stc-1 by Raisbeck et al. 9 The loge' concentration variations beck to 2.5 million years

are within the limits of %30% (Fig.l), although a larger scatter can be seen during the period of 1-2.7 million years compared to the last one million years (Fig.3 in ref. 1). In two cores KH70-2-7 and KH70-2-5 (Fig.2 and 3) taken frox North 'Pacific.basin, a catastrophically large change of more than

. . an order of magnitude in 'OB~ concentrations is seen at around . , I' , , .. 2.9 and 3.8 million years BP, respectively. This shows that a drastic change of 'OB~ flux or sedimentation rate happened during this period in the North Pacific. The possible causes for this variation would be dilution with detrital materials cosmic-ray intensity change paleo-climatic change change of sediment focussing change of sedimentation process . geophysical hiatus and others. Either microscopic observation or chemical element distribution does not show any clear evidences for dilution with influx of detrital materials. The change of an order of magnitude could hardly be caused by a change of cosmic-ray intensity or paleo- climate. In Fig.4, the 'OB~ concentration variations in two cores are plotted against age. The variations seem to have a time lag of 0.5-1.0 million years between two cores. At present time, the causes for the catastrophic change are in open question. References

1. S. Tanaka and T. Inoue, ~arthPlanet. Sci. Lett. (in press) 1979. 2. T. Inoue and S. Tanaka, Nature 277 (1979) 209. 3. S. Tanaka et al., Earth Planet. Sci. Lett. -37 (1977) 55. 4. S. Tanaka et al., (in preparation) 1979. 5. T. Inoue and S. Tanaka, INS report INS-TCH-11, 1978. 6.' Y. Fujita et- .al., .Nucl. Instr. Methods -128 (1975) 523. 7. N. Takano et al., Int. J. Appl. ad. Isotopes (in press).1979. 8. B. S. Amin et al., Geochim. Cosmochim. Acta 2 (1975) 1187.

9. G. M. Raisbeck et al., (submitted to Nature) 1979. Table 1. Cores investigated

Core No. ~ocation P ( 'OB~ preclpitatlon rate) Ref.

...... , ...... q3nv1n-'2 -3 -I 1. KH 70-2-7 33ON ijoP" : 1 11 Figure Captions

Fig.1 1°8e concentration Variations with time in five.~orth Pacific cores; the numbers (1-5) denote the sample cores whose sampling locations are shown in Table 1. Mean

decay lines are shown by solid lines, with 230% limits of variation by dotted lines (taken from ref. 1).

Fig.2 IUBe concentration variations in core KH70-2-7.

Fig. 3 1°8e concentration variations in core KH70-2-5.

Fig.4 Comparison of the l08e variations with time in two cores, KH70-2-7 and KH70-2-5. ''Be Age ( million, years ) 600 800

DEPTH (cm) i ' -. . - - ;A lo K H 70- 2-5 -

D . .I:(:,. . . :,m,1, . .. .,);'.:.r.l:".. ,, *"'\ 6'*...:" : . . ...; . .: l..3: .* . . ., , . .. , ...... i'. rn I I I f f'.. i E 8 U. .. E n mu '0.- . '' - \ 0.1 aJ m w 0 v age (rn.y)

0.5 1.0 1.5 2.0 2.5 3.0 3.5 4. 0 I I 4, 1 ,I. 0.01. I I I, 0. 200 400 600 800 1000 DEPTH (cm 1 AGE (rn.y.1 ''BE MEASUREMENTS WITH A CYCLOTRON AND THEIR APPLICATIOS TO THE STUDY OF AIR AND I\;.4TER bL4SSES IN THE ENVIROS~~EF;T G.M. Raisbeck and F. Yiou - Laboratoire Ren6 Bernas, Centre de Spectrometric NuclSaire et de SpectromGtrie de Masse, 91406 Orsay, France.

As is evident from the of this meeting, c~smogenic7~e has been used extensively as a tracer in studies of the movement of air and water masses in the environment. Among the properties that make this isotope useful in this respect are that it has a continuous, well understood source function, is globally distributed, is fairly abundant (because it is formed from nuclear interactions on oxygen and nitrogen), and has a $ray decay mode whicg is convenient for measurement. The isotope 'OB~(half-life = 1.5 X 10 years) shares a11 of these properties except the last. In addition, its essentially infinite half life (compared to the time scale of the air and water movements in question) ma.ke it potentially very valuable, either alone or in conjunction with 7~e,for such studies . Until recently, detailed studies with ''Be were precluded because of its very low level of radioactivity. We have recently developed a cyclotron technique (1,2) that now permits us to measure as few as 10 7 atoms of this isotope, thus making its determination in environmental samples completely feasible (this quantity of loge is found for example, in 400 ml of rainwater, 15 1 of ocean surface water, 250 m 3 of ground level air, and less than 10 m 3 of stratospheric air). 7 One of the main limitations in using Be as a "tracer" of environmental samples, are its large concentration variations, in both time and place. These can be due not only to decay , but also to mixing, to "washout" (in the atmosphere) or adsorpion on particles (in marine environments). Because it has identical chemical propertips,, ''Be will have the same variations , evcep t for decay. Thus the ratio Be / 7Be will be much more constant, and ths variations can be interpreted directly in terms of the "agew of the sample in which it is measured. This, in turn, will often be useful in identifying the origin of the sample. As a specific example, we consider the problem of identifying the movement of stratospheric air into the troposphere. Because the residence tine of aerosols in the stratosphere is thought to be about an order of magnitude longer than in !be troposphere ( q, 1 year compared to 30 days), the equilibrium ratio of Be/ 7 Be(in the absence of transfer) can be estimated to be about four times larger in the str~bospqerethan in the troposphere. (In fact, an accurate measurement of Be/ Be in the stratosphere, combined with knowledge of its production ratio, would give an excellent estimate of stratospheric rr8ideqce time for these aerosols). As we shall show, the measurement of Be/ Be should. thus be able to unambiguously identify even a few.percent contribution of stratospheric aerosols into the troposphere (or into precipitation). One obvious application of such a procedure would be in trying to resolve the question of the importance of the input of I stratospheric aerosols in polar regions. (A question that is of direct interest to us for interpreting possible 'OB~variations in polar ice cores (3)). In addition, of course, the widespread monitoring of this ratio at different locations and times, could be expected to give very valuable information on the global,.featuresof stratospheric movement into the troposphere. 7 Similar considerations to the above will hold in the use of '~e. measurements in ocean water.

References 1 - G.M. ~ais'beck,F. Yiou, M. Fruneau and J.M. Loiseaux, ~cien'ce,-202, 215, (1978). 2 - G.M. Raisbeck, F. Yiou, M. Fruneau, M. Lieuvin and J.M. Loiseaux, Nature, -275, 731 (1978) ; Earth Planet. Sci. Lett. -43, 237 (1979) ; Geophys; Res. Lett.. (in press) Nature (in press). 3 - G.M. Raisbeck and F. Yiou, 10th International Radiocarbon Conference, Bern-Heidelberg, Aug. (1979) ; Conference on the Ancient Sun, Boulder, Colo, Oct. (.1979). Conference on "Natural Radioactive Isotopes of Beryllium in the Environment" October 1-2, 1979, New Haven

Department of Geological Sciences University of Southern California Los Angeles, Calif orlr-lia 90007

The cosmogenic l0~ewas first identified mcre than 20 years ago in deep-sea sediments [l]. To date, the published results on 'OB~ in the marine environment include only about 110 analyses on 18 sediment cores (15 of which from the Paci- fic), less than 50 analyses on 11 manganese nodules (all but one from the Pacific), and 2 measurements made on seawater . samples. However, more than half of these measurements were nade within the last 3 years or so. .And the pace is certainly to be further hastened by the accelerator techniques C23 and refinement in ultra low-level beta countin,g L31. 10 The two best- know^ applications of marine Ee are: (1) as a geochronological tool and (2) as .an indicator of cosmic- ray intensity and/or geomagnetic-field variations in the past. In this connection, the following summaries are made from the data available. (1) Input Rate of ''Be to the Ocean (a) Global average: The average 10Be deposition rate esti mated from 'OB~ contents. in 11 dated (by the uranium- series and paleomagnetic methods) sediment cores is -2 -1 2.3*10-~atoms cm sec (Table 1). This is corn- parable to the atmospheric production rate of 1.8 atoms cm - 2 sec-1 obtained by model calculations of cosmic-ray interaction rates C41. There is the possibility that both estimates represent lower limits I51. But it seems clear that the mean residence time of "~e in the ocean (with respect to removal to bottom sedi-'

ments) should be short compared to its mean life of about 2 m.y. (b) Latitudinal variation: How much the ocean circulation does' to homogenize the latitudinal input variations of ''Be C61 is unclear. Though there appears to exist a 10 correlation between Be deposition and its stratospheric fallout pattern, the data are too scattered to be 'conclusive (Fig. 1). Note further in Fig. 1 .that

i the low l0I3e deposition in the two Arctic cores (FL- 106, -277) can be explained by causes other than atmospheric input variations C71. (c) Time variation: During the past 2.5 .m.y., va~iations in the ''Be input as reflected by the sedimentary rec6ds (Fig. 2) do not exceed +40% for period of -10 5 years. This may also be taken as the upper limit for fluctuations in cosmic-ray intensity C83. (.21 Ocean Water No vertical profiles of 9~eand 'OB~ are available. From the consideration of "concentration factors" for 'Be (Table 2) 191, the oceanic distribution of the two isotopes may resemble Ba and Ra more than the other Group IIa elements Ca, Mg and Sr, in being affected by the particulate (biological) cycles. If so, a two- to five- fold increase of Be concentration from surface to the deep ocean is possible, and the deep concentrations should reflect bottom water renewal time in different ocean basins (e.g., [Pac.] > 1Atl.l). The residence time of Be in the surface waters is of the order of 2 - 9 years, that in the ocean as a whole is3water mixing time of -10 3 years.

