tool provide in here the a the series of 14( concentrations in the atmosphere the ocean between surface 28 N between and 45 S and and

Transcription

1 [RADIocARBON, VOL. 38, No. 3,19%, P ] FURTHER APPLICATION OF BOMB 14C AS A TRACER IN THE AND AT OCEAN MOSPHERE REIDAR NYDAL and JOR UNN S. GISLEFOSS Radiological Dating Laboratory, The Norwegian Institute of Technolo N-734 Norway gy, N-734 Tr ondhelm NTH ABSTRACT. Bomb 14C from nuclear tests in atmosphere has proved to be a carbon particularly cycle. useful We tool provide in here a stud ca. 3-yr of time series of 14( concentrations in atmosphere ocean between surface 28 N between and 45 S 71 N and and 45 N. in More recently (since 199), a north-south profile surface also waters has been of obtained Atlantic for Ocean. 14C in The measurements were performed using ing of conventional large samples (4 technique to 5 liter of beta C2) in count- CO2 proportional counters. These data show that sphere is leveling 14C concentration off with in atmoa time constant - of.55 yr-1, and is now approaching that of ocean surface at lower latitudes. Additional tracer studies have been concerned especially with penetration of gian bomb and 14C into Greenland deep seas ocean. are of interest The Norweas a sink for atmospheric CO2 and also a source During of water for last five deep Atlantic years, several 14C Ocean. depth profiles have been measured from Fram (62 N), Strait using (79 N) to AMS south of technique Iceland available at University of Arizona AMS F Facility. and compare a few of it profiles with those produced b by GEOSECS expedition in 1972 The and profiles show TTO that expedition water in descending to deep Atlantic Ocean is originating depths mainly in from intermediate Nordic Seas. and However, surface ventilation rate of Norwegian Sea deepwater ponent is in too slow transfer to be an of important water over com Greenland-Scotland Ridge. INTRODUCTION About 1 nuclear tests, with a total strength of 5 MT (TNT equivalent), ' atmosphere were carne d out from in 1945 to About two-thirds of that energy was released at itudes; higher mainly norrn over lat- Novaya Zemlya and mainly during fall of 1961 and 1962 (UN The main R Report atmospheric 1964). testing pr ograms came to an end with Test Ban Treaty of 5 (The Moscow August 1963 Treaty). France and China did not accept Treat Y immediate) and smaller Y continued bombs to in test atmosphere. France h had its main testing period from 1966 to 1968 se and out in carried Pacific Ocean. China carried out several tests over Asian until Continent 198, (Lop with highest activity during period 1968 to The total se contribution post-1963 from bomb tests was ca. 12% of total power released into nuclear atmosphere testing. from We present here a furr contribution to carbon cycling research program, based on b 14 a tracer, omb which C as was initiated in our laboratory i in From a noted doubling of at norrn natural latitudes 14C level in 1962, dispersion of bomb 14C in atmosphere and been ocean studied surface extensively has during g subsequent year s (Nydal and Lovseth 1983; present, Nydal et al. 14C 1984). concentration At in atmosphere is approaching that of ocean itudes. surface at lower The tracing latinterest is now main) y concerned with continuing into penetration deep of ocean. bomb The arctic regions 14C are especially in focus as y constitute sinks for CO2. Several atmospheric cruises during recent years have been used to measure 14C depth gian profiles and in Greenland Norwe- Seas. These data are used to study uptake of CO 2 in net Nordic S transfer Seas and to its deepwater in Atlantic Ocean (Nydal et a1.1991,1992; Gislefoss et a1.1995). SAMPLING AND MEASUREMENTS The CO2 sampling and conventional 14C measurement of atmospheric CO and are 2 ocean described surface water in detail in earlier articles (Y N dal and vseth 1983; Nydal et a1.1984). In brief sum- 389