(3) Deep-Sea Sediments There is a more or less "uniform" ''Be concentration

of about 2.5-10' atoms cm-3 ( 4.5 dpm kg-') in surface sediments. This is suggested by the decent to pood correlation between ''Be deposition rate and sedimenta-

tion rate (Fig. 3) - more decent than the 'OB~ deposition . ' rate - latitude correlation (if any) shown in Fig. 1. The Fig. 3 correlation points to the danger in assigning a constant ''Be flux based on either the average global input

value or its latitude-corrected version in zgc determiria- .tian, as some authors.have done. Dating sediments using such so-called "constant 'OB~ flux model" should adopt the method of ~ackett[lo] for 231~,a.That is, the actual standipg crop of ''Be in a core should be determined. Most of 'the 'Be in sediments (2-3 ppm) are of detrital origin. (4) Pelagic Manganese Nodules The slow growth rates of nodules make the relatively long-lived ''Be more suitable.as a chrondmeter than nuclides like 230~hand 231~a. Exponential decrease in

''Be ''Be concentration . .with depth is a common feature in nodules, giving rates of growth on the order of mmls/m.y. Nodules contain 2-9 ppm 'Be, and in a given nodule, 'Be contents are rather uniform. 'Be in nodules is primarily of hydrogenous origin. Because of their slow accretion rates and their containing small amounts of detrital dilutants, nodules have much higher specific ''Be activi- ties than sediments. Typical values are tens of dpm/kg.

. But the total inventory of ''Be in nodules are small - only about 1/20 of that produced in the atmosphere is collected by them (Table 3, Fig. 1). The suggested ''Be flux - Mn flux correlation C2bl and its .implications remain to be explored.

References Cited 1. Arnold, J.R., Science 124 (1956) 584. Goel, PiL., D.P. Kharkar, D. Lal, N. Narsappaya, B. Peters and V. Yatirajam, Deep-Sea Res. 4 (1957) 202. 2. a. Raisbeck, G.M., F. Yiou, M. Fruneau and J.M. Loiseaux, Science 202 (19781215. b. Turekian, K.K., J.K. Cochran, S. Krishnaswami, W.A. Lanford, P.D. Parker and K.A. Bauer, Geophys. Res. Lett. 6 (1979) 417.

3. Fujita, Y., Y. Taguchi, M. Imamura, T. Inoue and S. Tanaka-, Nucl. Instrum. Methods 128 (1975) 523. 4. Amin, B.S., D. La1 and B.L.K. Somayajulu, Geochim. Cosmochim. Acta'39 (1975) 1187. 5. Raisbeck, G.M., F. Yiou, M. Fruneau, J.M. Loisearx, M. Lieuvin and J.C. Ravel, Submitted to Nature (1979). 6. Lal, D. and B. Peters, Handbuch der Physik, K. Sitte, ed. (Springer, Berlin, 1967) Vol. '46/2, 551. 7. Finkel, R., S. Krishnaswami and D.L. Clark, Earth Planet. Sci. Lett. 35 (1977) 199.

8. Inoue, T., S. Tanaka, Nature 277 (1979) 209. 9. Lowman, F.G., T.R. Rice and F.A. Richards, In "Radio- activity in the Marine Environment" (Nat'l Acad. Sci., Wash. D.C., 1971) 161. 10. Sackett, W.M., in "Symposium on Marine Geochemistry" D.R. Schink and J.T. Corless eds. (Univ. Khode Island, 1965) 29. CORE LAT I BAS IN SED, RATE loB5 FLUX (CM/KY) (10-nToms cM-2sEc-1)

NOVA- 111 -16 o0 PAICFIC 0.16 1. 4 KH-68-4-18 2'~ PACIFIC 0.35 2.2

I .. NOVA- III -13 4'N ' PACIFIC 0.14 0.6

I MSN-147G 8'N PACIFIC 0..'19 . 1.3 20°s PACIFIC 0.24 0.7

21's INDIAN 0.10 1.9

ARI ES-40P 30°N PACIFIC 0.37 3,2

V20-108 45'~ PACIFIC 1,13 10

MSN-96 58's PACIFIC 0.26 2.3

FL-106 78'~ ARCTIC . 0.24 c.95

FL-277 84'~. . ARCTIC. . .. 0 .09 S.6. . TABLE 2

CONCEKTRATI ON FACTORS

PHYTOPLANKTON ZOOPLANKTOPI

BE 1,000 15 BA 17,000 900 RA 12,000 190 P 34,000 13,000 AL 100,000 100,000 NODULE LAT , BAS I N GROWTH RATE - ~OBEFLUX !MM/MY) ATOMS CM -2SEC -1)

A47-16-3 9'~ PACIFIC 2 0,044 TECHNO-1 13Os PACIFIC 2,8 0,054 C57-58-2 15'~ PACIFIC 2 0,067 DODO 15-1 lgON ,PACIFIC 3;0 0,044 MN-139 20°N PACIFIC 1,3 0811

ARIES J2D 21'~ PACIFIC 3,3 0,067 ARIES 13D 21°N PACIFIC 2,4 0,12

ARIES 15D 21°N PACIFIC 2, 3 0,067 TRIPOD 2D 2lobl PACIFIC 6.13 0'.160 RC16-Dl0 28's ATLANTIC '4,5 0,086

ZETES. . 3D . . 40'~. .. PACIFIC...... 1,3 0,120 Figure Captions

10 Fig.. 1: Measured depos-tion rates of Be in marine sediments (triangles') and Mn nodules (circles). The calculated rates (curve) is from C71. Mean global ''Be production rates estimated in' 141 (long dashes) and in C51 (short dashes) are shown.

Fig. 2: ''Be in two Pacific cores. ~ottedlines represent 340% limits of variation. From C81.

Fig. 3: ''Be deposition rates vs. sedimentation rates in 11 cores, with linear regression lines.

"~eAge ( million years )

VOL. 6, NO. 5 GEOPHYSICAL RESEARCH LETTERS MAY 1979

THE MEASUREMENT OF "BE IN MANGANIISL.: NODULES USIN(; A TANI)I:!I VAN DE GRAAFF ACCELERATOR

K.K.Turekian and J.K. Cnchran Dept . of Geolc~gyand Ge(1pl1ysic.s Yalc Universi~\,New Haven. Ct. Oh520

S. Krishnaswami Physical Research Labt~ratory.Nivrangp111a Ahmedabad 380009, India

W.A. Lanford. P.D. parkcrand K.A. B.111cr Wright Nuclear Structure Laboratory )'ale University, New Haven. Ct. 06520

.-lhsrrat~l. Tl~cp~crall~ rate of a niaIIgsnesc 110tlt11e~I~IIII the S{~uthAllsrilir. tcrniinal wl~crcit is stripped by passin; tlirot~pliI~VO 1Ili11 C:II~#BI~fali!. I,! OCC~II(Kit) G~andeRidgel 113s hcen deterniined using a "BC pr~~lilcas form i3c4+ wl~iil~is IIICII accelerated OUI of tllc taliden1 and monicillum measured with a Tandcni Van dc Graal.faccelerator. This rate is 4.5-mrnlhfy. selected by the 90' analyzing magnet. The terminal \,oltagc and anal\.zi~lg A correl3tioli may exist between tlic "Be flus and thc MI) flus to nodules. magnets are set to accelerate '"BC to 40 hlcV. TIIC(I,-I,*:II{III ~ystenic~~n.c8>!1:?

It lias brcn hic1\\*11since IIIC pioneel ing \~OIk .>iArnold and his COWOI LCIJ llle curlent5 oi i111e11bcbca~iis (buill 25 'BCJ ~IIL!8 Al:.L ~ounte~tcIc~;tq>c (Arnold, 1956: Merrill el al., 1960) and Lal and his coworkers (Goel el al., which is used to count directly the individual parlicles in ver). well bcsnis 1957) tha~cosmogenically produced "BC miph~he a useful radionuclidc (such as "Be). In tliis counter telescope. particles pass throudi a thin (,>I tlic SI~ILI! 1ii:11i11eS!.SICI~~S. 11sl~alt !ifL. i> ncnv ~II~I\\,I~I<> he I ,5 s 10" (AI:) siliso~lSII~~~I~C-~AI~~CI d~t~ilo~(17 pni tl~~cll211d 111~~11>i,'/i i:; r \,eals jl'iou and Ibisheck. I<)::) and i~sglc>bal proJuction rate appeals to thick (t) dc~cciur(- 150 pnl thick). Becausc \he amuulit CIIL.I~~I~~JI !!i be about 1.8 x lo2 atoms/cm2 sec (Aniin, Lal and Somayajulu. 1975). the AE detector by a particle of a given 11>ta1energy is a function of the The 3~1~31flus to any spot on the ocean surf~ceis determined by the nuclear charge and mass of tlie particle. this counter telescopc ~311 suppl! WIG to 11131 sp>I iron1 tlie stratosphere wllcrc ''Be is primarily uniquely identify any ligllt isotope which hasenougli energy to pass 1111,~itpli prodo,-ed. This flus is espected to be 3 ni3simum bet\\wledge of thc disrribu~ion of fallout from siratospherically injected (AE) is plotted agaiiisr the total energy of the particle (AE+E). TIIC bomb radionuclides. Thc localized flus of ''Be to the ocean floor. hwv- diagonal line corresponds to particles that are completely s~oppcdi:; ll!i. ever, is deterniined by tllc eslent of Ilonlorcniz3tion in llle tvster colu~~i~iI'rc>n~ dr~eaor. so 01a1 AE = (AE+E).(Once 3 psrticlr has suil'iiient e~ic~ned at an angle oi 30' with resp.-~.l I,> lilt used to infer 311 accuniula~io~irale of that deposi~.'Illis has been at~emp~cd incident beam and detected only a small fractii~nof tlie 11uill.i proJu.~~!i!: on deep-sea cores (Amin. Lal and Somayajulu, 1975; Goel el al.. 1957: reactions between the beam and target nuclei. Careful esaniinatii~noi tlii; .Amin. Uarkar and Lal, 1966: Tanaka el 31.. 1968: lnoue and Tanak3. spectrum shows tlie separ3lion of lithium. ber).llium sfid hor~~iinlv -1976. 1979: Somayajulu, 1977: Finkel, Krishnaswami and Clark. 1977), various isotopes. In particular a weak band of ''Be events can be seeti cln but its most resounding success has been in setting the chronology of tlie upper right hand side of tlie niucli niore intense bind of el3s1ic and manganese nodules (Somayajulu. 196?: Krishnaswami. Somayajulu and inelastic 'Be events. Moore, 1972: Bhal el al.. 1973) Sharma and Somayajulu, 1978; Ku, During measurenienls of the "Be bcani intensit!.. the deteao~telC- Ornula and Chen. 1978: Krishnas\vanii el al.. 1978: Cuichard. Reyss and scope system is moved 0'.'and 3 1 7.25-nig!cni2 aluminum I;,il is pi11 ill Yokoyama, 1978). I'I,>IIIof the detcctt~rs. This LI~S~II~CIis si~itkic~i~l! tliick I,> s1.q> 211 The number of nieasurements on manganese nodules is small-limited Ilca\.icl. ist>topcs in tlic hcanl will1 JIIC prtbpcr ni~>nicnti1ni~;l1arg1'13:i~j IS> by the sample size required, tlie tedious radiochemical separation and the get througll tlic analyzing niapnet. e.g.. 52-Me\' I2c.77-!I<\' ''0. ~II. need for excellent low level p counters and long counting linics. Eve11 wit11 40-MeV "B. If these nuclides \\,ere prcscnt ill the i~lciden~bean1 e\en 11 these restraints there are now I0 nodules, all of them froni the Pacific. 111a1 tllc level of 10" timcs thc 9~eintensity and werc slllr\ved to strike tllc have been assayed for "Eie by several different laboratories using radio- counter telescope, they would have sw~nipedtllc detectors. Any *BC it~lis chemical techniques. (44.4 MeV:) which do get throudi both tllc 90~-anslyz1n~niagnel and tllc As was first shown by Raisbeck et a). (1978) on the Orsay cyclotron. aluminum absorber are easily distinguished from the ''Be ions (30 hlz\') "Be measurements can now be made niorc rapidly and on smaller samples on the basis of the separation in Figure 2 and their ve? diiferen~rcsiJu~I by using nuclear accelerators. We report the results of measurements made energy. using the Yale MP tandem Van de Graaff accelerator to analyze "Be in "~e/'~)c ra~iomeasurements for each particular sample were made hy manganese nodules. alternating observations of thr '~e(Faraday cup) and "Be (AE-F telc. Ihe tandem Van de Graaff is shown schematically in Figure 1. The ion scope) beam inte~isitiesfro111 111il1~niple. The aIternati~>nsfi~>ni 'BC 10 source is a cesium sputter source (Middleton and Adams, 1974). A samplc "Be to 'Be etc. were made by changing the niagncli; IicIJs in IIIC ion consisting of 10 to 30 mg of Be0 is pressed onto the inside of a stainlcss source inflec~i~rnnlsgncl. the rnagnelic quadrupalc jus~af1t.r 111z az,.zlt*~- steel cone. (Up to 12 different sample cones can be mounted in the ion ator,and the 90°analyzing magnet, while keepine, the accelerlting potential source at one time so that different samples can be analyzed by simply constant. The terminal voltage ofthe accelerator was regulated to k 2 kc\' rotating the sample changer wheel.) The sample is then bombarded by a using the generating voltmeter. Typically sin nieasurenients were made of cst beam, producing a variety of negative ion beams. By adjusting the each isotope for a particular sample ill 3 period or 30 to 45 minutes. The ion-source magnet, the 'B~o'beam or the 'OB~O' beam is selected and rapid comparisons between ''Be and 9~eand the rapid comparisons injected into the accelerator. In the measurements reported here, the between samples (with repeated checks) are important to monitn~any typical 'Be0' beam injected into the accelerator was 100 to 300 namps. changes in the conditions of the ion-source. tlle acceler;lto~.or the beam The negative BeO- beam is then accelerated to the positive high potential transport system. Our experience has been that the ratio of "Be counts (AE.E telescopc) Copyright 1979 by the American Geophysical Union. to the '~ebeam (Faraday cup) stays constant to within rhc statistic~l

Paper number 9L0572. 0094-8276/79/059L-0572$&l. 00 416 Turekian et a).: Manganese Nodules

YALE MP TANDEM ISOTOPE ANALYSIS SYSTEM

Fig. 1 : A schematic representation oi the isotope analysis system based on he Yale MP tandem accelerator. Sa~iiplcsto be 31iaIyzcd arc mtrulited in tlie ion source (1111 lel't) 3115 bonibsrded wit11 CS iolts. Kesulti~igBcO' icil~s are mass selected and accelerated to the high voltage ~erminalof the tandem accelerator. At the terminal, these ions pass through thin carbon foils re- nioving their electrons fcrrmi~igpositivc ions. Thcsc ions arc accelerated 11, tlie detectiull syste~il(~111 ridit) consia~it~gt~l'a reniovahlc Faraday CUP 311d a AE-Ecounter telescope. See tent.