2 39 R. Nydal and J. S. Gisle f oss wary. 1) sampling at ground level in troposphere is performed with a solution of 1-2 liters NaOH (2%), which is exposed to open air for several days. An absorption time of 7 days was generally applied from , but this was changed subsequently to 3-4 days. Depending on air ventilation, between 3 and 6 liters of CO2 are now absorbed over a period of 3 to 4 days; 2) sampling from ocean surface is generally based on recovery of 2 liters of seawater, col - lected at a depth of 5-1 m through inlet of pumping system of ship. After acidifying water (with H3P4) to a ph value of ca. 3, a relative amount of between 4 and 6 liters of CO2 is extracted in a flushing Procedure on board ship and absorbed in a bottle of.75 liters NaOHi solution (2%); 3) conventional 4C measurements were performed by beta counting 3-5 liters of NIST CO2 in proportional counters. The 14C/13C ratio is measured relative to modern standard ( HOxII) and normalized for isotopic fractionation effects. The final 14C enrichment for each sample is calculated and quoted in per mil excess above pre-bomb level, according to formula defined by Stuiver and Polach (1977) A14C = b14c - 2(S13C + 25) b14. 1 (1) Until 198, a counting time of 1-2 d per sample was used to ensure an analytical precision of ca. 1%' g (1 Q). After 198, counting time was increased to 4-5 days to obtain a better precision (4-5%o). In earlier research from this laboratory, A14C values in time series of bomb 14C measurements were not corrected for radiometric decay of modern reference standard, which is defined to equate AD 195 with atmosphere. Compared to earlier limit of error (up to 1%o) in those measurements, disregard for decay of reference standard could be considered unimportant. However, after a decay period of 4 yr, associated error of 5%o is now comparable to precision achieved in radiometric measurements. A retrospective correction for this decay has refore been adopted for all our 14C measurements, using approximate correction formula A14Ccorn = A14Curicorr + 1[e(1-t) _ 1] (2) where?, is 1/8267 (573 Yr half-life) and t is year of sampling. The corrected 14C data are also available from CDIAC database held at Oak Ridge National Laboratory, Tennessee, USA. The 14C deep-sea profiles were initially achieved using large samples (1-2 liters seawater) and radiometric 14C counting technique. This collection procedure was very time-consuming. However, recent availability of accelerator mass spectrometry (AMS) for 14C measurement allows sampling to be based on much smaller water samples (.5 liter). This technique brings exciting new $ possibilities for tracing 14C in deep sea (Gislefoss 1994; Gislefoss et al. 1994). The AMS measurements reported here were carried out at NSF-Arizona AMS Laboratory in Tucson, Arizona. The procedure involves CO2 samples between 1 and 2 ml being converted to CO over hot Zn, and furr reduction of CO to graphite over an iron catalyst at 625 C (Slota et al. 1987). The graph ite powder is n pressed into an aluminum target holder for AMS analysis. The 14CP3C isotope ratio of sample graphite target is measured and compared to that recorded by reference standards. An analytical precision of 4-5%o is indicated by replicate analyses using independently pre- Pared targets. Details of experimental procedures are given by Linick et al. (1986) and associated calculations are quantified by Donahue, Linick and Jull (199).

3 Bomb 14C as a Tracer in Atmosphere and Ocean 391 ATMOSPHERE In early 196s, our program of 14C measurements in troposphere was set up collected based on at CO2 a total of 14 stations sited between Madagascar (21 S, 47 E) and Svalbard (78 N, (Nyda11968; 13 E) Nydal and Lovseth 1983). Over years, it has not been possible to continue large number wit of stations, and from 1978 onward, sampling network was reduced to at two Fruholmen, stations Nordkapp (71 6'N, 23 59'E; 7 m above sea level ( asl )) and Izana (28 22'N, Tenerife 16 3'E ', 24 m asl). The curve for Fruholmen is complete between 1963 and curve measured at Canary Islands is largely complete over period , but measurements comprises from two neighboring stations; one at Izana, Tenerife and or at Mas Gran Palomas Canaria (27 45'N, 15 4'W; ' 1-1 m asl). U d -2 _ Year Fig in troposphere and ocean surface Measurements only from Nordkapp (71 N) and Ten- erife (28 N) after A summary curve of d14( at all our tropospheric collection stations is given in Figure 1. large seasonal The variations in troposphere values between 1963 and 1968 are caused mainly meteorological by influx of 14( from a concentration in stratosphere 1 to 2 times higher that during early period (Feely, Katzman and Tucek The main exchange of CO2 between sphere stratoand troposphere occurs during spring and summer, when tro po increases pause height toward higher latitudes. The magnitude of seasonal variation in tropospheric 14C centration conleveled off during first years, until a furr slight increase occurred from as to a result of French and Chinese tests. For subsequent years, curve follows a more exponential regular decrease, with a decay constant of.55 1 calculated at Nordkapp from Data to in Figures 1 and 2 show that past AD 198, re is a fairly close agreement between curves recorded at Fruholmen (71 N) and Izana (28 N). Both collection stations enjoy clean relatively air and are within rapid circulation cells of troposphere (Meijer et al. 1994).