aicuraiy of tlic '"B' counts (- 207) 31111~i~lglltlie iltle~~sityt~ibt~tll hc;~nis may change by as much as a factor of two or more in a period of hours. < Calibration standards were prepared by processing, separately, two Fig. 3: Plot of 'OBei9~eratios as determined by 0-counling of samples whole small nodules for "BC deterniination by 0-cou~iting.This was done prepared from whole manganese nodules using 9~e0carrier wirh subse- to simulate tlie chemical prticessinp hc used in tl~cal~alysis oi actual quc1.8t di111tiannswit11 '~~0. ' samples. With tlie amount of 9~e0carrier addcd. the resulting "i3ei9~e ratios were determined to be 6 s 10"~and 6.5 x 10-1° wliicl~were treatcd The separation of ''Be was made hy dissolution in 6S IiCI ill tltc ' as indistinguishable on the hasis of cou~itingerrors. These standards were presence ni 1i,O2. To tlte solution BeSO, carrier was added equi\'alent to diluted with 'Be0 and tlic 'OBet9Be ratios measured to obtain the working 36.6 nig BrO. Subsequent separation of Be was made essentisll! io!l,w i::~ cur\.? shc1wn in Figure 3. These results show that within the 10~e/9~cthe procedure of Amin. Wi3rk3r and LI (1966) except 11151 nc> sep:!r3iit>lii range from 3 x lo-" to 6 x 10"~the accelerator response is linear. The from U, Th or Ra were specificall!; made. Tlie final precipitates \\'err' present level of "Be sensitivity is about I count per 100 seconds: on tlic chemically pure Be0 but were not radioclieniisally pure. basis of our comparisons wit11 the calibration standards. this corresponds The accelerator measurements of the ''Be/'Be ratios in tile wniple; to an abundance of - 4 s 10' atonls of ''Be in the sample, correspondin; were made relative ttr the radiurnetrically determined staiidard whicli \v3s to'a "~e/~Beratio of 5 I x 10"' for the sample cones used. N'e belicvz rull at the same time in tl~emsrtner alread! described. Thc rzsul~rrr.:rc that this sensitivily is limited by cross-contamination among the samples converted to ''Be concentrations. The plot of the log 10~e/gwith deptli mounted in the ion source. We are presently adding baffles to the source in this nodule is shown in Figure 4. The growth rate oi tl~rnodule is to reduce this effect. It is expected that this sensitivity can be improved b! 4.5 mm/hly. Tl~isis !lit first 'OB~growth rate deterniined ior an Atla~ltii as much as a f3c1or of I00 i~tthe future. Ocean manganese nk>duls. For our first measurements we have pr~ccedcdconservatively. A hrgc Figure Sshowsa plot of the flus of "Be fluses than w

Table 1: "Be measurements on nodule RC16-Dl0 (28'25's. 40~15'\\'. -l.;SS ni I

'0~c/9~e* Attinis Sampling Sample (with 36.6mg 10Bc/g dpnt Depth Size Be0 carrier) lo error* nodule 'OBe/kg (mm) (grams) (lo-"') (%) (10") nodule Fig. 2: A typical AE-E calibration spectrum for the counter telescope detection system. This spectrum was recorded with a 44 MeV 9~ebeam incident on a 'OB target with the telescope at 30'. The elements He through B are clearly separated and labeled; the separation of the isotopes of Li, Be and B can also be discerned. 'Based on ten minute count Turekian et a].: Manganese Nodules 419

able 2: I0~ein manganese nodules

I Water Growth 'OB~flux Depth rate (lo4 atoms/ Code Nodule (mj (mm/My) em2 y) (Fig. 5) Rererencc

Zetcs 31) 3000 1.3 3.8 A Bhat el al 40'1 6%. (19i3) 170°?0'li Tripod 21) 3000 6.3 5 .O A Bl~atet al 20'45'~. (1973 113'47~ 1)01)015.1 4160 3.0 1.4 B Somayajulu 19~23'~. (1967) 162'20~' ARIES 15D 1278 2.3 2.1 C Sharnis and 20~47.3'X, Soma!.iijulo 173'20'~ (1 9%)

ARIES l2D 1623 3.3 2.4 C Sharma and ' ?0@45'X.' SU~XI~>!!!III 173'26.4'~ (197b;) TECHNO 1 4020 2:8 1.7 D Guichard et al 1~'o?'s, (1978) 148'62%'

Fig. 4: Plot of log ''Be concentration versus dpeth in manganese nodule A47-16-3 5049 -2 1.4 ' E Krishnaswanii from the d3t3 of Tablc I. 9O?.3'N, et a1 (1978) . 151°11.41\' correlation hetweeli 11odule gruwtll rate and flus of 101)2 to the grouting C57-58.2 4635 -2 2.1 E Krishn;lswani~ nodule.we can further relate the 'OBC flux to the Mn flux by the following 15O19.5'~, et al (,1978) transformation. We assign an average density of 2 g/cm3 to the nodules 125'54.41~ and an average h11i conce~itr~li,~~iof 225. This yields a ''Be alom llus ti1 ARIES 13D -3000 2.4 3.7 F Ku ei a1 (19761 hln atom flus ratio or' l .5 x 10'I. 20°45'~, 'Ihe average ''Be areal flus to tlie ocean floor should be determined by 173~40'~ the global areal production rate (= 5.4 x 10~toms/cm2y). The Mn flus estimation based on sediment and MI) nodule analysis made hy Bender. Ku hln-139 3910 1.3 3.5 F Ku et a1 (197s) and Broecker (1970), tirisli~iaswanii(1977), and Bender, Klinklianimcr ?oOO1'N: and Spencer(1977) is about 10Ib atoms/cm2 y and the value inferred from 138'36~' hydrothermal sources by Weiss (1977) using the '~ehlncorrelatio~l in RC16-Dl0 438s 4.5 2.7 C Present work deep-sea thermal areas and the '~eflus Illrough the oceans estimated by 2s0-"s> Craig, Clarke and Beg (1975) is about three limes this amount. This yields 40'45~: an oceanwide average 'OBe/hln flus ratio of 5 x lo-" to 1.7 x 10". within a factor of three nf the manganese nodule estimate. Since most of the All values are calculated using a ''Be half life of 1.5 r lotiy and a nodule ''Be and Mn fluxes are to the sediments rather than the nodules, this density of 2 g/cm3. No corrections for turning over of the nodule during provides no proof of tlie singular association of 'OB~deposition on the growth have been made. oceali floor with hln depnsitio~ibut it does suggest it as a possibility. Ack~~owledgniet~rs:We acknowledge the curator of the Lamont-Doherty Geological Observatory collections and grants ONR-NO00 14-75-C-02 10 and NSF OCF-76-18049 to LDGO Cor providing us with nodule. Research supported in part by U.S.D.O.E. grant EY-76C-02-3074 and NSF grant OCE-7602039.

References

Amin. B.S.. DP. Kharhr and D. L1. Cosmogenic 'OBC and 2b~lill marine sediments. Deep-Sea Res..l3. 805824. 1966. Amin, B.S.. D. Lal and B.L.K. Somayajulu, Chronology of marine scdi- ments using the ''Be method: intercomparison with other metlii)ds. Ceochin~.Cosmochini. Arra.39. 1187-1 192. 1975. Arnold. J.R., Beryllium-10 produced by cosmic rays. Scie1ri~c.124.581 585, 1956. Bender, M.L.,T.L. Ku and W.S. Broecker, Accumulation rates of manpnw in pelagic sediments and nodules. Earrh ~la~iri:'~ri.1.rirers.R. 142-14s. 1970. NODULE GROWTH RATE (cm/My) Bender, M.L., G.P. Klinkhamnler and D.W. Spencer, Manganese in seawater and the marine manganese balance, Deep-Sea Res..24. 799-812, 1977. Fig. 5: Plot of ''Be flux against the growth rate for manganese nodules Bhat, S.G., S. Krishnaswanli. D. Lal. Rama and B.L.K. Soma).ajulu. Radio. listed in Table 2. metric and trace element studies of ferromanganese nodules, in Pruc. 420 . Turekian' et al. : Manganese Nodules