4 392 R. Nydal andf. S. Gislefoss Radi togical bating L'ab.. Tron helm, ` Fruh lmen, ordkapp (71 N Omasl) Om o Year Rad gical acing Lab. Trondheim, 994, Izan, Teneri e (28 Nt ; 24OOm.s1) 15 1 ' - ' Year Fig in troposphere at Nordkapp (71 N) and Tenerife (28 N)

5 Bomb 14C as a Tracer in Atmosphere and Ocean 393 It is interesting to speculate on a possible signal in 14C from Chernobyl event of 26 April Two values of 3-4%o above ambient level at Fruholmen coincide with time of this event (Fig. 1). Due to wear patterns prevalent at time of explosion, a low-pressure feature passing over Chernobyl subjected norrn Scandinavian countries to radioactive fallout a few days later. Even though 14C from Chernobyl event can be regarded as negligible on a global scale, it is not surprising to find a more local and transient increase in this area. Direct comparison of temporal 14C concentrations recorded from two operational stations between 198 and 1992 (Fig. 2) reveal seasonal variations with summer maxima and winter minima. The higher amplitude is found at Fruholmen, where temporal variations are also more regular. A major cause of this pattern is dilution of 14C2 by excess of inactive CO2 (14C-depleted) discharged to atmosphere as a result of greater combustion of fossil fuel during winter. This dilution effect levels off during summer season, due to atmospheric mixing and ongoing exchange between atmosphere and or carbon reservoirs (ocean and biosphere). A comparison with 14C concentrations recorded at sampling stations in central Europe (Levin et a1.1995) also indicates that seasonal variation is greater in areas with higher combustion of fossil fuel. A model simulation of later (198 to 1992) seasonal trend in tropospheric 14C concentrations recorded at Fruholmen and Izana has been attempted using function F(t) = Asin 2n(t-t)+B e_(to). (3) The parameters A, B and to (Table 1) are determined via a least-square fit, whereas decay constant k is calculated independently from a longer period at Fruholmen (1973 to 1992). To comply with present observations in ocean surface, and to avoid spurious effect of additional parameters during a relatively short period, ultimate level for function was chosen to be zero. With an amplitude of 6.6 ± 1.4%o and a decay constant of.55 yr-1, function gives better fit to Fruholmen, Nordkapp data. For curve at Izana, Tenerife, amplitude term (A) was reduced to approximately one-third of that used for Nordkapp. The seasonal variation and some of 14C data at Izana are more irregular than at Nordkapp, and do not always fit well with calculated cycle. This latter observation reflects special meteorological condition at Izana, as discussed previously (Nydal 1968). TABLE 1. Parameters Obtained in a Least-Square Procedure (Marquardt-Levenberg Algorithm) for Function (2) to fit Data Sets from 198 to 1992 from Nordkapp and Tenerife Parameter Parameter explanation Nordkapp Tenerife A Amplitude of yearly oscillations 6.3 ± 1.3%o 1.9 ± 1.%o t Time at turning point of cycle ±.3 yr ±.9 yr B The A14C value at time to ± 1.2%o ± 1.6%o Mean deviation from curve 6.9% 7.3% A small contribution to seasonal variation of 14C in troposphere could be from a still enhanced concentration of 14CO2 in stratosphere. According to Tans (1981) amount of bomb 14C input to stratosphere might have been underestimated significantly from sampling flights that took place after cessation of nuclear testing. A few bombs tested in upper atmosphere were significantly larger than average and, in such instances, induced radioactivity may have reached greater altitudes than expected. For example, a single hydrogen bomb over Novaya Zemlya on 3 October 1961 had a recorded 58 MT yield (SIPRI Yearbook 1975). Furr-