Symp. ~ydro~eochemist~and Biogeochemistry, I (The Clark Co.. bI. D. and B. Peters. Cosmic ray-produced radioactivity on the earth. in: Waglington, D.C.). 44346,.' 1973. Hand buch der Pliysik. K. Sitte, Ed. (Springer, Berlin, 1967) vo. 4612. Craig. H.. W.B. Clarke and M.A. Beg, Excess 'HC in deep water on the East 551-612. Pacific Rise. Eorrlr Phrrct. Sci. l.errers.26. 125-132, 1975. erri ill, J.K., E.F.X. Lyden, M. Honda and J.R. Arnold, The sedimentary Finkel, R., S. );rishliasu.nmi and D.L. Clark. ''Be in Arclic Ocean sedi- geochemistry of the beryllium isotopes, Geochir~r.Chs~rroc'hirrr. Acro. ments. Earrlr Phrrcr. Sci. I,rrrrrs..?5, 199.204, 1977. 18. 108-129. 1960. Goel,P.L.,D.P.Uarkar.D. LI,N. Narsappaya, B. Peters and V. Yatirajam, Middleton, R. and C.T. Adams, A close to universal negative ion sourur, The beryllium-I0 co~icentratianin deep-sea sediments, Ilrrp-Sea Rcs..4. Nrrcl. Insr. Merh.. 115, 329-336, 1974. 202-210. 1957. / Raisbeck, C.M.. F. Yiw: M. Fruneau and J.M. Loiseaux, Beryllium-lO Guichard, F.. J.-L. Reyss and Y. Yokoyamn, Growtl~rate of mangahesc mass spectrometry with a cyclotron. Science.202, 21 5-21 7. 1978. nodule measured with ''Be and "~1,Norure,272, 155-156, 1978. Sharma, P. and B.L.R. Somayajulu, Growth rate and compositic~nof two Inoue. T. and S. T~nnka,I0Mc in marine sediments.,tbrrh Plorrer. .Ti.;. ferromanganese nodules from the central North Pacific. in Sur la ?enhsc I.errcr,.?V, 152.160. 1970. des nodules de manganise (Colloque internationale du CNRS, h'o. 289. Irioue. T. and S. T~naka.''Me in marine sediments. Earth's environment 1978) in press. and cosmic rays. Naf1trc;277. 209-2 10, 1979. Somayajulu. B.L.K., Be-I0 in a manganese nodule, Scierrcc,l56, 1219- Rrisli~iasu*~nii.S.. Authipenic rans sit ion elements in Pacific pelagic clays. 1220, 1967. (;c,~~.lrit~r.(i~srrroc~lrirr~. Ac.r1~.40. 425434. 1976. Somayajulu. B.L.K.. Analysis of causes for the beryllium-I0 \rariations ill Krishnaswmii.S., B.L.R. Somayajulu and W.S. Moore, Dating of manganese deep sea sediments. Groc.lrirrr. Cosn~oclri~tr.Acro.41. 909-913, 1977. nodules using berylliunl.10. in Papers from a conference on Ferro- Tannka, S., K. Sakamoto, J. Tahgi and M. Tsuchimoto, Aluminum.26 and . ni31ipanese deposits on the ocean floor. D.R. Horn. Ed. (National beryllium-I0 in marine sediment. S~~ierrcc.l60.1348. 1968. S;icn,.? F,NII~~:I~IOI~\I 17.1 21. 1072. b'ciss. R.F., Hydrotliernial n1ang~11escin thc deep sea: scavenging reside~li? Rrislinasu~anii. S., J.K. Cochrul. K.K. Turekian and M.hl. Sarin. Tinic time and ~n/"e relationships, Eorrlr Plorrer. Sci. Lc11ers.37. 257-262, ' scales of deep-sea ferromanganew nodule growth based on ''Be and 1977. alp113 tr3ck distributions and their relation to uranium decay series Yiou. F. and C.M. Raisbeck. Half life of ''~e, Pirj~s.RPI.. .l,errcrs.29. n1c3surcn~e1lts.in Sur 13 g~~i?~~d~j~iodules de niangane'se, Colloquc 373-375. 1972. Internatio~~aldu CNRS. no. 289. Ku. T.L.. A. Omura and P.S. Chen, Be" and U-series isotopes in nodules froni the zentr3l Nortli Pacific. in Marine geology and oceanography of .tlie ce~lrral P3~ific nlanpnesc noduk province, J.L. Bischnfl' and (Received February 27, 1979; D.Z. Pipr.1. IIJj. ill prcs,. 197s. accepted April 5, 1979.)

VOL. 6, NO. 6 GEOPHYSICAL RESEARCH LETTERS JUNE 1979

Correction P

In the paper "The Measurement of ''Be in Manganese Nodules Using a Tandem Van de Graaff ~ccelerator"by K.K. Turekian, J.K. Cochran, S. Krishnaswami, W.A. Lanford, P.D. Parker and K.A. Bauer (Geophysical Research Letters, 6(5) 417-420, May 1979), the third paragraph, column 2, page 418 should read:

Figure 5 shows a plot of the flux of ''Be to a nodule verFus the nodule accumulation rate for all available measurements (Table 2). Eight of the eleven points can be'seen to lie close to a line going through 'the origin. The other three points show higher 'OB~ fluxes than would be predicted from the correlation implied by the line. Two of these are the data points of Ku, Omura and Chen (1978) for nodules from the same area as several of those falling on the line. The reason for this disparity is not obvious and may be the result of interlaboratory variations. Except for one point measured by Guichard, Reyss and Yokoyama (1978) all the 'OB~ measurements were determined by the group at the Physical Research Laboratory or standardized with analyzed samples from them. The other data point falling off the implied correlation line represents a nodule from 40°N, whereas the other data points are nodules from lower latitudes. The significance of this, if any, is not yet clear.

(Accepted June 1, 1979)

Paper number 9L0843. 0094-82761791069L-0843$01.00 * Measurements of Distributions Using a Tandem Van de Graaff Accelerator

P. D. Parker, W.A. Lanford, and K. Bauer

A. W. Wright Nuclear Structure Laboratory Yale University, New Haven, Connecticut 06520

J. K. Cochran and K. K. Turekian

Department of Geology and' Geophysics Yale University, New Haven, Connecticut 06520

S. Krishnaswami

Physical Research Laboratory Navrangpura, Ahmedabad, India

ABSTRACT

We have been using the Yale MP Tandem as a very s~nsitivespectrometer 10 9 to directly measure the ratio of Be/ Be atoms in samples of geophysical interest. The procedures used to measure absolute l0E3e contents of unknown samples are described.

*This work is supported in part under U. S. Department of Energy Contract No. EY-76-C-02-3074, and NSF grant OCE-76-02039. We also acknowledge the curator of the Lamont-Doherty Geological Observatory and grant ONR-N0014-75-C-0210. :>..

It has been 40 years since Alvarez and ~orno~')directly rnkasured the pres- 3 ence of He in natural He using the old Berkeley 60 inch cyclotron as a very seilsitive

mass spectrometer. This novel use of a, then, forefront accelerator was made

during the first few months of operation of .the accelerator. More recently, Muller, 2)

also from Berkeley, has re-emphasized the potential importance of using nuclear

,accelerators as ultra-sensitive mass spectrometers. Since Mullerls 1977 paper,

groups at Rochester, McMaster, Grenoble, Yale aid elsewhere3) have demonstrated

the ease with which accelerator based mass spectroscopic measurements of long lived

radioisotopes can be made. While much of this effort has been directed toward re-

finements in measurements of 14c: 13c: 12c ratios for archaeologic dating and

climatic research, for small samples, the advantages of counting atoms as opposed to

counting decays progressively increases the longer lived the isotope. In particular, 6 the practical and convenient usefulness of 'OB~, which has a 1.5 x 10 year half-life,

for dating over a span of millions of years or as a tracer of geophysical processes

hinges largely on the successful implementation of very sensitive mass spectrographic 10 methods of analyzing samples for Be.

There are already two groups which have reported measurements of 'OB~

contents of natural samples determined using nuclear accelerators as mass spectro-

graphs, and a number of other groups are also working in this area. Raisbeck and 10 . collaborators have used the Greqoble ~yclotroo~)to measure Be in sea water and

in ice cores. At Yale, the Tandem Van de Graaff has been used to measure 'OB~ in

marine samples .5) a Figure 1 shows a schematic diagram of the experimental setup used to measure the l0BePBe ratio. This arrangement is similar to those used at other tandems

for 14c analysis. 3, Samples to be analyzed (in the form of BeO) are loaded into the

cesium sputkr ion source of the Yale MP Tandem Van de Graaff. In the ion source,

. . the sample being analyzed is bombarded with a 20 keV Cs beam sputtering off 'various

ions and atoms. The most prolific negative ions produced are B~O-ions, which are

extracted, accelerated to 20 keV, analyzed by the ion source magnet and injected into

the accelerator. The accelerator is typically operated with -10 MV on the high volt-

age terminal. The injected B~O-ions are accelerated to the high voltage terminal

where they pass through carbon foils and, as a result, a& stripped of their electrons. 4+ This causes the molecular ion to dissociate. The resulting positive Be ions are

then further accelerated by the positive high voltage to the high energy end of the

tandem where they are momentum analyzed and pass into the detection system. See

Fig. 1.

The detection system consists of a removable Faraday cup for measurement 9 of intense ion beams. (such as Be) and a AE-E silicon surface barrier detector Me- 10 scope for measurement of individual ions in very weak beams (such as Be). When

the telescope is being used to count individual 'OB~ ions, an absorber of 17.25 2 mg/cm A1 is placed in front of the telescope. This absorber stops all stray ions of

elements with Z greater than that of Be (most importantly B and 0) without attenuatiug

the number of Be ions.

Because for a given energy the rate of energy loss of light ions is a function of

their A and 2, the AE-E telescope provides additional data with which to identify ions

and, hence, to eliminate backgrounds. To calibrate the AE-E identification spectrum,

the counter telescope was rotated to 30 degrees from the beam axis, and a 'OB 9 target was inserted in front of the telescope and bombarded with a 44 MeV Be beam.

The resulting AE-E spectrum is shown iu Figure 2 where the loci of Li, Be and B

0 isotopes are clearly evident. [The line at 45 in Figure 2 corresponds to ions which stop in the AE counter.] Inspection of Figure 2 shows, for example, lines corre- 7 9 8 sponding to Be, Be and 1°~e.No Be is seen, & expected, because it is unstable to decay into two alpha-particles. Ouce the AE-E spectrum is calibrated, it can be used to unambiguously identify.ions incident on the telescope if they have sufficient energy to pass through the AE detector (17 pm thick) and stop in the E counter.

A typical AE-E spectrum observed when the telescope was being used to count 10 Be ions is showu in Figure 3. This spectrum was measured with the telescope at

0 degrees and the A1 absorber in place. The dashed lines correspond to the loci of the various isotopes as determined from the calibration spectrum (Fig. 2). Although 10 not completely clear from this figure, almost all the ions observed are . . Be. There 9 are a few Be ions, which, because of the magnetic analysis before the telescope, come at a higher E (and a lower D).A group containing a few Li ions is also ob- 4 served. A contiuuous spectrum of He is also seen. Because these are produced by interactions in the absorber, they do not have to be mono-energetic as do those ious which have passed through the analysis system.

In such an experimental arrangement, backgrounds caused by miss-identification of ions are essentially negligible. The ions counted in the AE-E telescope as 10 10 Be are, indeed, Be. The other ions which somehow pass through the analysis system and are incident on the telescope are relatively few and are clearly distiu-

@shed from 'OB~ by their position on the AE-E spectrum.

We do, however, observe a background when we analyze a sample which is 9 expected to be pure Be. This background fluctuates in different measurements and is believed to result from cross-contamination in the ion source. We are presently attempting to reduce this background by appropriately bafniug the ion source.

The procedure we use to determine the l0BeaBe ratio for an unknown is to 10 9 repeatedly measure the Be counts in the telescope and the Be current in the removable Faraday cup. The switch between these two ion beams is made by changing only three magnets: the ion source magnet, the analyzing magnet and the quadmpole lens; see Fig. 1. It typically takes 30 seconds to change these magnets. We cycle 9 back and forth between 1°Be counts and Be current. repeatedly until sufficient statistical accuracy and reproducability is achieved.

The absolute 'OB~PB~ratio is determined by comparing the measured ratio of ''Be couu&?Be current for the unknown with the ratio measured for a known standard(s) which is in the ion source at the -same time. The samples in the ion source are mounted on a-remotely rotatable wheel which allows us to change source samples rapidly add easily without making any other chauges in the analysis system. initially, we tested this procedure by measuring the l0Be?l3e ratio for a series of 9 known samples (made by dilution of the standard with pure Be) and established that the procedure outlined above was accurate. 5.6) 9 One important point in this procedure is the measurement of the Be beam current -after acceleration and magnetic analysis. This is done, as opposed to the 9 experimentally easier procedure of measuring the Be current at the entrance to the accelerator, because both the ion beam current and emittance vary from cone to cone.

A measurement of current before the accelerator does not as fully reflect changes in emittance, as does the measurement after the analysis system. A second important point is that by making measurements relative to known

standards, instead of atternptiug to measure the absolute 1°Be?Be ratio directly,

one avoids possible isotope fractionation effects in the ion source sputtering process

or in transmission through the accelerator.and analysis system.

It is perhaps worth noting that it is quite straightforward to make measure- 10 ments of Be which are geophysically interesting, using only fairly standard equip-

ment available in most tandem laboratories. In fact, to get started we did not have

to make any changes in our laboratory equipment. The only %on-standard1' item

needed was the capability of stabilizing the high voltage of our accelerator with a

generating voltmeter. We had built in this capability several years ago, primarily to

,.#, d. voltage condition automatically, but this is the first time we had used'it to run beams . ,<.,~-. without reference to slit currents. Figure 4 shows a plot of ''Be counts -vs tennipal

voltage, showing that it is necessary to regulate to of order * 5 kV, which is easily

'?? attainable.

. J At the present time we are making a number of modifications to increase the

sensitivity of our measurements. These improvements include the installation of

0 an electrostatic analyzer following our 90 magnetic analyzer and the modification

of the 'accelerator ion source to provide better baffling between the different

sample cones and to allow the use of 'smaller~arn~leswith better beam emittance.

Our initial applications of this method to geophysically interesting problems . . and samples have been centered on the study of manganese nodules, and the detailed

results of this work will be described by Kirk Cochran in the next talk this morning. . . 9 It is clear that there is a wide variety of geophysical problems which can be explored using this new method for measuring the distributions and concentrations of ''Be and other long-lived radio isotopes. One problem of particular interest to us (and of fundamental importance to the interpretation of all su'ch measurements of cosmogenically produced long-lived isotopes) is the determination of the history of the time dependence of the cosmic-ray flux responsible for producing these isotopes in earth's atmosphere. This problem has important geophysical and astrophysical implications, and we are 'currently setting up to investigate 26 it here using the method of double-dating (analyzing both and Al) which seems to be particularly well suited to this problem. REFERENCES 1. L. W. Alvares and R. cornos, Phys. Rev. -56 (1939) 379 and 613.

2. R. A. Muller, Science 196 (1977) 4S9. , . - 3. See Proc. of the First. Conference on Radiocarbon Dating with Accelerators, . . April 20-21, 1978. Ed. H. E. Gove, University of Rochester, and references

therein. 4. G.M. Raisbeck, F. Yiou, M. Fruneau and J. M. Loiseaux, Science -202 (1978)

5. J. K. Cochran, S. Krishnaswami; W.A. Lanford, P.D. Parker and K.K. .' Turekian, Trans. Am. Geophysical'~n.',59- (1978) 1117:

. 6. K.K. Turekian, J.K. ~ochran,S. Krishnaswami, W.A. Lanford, P.D.

Parker and K. Bauer, Geophysical Research Letters -6 (1979) 417. FIGURE CAPTIONS

Fig. 1: A schematic representation of the setup used to measure isotope ratios. The

samples to be analyzed are introduced into the ion source at the left, B~O-

ions are extracted and injected into the accelerator. At the high voltage

terminal, these ions pass through carbon foils removing the'ir electrons.

The resulting positive ions are then further accelerated and mWetical1,y

atialyzed before passing into the detection system.

Fig. 2: A typical AE-E calibration spectrum used to identify the loci of particular

isotopes. b Fig. 3: A typical AE-E spectrum recorded while using the detector telescope to count 10 Be ions.

Fig. 4: Plots of 1°Be count rate -vs ion source magnet field and terminal voltage. YALE MP TANDEM ISOTOPE ANALYSIS SYSTEM

High Voltage 40MV

, 6Absorber E and AE Detectors

I I I I

TERMINAL VOLTAGE

TERMINAL VOLTS (MV) ''BE IN POLAR GLACIERS:

A TREE RING ANALOG FOR THE ENTIRE PLEISTOCENE

Mt. Soledad Radioisotope Laboratory

scripps Institution of Oceanography

University of California, San Diego

La Jolla, California The polar ice caps, which are more than 4000 m thick in some places,

hold an almost undisturbed record of terrestrial conditions during the

Pleistocene. More inert than even the abyssal sediments, the Greenland and

Antarctic ice caps harbor no significant flora or fauna to cause bioturbation,

and, because the mean annual temperature is well below freeaing (e.g. -24" C

at Camp Century), are also not subject to alteration through melt water

percolation. The major diagenesis of trace constituents occurs via vapor

phase transport of volatile components beforeconsolidation seals off the

network of cavities which initially permeate the snow pack. The deeper ice

is subject only to solid state diffusion.

The snow scavenges from the atmosphere, during nucleation of the snow

crystals and as it falls, non-volatile components including terrestrial

and cosmic dust and atmospheric aerosols. Atmospheric gases are held in

: the latticework between individual snow flakes and are eventually trapped

. in the ice as the growing weight of overlying snow increases the density of

the firn and seals the cavities from further contact with the atmosphere.

The glacial ice thus holds a record of the past composition of the

atmosphere and its trace constituents. One of the major impediments .to the

full utilization of this record is the lack of a technique for assigning ~ time scale to ice samples which are older than the 14c limit of about 25,000 years. In this paper I will discuss the possible use of 1°~eand

36~1,both as chronometer species for dating polar ice and as general tracers

of geophysical and astrophysical processes during the Pleistocene.

Theoretical time scales have been obtained for glaciers by considering

the dynamic response of the glacier to the applied stress (1). The simplest

model (2) assumes a constant vertical strain rate, neglects bottom melting and assumes that the accumulation rate and glacial thickness are temporally

as well as spatially constant. The following relation then holds between

the distance from the bottom of the ice cap, y, and the age of the ice, t,

H is the thickness of the ice cap, X is the accumulation rate, and T equals H one year. All lengths are expressed in meters of ice equivalent.

From this it readily follows that

*=-- A~ dt HT' .' or that the thickness of an annual layer is

The annual layer thickness decreases as one approaches the bottom of the .: .: . glacier. This thinning of the annual deposition layers, is an important factor

when considering the time resolution available in ice core studies. For

example, the 1367 m Camp Centruy core extends back to about 125,000 years;

the oldest 70,000 of which:'occur in the lowest 50 m (3).

There are currently several experimental techniques available for

. assigning a time scale to the polar ice sheets. These technqiues include

classical stratigraphy, stable isotope stratigraphy (6 180), micro-particle

stratigraphy, chemical stratigraphy, and radioisotope dating (1). The strati-

graphic techniques all rely on the counting of annual layers which are

delineated by seasonal differences in some chemical or physical property of

the snow deposit. These techniques are therefore limited by diffusion and thinning of layers - both processes which tend to obliterate seasonal

variations - to samples not much older than 10,000 years (1) except

possibly for 6 180 variations which could extend back several hundred

thousand years.

Radioisotope dating, on the other hand, being an integrative rather

than a differential technique, is in principle limited only by the half-life 14 of" the isotopic chronometer species being employed. C (tl,? = 5730 y) is

the longest lived isotope which has been used for dating ice cores (5) and can

be sued for samples as old as about 25,000 years. It has already been

shown that the ice caps are much older than this. Three cores have been

obtained through ice caps to bedrock, from Camp Century, Greenland, (1376 m)

(3), fro the Devon Island ice' cap (299 m) across Baff in Bay from Greenland

(5), and from Byrd Station (2110 m) in the Antarctic (4). All these

cores extend through the full Wisconsin glaciation and contain ice more than 100,000 years old. Planned drilling in Greenland, where ice

thicknesses are even greater than at the above sites, promises to produce

even older ice. 5 There are a number of radionuclides available for dating in the 10 6 to 10 year time span. These are listed in table 1 along with the estimated

production rates. The cosmic dust rates assume an influx rate of 1000 tons

per day, which is probably an upper limit to the actual influx rate. Cosmogenic nuclide rpoduction rates and cosmic dust influx rates are from La1 and

Venkatavaradan (21), except fro 1°~e,which is from Amin et a1 (22). The uranium value was calculated by assuming a trace element input of seawater

composition normalized to a measured Na concentration of 50 ppb, which is

typical for polar ice (6). As can be seen, the concentrations

are so exceedingly small that ton sized samples would be required for measurements by even the most sophisticated high sensitivity I .. I counting techniques. Although near heroic efforts have enabled Prof. H. Oeschger

Table 1 Radionuclides in Polar Ice

2 Deposition Rate (atomlcm sec) Ice Concentration* .Isotope 5/2 Cosmic Dust Atm. Prod. atom/kg dpm/ t on

5 6 234~/238~2.5~10 y . terrestrial 1.1~10 ,006

*assuming an accumulation rate of .26 g/cm2y

14 and his gioup at the University of Bern to extract C samples from ton

quantities of in situ melted ice, it is clear that such large sample sizes

severely limit the use of countirg techniques'for the application pf these

isotopes to developing a glacial chronology. However, recent developments

in high energy mass spectrometry using cyclotron or van de Graaff

accelerators haw increased by several orders of magnitude the sensitivity

with which long-lived radioisotopes can be determined. Although the other isotopes in the table also warrent investigation as glaciological - chronometers, the following discussion will concentrate on the application 36 of the ''Be/ C1 pair to ice dating and the study of Pleistocene processes. 6 Because its half-life (1.5~10y) makes it ideal as a dating tool and

a tracer for the Pleistocene, ''Be has been a tantalizing isotope to earth

scientists ever since its first detection in Pacific sediments by Arnold in

1956 (18). The major impediment to its widespread use as a tracer has been

the extremely low concentration with 'which l0I3e is found in natural

reservoirs. The global mean production rate by cosmic ray spallation in 2 . atmospheric nitrogen and oxygen is estimated to be about 1.8x10-2atoms/crn sec ' 2 or 3.5x10-~ dpm/cm y. High sensitivity counting techniques require a minimum 10 of a few times 10 atoms to obtain numbers with a 10-20% analytical precision.

Detection even at these modest levels of precision places severe demands on

both the radiochemical separation procedures and on the high sensitivity

counting techniques used. The use of cyclotron or van de Graaff accelerators

for isotope determinations by high energy mass spectrometry has increased 8 ''Be sensitivity by at least two orders of magnitude to between 10' and 10

atoms (7).

There is a growing body of experimental information with which to

estimate fallout rates of 'OB~ in polar regions. Figure 1 presents all

available ''Be data - in sediments, ice, and rain - which could be translated 2 into an atoms/cm sec deposition rate, plotted as a function of latitude. Also

plotted is an estimate of the 'latitudinal variation of the .''Be fallout rate

based on the curve of La1 and Peters (7). The dashed line is the calculated

mean global ''Be rate. The interpretation of the observed variation

in 'OB~ deposition is not yet clear (9). This variation may be due to either

latitudinal dependence of the fallout rate resulting from the mode * of

injection Of stratospheric 'OB~ into the troposphere or, at least inthe

case of sediments, to chemical and biological effects influencing the

precipitation of 1°~e. 1. B.S. Amin, D. Lal, and.B.L.K. Somayajulu. Geochim. et Cosmochim Acta 39 (1975) 1187. 2. R. Finkel, S. Krishnaswami, and D.L. Clark. .EPSL 35 (1977) 199. 3. T. 1noue and S. Tanaka. Nature 277 (1979) 209. 4. S. Tanaka, T. Inoue, and. M. Imamura. EPSL 37 (1977) 55. 5. G.M. Raisbeck, F. Yiou, M. Fruneau, J.M. Loiseaux, M. Lieuvin, and J.C. Ravel. (unpublished work) '6. R.M. McCorkell, E.L. Fireman, and C.C. Ladgway, jr. Science 158 (1967) 1690. ~ 7. G.M. Raisbeck, F. Yiou, M. Fruneau, M. Lieuvin, J.M. Loiseaux. Nature 275 (1978) 731. 8. G.M. Raisbeck, F. Yiou, M:Fruneau, J.M. Loiseaux, M. Lieuvin, J.C. Ravel, J.D. Hays. (unpublished work). 36~1has been much less studied than l0Iie; however, its half-life 5 = 3.1~10y) suggests that it would also be extremely useful for (tl/2 glaciological studies. There are two natural production mechanisms for 36~1 in the terrestrial environment: cosmic ray spallation of atmospheric argon and thermal neutron capture by 3561 in surficial salt deposits and saline waters. The thermal neutrons involved are derived mainly from secondary neutrons produced by cosmic ray interactions in the atmosphere. 2 Neutron capture in seawater produces approximately 0.12 atomslcm sec.

When mixed through the entire ocean, this leads to a 36~1/~1ratio of about

5x10-16, which is just below current detecti~onlimits. The neutron capture mechanism is not significant for glaciological studies, because any small amount of such activated chlorine, once transported to the ice caps, would be masked by the much higher specific activity spallation 36~1deposited directly from the atmosphere. La1 and Peters (8) have estimated the mean 36 global production rate of C1 by spallation of atmospheric argon to be 2 1.1x10-~ atomslcm sec, about an order of magnitude less than that of 1°~e. 10 As with Be, it is only since the development of high energy mass spectrometric techniques that it has become possible to contemplate extensive application of

36~1to glaciological problems. Current van de Graaff techniques have a 5 detection limit of about 10 atoms of 36~1(10).

There is much less experimental data for 36~1than for 1°~e.Schaeffer 36 et a1 (11) reported C1 concentrations in rainwater and Bonner et a1 (12) attempted measurements in lakes and rivers in the western U.S. Both groups concluded that the 36~1they had detected was primarily bomb produced. Bagge and Wilkommen (13) have reported measurements of 36~1produced by neutron capture in a saline lake. Tamers et a1 (14) have studied various groundwater and saline deposits and reported measurable 36~l.concentrations. It is difficult 36 to derive an estimate of the atmospheric C1 production rate from these

reports,'because of the large neutron capture component to the production

mechanism in most of the samples studied. ~ishiizumiet a1 (15) have ,

reported one measurement of 36~1in ice at 70' S from the Antarctic Yamato

Mountain area. There are uncertainties about the age of this ice and about 36 its accumulation rate which preclude the calculation'of a C1 deposition

rate. However, the measured concentration is consistent with the estimate

of La1 and Peters (8) after correcting for an assumed latitudinal fallout

dependence similar to that observed for "~r. Although uncertainties in

interpretation remain, work to date has demonstrated that by using high 10 energy mass spectrometric techniques, 36~1and Be measurements can be :

made in kg quantities of melted polar ice.

An important consideration in the application of either l0J3e or 36~1

to the dating of polar ice is the relationship between changing snow

accumulation rates and isotope deposition rates. It is possible to consider

; two extreme models. In the first model one can assume that the concentration

(atoms/g snow) of isotope remains constant. Then one can write

where A. and Ah are the isotopic concentrations in the ice.at the time of

deposition and at depth h in the glacier resp ively. Xis the radioactive

decay constant and t is the age of ice at depth h. In this model the h de*h-age relationship remains valid even if water or isotope accumulation

rates vary.

In the other extreme case, 'one can assume that the isotopic deposition

rate, atoms/cm2 sec, remains constant. In this case one can write 2 where S is the integral isotope concentration in the core in atoms/cm , 2 F is the isotopic deposition rate in atoms/cm y, and Xis the radioaactive

decay constant. Integrating over the whole core one obtains

or F = ASoD 2 where Sm is the integ.ra1 number of atoms in the core per cm .

There is,at the moment,insufficient data to decide between these two

models, although 'OS~ (16) and trace element measurement. (6) in Antarctica

tend to support the idea that, at least for the Antarctic ice cap, deposition

rates may be proportional to accumulation rates. It is, of course, possible

that neither model is strictly correct and that isotope deposition rates may

be a more complicated function of other climatological and geophysical .- . parameters. For dating purposes such variations can be corrected for by

measuring the ratio of two isotopes with the same production and fallout 10 patterns. For glaciological studies, 36~1and Be are an ideal pair. Both

are produced by reactions which vary in a similar way with bombardment

energy throughout the atmosphere. Both probably behave in a similar fashion

in the lower atmosphere so that fallout patterns could be expected to be 10 5 similar. The 36~1/Be ratio decays with a half-life of 3.9~10 years which

is ideal for the dating of ice caps estimated to contain ice as old as one

million years. 36 The. importance of ''Be and Cl measurements does not lie soiely or even primarily in the development of dating techniques. The deposition record of these isotopes which is held in the polar ice caps must also reflect any changes in atmospheric mixing modes such as stratosphere-troposphere exchange and any changes in atmospheric"isotopic rpoduction rates. The second factor is of special interest in relation to 14C measurements in dendrochronologically dated tree rings. These measurements have demonstrated a' complex history of variation of the atmospheric 14c content, with both short term (100 years or less) and long term ( thousands of years) periodicities.

Proposed explanations for these fluctuations have been discussed in detail by Damon et a1 (17). Broadly, the possible causes can be grouped into fluctuations due to geochemical which perturb the atmospheric content and/or CO exchange rates between the various terrestrial reservoirs 2 and fluctuations which affect the atmospheric rate of 14c . 36 ''Be ''Be and C1 are expected to have short atmospheric residence times, probably equal to the aerosol residence time of about two weeks. For this reason and the lack of any input of these isotopes into the atmosphere from other reservoirs, ''Be and 36~1will not be influenced by the same kinds of reservoir exchange processes which could influence 14c levels. Both ''Be and 36~1contents are expected, however, to be affected by changes in atmospheric production rates, whether caused by changes in the cosmic ray flux, in the extent of solar modulation, in the strength of the terrestrial magnetic field or by more exotic possibilities such as the radiation associated'with matter-antimatter annihilation which has been proposed to explain the Tunguska meteorite event (17, 19). A comparison of 14 ''Be ''Be and 36~1results with C fluctuations 'can help to separate geochemical from cosmic ray causes for the 14c variations. Because severe thinning of

layers only occurs towards the base of the ice caps (I), the time resolution

available for such studies in ice cores using high energy mass spectrometry 14 will be sufficient. However, the C fluctuations observed are of the order

of 10% or less. A 10% variation in l0B;or 36~1might just be detectable in

ice samples with theqrecision now available. There seems, however, no -a priori reason to doubt that future improvements will lead to analytical precision competitive with that now available for 14c. Once mapped, ''Be

and 36~1fluctuations will provide an additional technique for dating ice

which is too young to be dated by techniques relying on conventional

radioactive decay.

''Be and 36~1also offer the possibility of looking for production

rate fluctuations associated with magnetic reversal events such as the

Lake Mungo event at 25,000 years and the Laschamp event at about 35,000

years (20). In addition, depending on what is ultimately learned about the

-.. age of the ice caps, it might also be possible to investigate events as old

as the Brunhes-Matuyama boundary at 690,000 years or perhaps even older.

There are other polar phenomena such as the relation between

Antarctic meteorites and the ice in which they are imbedded and the elucidation

of glacial flow patterns which will certainly profit from the application 36 of ''Be and C1 measurements. 12. References

1. C.U. Hammer, H.B. Clausen, W. Dansgaard, N. Gunderstrup, S.J. Johnsen

and N. Reeh, Dating of Greenland ice cores by flow models, isotopes,

volcanic debris, and continental dust. Journal of Glaciology 20 (1978) 3 - 26. .2.-R. Haefeli, Contribution to the movement and form of ice sheets in the Arctic and Antarctic. Journal of '~laciolo~~3 (1961) 1133 - 1151.

3. K. Dansgaard, S.J. Johnsen, H.B. Clausen, and C.C. Langway, jr., Climatic record revealed by the Camp Centruy ice core. In --The Late Cenozoic Glacial Ages edited by K.K. Turekian (1971) 37 - 56. Yale University Press, New Haven.

4. S.J. Johnsen, W. Dansgaard, H.B. Clausen, and C.C. Langway, jr., Oxygen

4 isotope profiles through the Antarctic and Greenland ice sheets. Nature 235 (1972) 429 - 434. .~' 5. W.S.B. Paterson, R.M. Koerner, D. Fisher, S.J. Johnsen. H.B. Clausen, I~ W. Dansgaard, P. Bucher, and H. Oeschger. An oxygen-isotope climatic

record from the Devon Island ice cap, arctic Canada. Nature.266 (1977)

6. C. Boutron, M. Echevin, and C. Lorius, Chemistry of polar snows.Estimation

of rates of deposition in.Antarctica. Geochimica et Cosrnochimica Acta

7. G'.M. Raisbeck, F. Yiou, M. Fruneau, M. Lieuvin, J.M. Loiseaux, Measurement

of 'OB~ in 1,000- and 5,000-year,old Antarctic ice. Nature 275 (1978) 731 - 733. 8. D. La1 and B. Peters, Cosmic ray produced radioactivity on the earth, in

Handbuch der Physik, vol 4612, edited by K. Sitte (1967) 551,-612.

Springer, New York. 10 9. R. Finkel, S. Krishnaswami, and D.L. Clark, Be in Arctic Ocean sediments. Earth and Planetary Science Letters 35 (1977) 199 - 204. 10. D. Elmore, B.R. Fulton, M.R. Clover, J.R. Marsden, H.E. Gove, H. Naylor,

K.H. Purser, L.R. Kilius, R.P. Beukens, and A.E. Litherland, Analysis of 36 C1 in environmental water samples uisng an electrostatic accelerator. Nature 277 (1979) 22 - 25. 11. O.A. Schaeffer, S.O. Thompson, and N.L. Lark, Chlorine-36 radioactivity in rain. Journal of Geophysical Research 65 (1960) 4013 - 4016. 12. F.T. Bonner, E. Roth, O.A. Schaeffer, and S.O. Thompson, Chlorine-36

and deuterium study of Great, as in lake waters. Geochimica et Cosmochimica Acta 25 (1961) 261 - 266. 13. E. Bagge and H. Wilkormn, Geologische Alterbestimmung mit 36~1. Atomkernenergie 11 (1966) 176 - 184. 14. M.A. Tamers, C. Ronzani, and H.W. Scharpenseel, Observation of naturally occurring Chlorine-36. Atompraxis 15 (1969) 433 - 437. 15. K. Nishiizumi, J.R. Arnold, D. Elmore, R.D. Ferraro, H.E. Gove,

R.C. Finkel, R.P.' Beukens, K.H.Chang, and L.R. Kilius, Measurements of 36 C1 in Antarctic meteorites and Antarctic ice using a van de Graaff

accelerator. Earth and Planetary Science Letters (1979) in press.

16. G. Larnbert, B. Ardouin, E.Brichet, C. Lorius, Balance of 'OS~ over

Antarctica: Existence of a protected area. Earth and Planetary science Letters 11 (1971) 317 - 323. 17. P. Damon,.J. Lerman,'and A. Long, Temporal fluctuations of atinospheric

14c: causal factors and implications. In Annual ~evievsof Earth and

. .Planetary Sciences, edited by F. Donath, vol 6 (1978) 457 - 494. 18. J.R.,Amold, Beryllium-lo produced by cosmic rays. Science.124 (1965) 584 - 585. 19. K. ~'~rien,secular variations in the production of cwgenic isotopes

in the earth's atmosphere. ~oumalof Geophysical Research 84 (1979) 423 - 431. 20. M. Babetti and K. Flude, Geomagnetic variations during the late

Pleistocene period and changes in the radiocarbon time, scale. Nature 279 (1979) 202 - 205. . ,521. D. La1 and V.S. Venkatavaradan, Activation of cosmic dust by cosmic-ray particles. Earth and Planetary Science Letters 3 (1967) 299 - 310. 22..B.S. Amin, D. Lal; and B.L.K. Somayajulu, Chronology of marine sediments

4 using the 'OB~ method: intercomparison with other methods. Ceochimica

et Cosmochimica Acta 39 (1975) 1187.