6 394 R. Nydal and J. S. Gisle foss more, a recently observed AC value of %o at a height of km (39.16 N, E) by Nakamura et al. (1992) indicates a 14C excess in lower stratosphere that is still ca. 15%o above present tropospheric level. A progressively smaller part of this excess radioactivity will be transferred to troposphere during spring and summer each year. However, according to calculated residence time for 14C in upper stratosphere of 9.1 ±.2 yr, present contribution from this source of bomb 14C to amplitude in seasonal variation at ground level is <1%o. OCEAN SURFACE Prior to 1986, sampling of ocean surface water was carried out from several ships crossing Atlantic, Pacific and Indian Oceans (Nydal et a1.1984). During last ten years, however, our sampling program has been restricted to a single ship of Barber Line (MS Tourcoing), on its main global route from Europe, across Atlantic Ocean to Panama, across Pacific Ocean to New Zealand, northward to Japan and back across North Pacific and Atlantic Oceans. In some cruises, route to New Zealand has passed south through Atlantic around Cape of Good Hope to Indian Ocean, with a return route through Suez Canal and Mediterranean Sea. All of 14C measurements from ocean surface water that are shown in Figure 1 derive mainly from samples collected in region 45 N to 45 S. The mean trend from all scattered data approximates to a near-horizontal line at e14c = 1%o, which is close to present atmospheric level. A closer study of 14C in surface ocean layer shows a pattern of seasonal variations that are normally correlated with ocean temperature. The largest variations coincide with upwelling areas along continental margins, and smallest variations are recorded in those stable parts of open ocean least influenced by vertical mixing. Figure 3 shows latitudinal variation of e14c in surface of Atlantic Ocean from pre-bomb level until present. The pre-bomb level is established from data compiled by Broecker and Peng (1982), and data are reproduced from GEOSECS expedition (Broecker et al. 1985). Our data were obtained with Norwegian research vessel RV Andenes on a cruise to Antarctic in winter of (Table 2) and are supplemented by more recent measurements ( ) from Nordic Seas (Table 3). These 14C data concur reasonably well with GEOSECS data reported for both sides of Atlantic Ridge, and indicate that only small changes in A14C have occurred in ocean surface during last 2 yr. It must be emphasized that GEOSECS data show differences in magnitude between each side of Ridge (Fig. 3, I and II) and refore our more recent data have to be compared with geographically closest GEOSECS values. One of main trends in e14c curves for Atlantic Ocean is an approximate symmetry around equator. The most stable surface layers (which exhibit highest e14c values) coincide with high-pressure zones along both sides of equator at about 3 N and 3 S. A slight decrease in e14c occurs along equator, where upwelling water with a lower 14C concentration displaces surface water toward higher latitudes (Broecker and Peng 1982). There is, however, a more dramatic lowering of 14C enrichments towards higher latitudes, where more stable surface layer vanishes. The Arctic Ocean (including Nordic Seas) behaves somewhat differently from Antarctic, mainly due to geographic distribution of adjacent land areas. The Antarctic Ocean is unique in that total global circulation of ocean currents is not impeded (Pickard and Emery 199). Upwelling and downwelling of water both occur in this region (Foldvik and Gammelsrod 1988). The upwelling water displaces surface layers and dilutes its ambient e14c value.

7 zoo 15 goo U -9 v I L -6-3 (S) LATITUDE (N) Pre-1958, prepared by Broecker and Peng (1982) * , GEOSECS data (Ostlund et al. 1976) To Antarctic, Dec (this work) o From Antarctic, Feb. 199 (this work) Summer cruises, (this work) Fig. 3. A north-south profile of 14C in surface of Atlantic Ocean

8 396 R. Nydal and J. S. Gisle foss TABLE 2.14C Measurements in Surface of Atlantic Ocean from a Return Cruise to Antarctic with RV Andenes, December 1989 to February 199 Date Sample (yy.mm.dd) Location 814C 813C A 'N 5.7 A 'N 5.3 A 'N 4.1 A 'N 4.9 A 'N 4.4 A 'N 4.1 A 'S 4.3 A 'S 3.8 A 'S 3.9 A 'S 4.1 A 'S 4.2 A 'S 4 A 'S 4 A 'S 3.8 A 'S 4.3 A 'S 4.4 A-16(2) 'S 4.1 A-15(2) 'S A-14(2) 'S 3.8 A-13(2) 'S 3.7 A-12(2) 'S 4.2 A-11(2) 'S ±4 A-1(2) 'S 3.7 A-9(2) 'S 4.9 A-8(2) 'S 5.1 A-7(2) 'S 3.7 A-6(2) 'S 4.9 A-5(2) 'N 26 8'W A-4(2) 'N 23 5'W A-3(2) 'N 2 56'W A-2(2) 'N 14 4'W A-1(2) 'N 8 51'W The sourn limit of Nordic Seas is determined by shallow Greenland-Scotland Ridge, which serves to impede exchange of water with deep Atlantic Ocean. Toward Atlantic Ocean re is very little upwelling, and sinking water generally is replaced from Norwegian Atlantic and East Greenland surface currents. These features of circulation pattern, toger with a delay in downwelling caused by shallow Greenland-Scotland Ridge, explains fact that present 14C concentration in surface water of Nordic Seas exhibit a higher 14C concentration (14C = +5%o) than occurs at corresponding latitudes in Antarctic Ocean (A14C = -1%o). DEEP-SEA PROFILES IN THE NORDIC SEAS During last five years, our 14C measurements of deep-sea profiles have been limited largely to Nordic Seas, where exchange processes are rapid enough to be studied within a limited period of time. The Greenland Sea is assumed to be one of main source regions for deepwater

9 Bomb 14C as a Tracer in Atmosphere and Ocean 397 formed at higher latitudes (Smethie et a1.1986). The surface and intermediate waters sink as a result of surface cooling and deep convection during winter. A mixture of deepwater from Greenland Sea (GSDW) and Eurasian basins (EBDW) is furr assumed to be brought down through gaps in ridge, to form Norwegian Sea deepwater (NSDW) (Swift and Koltermann 1988; Bourke et a1.1993). An excess of water is also passing over Greenland-Scotland Ridge to contribute to deepwater in Atlantic Ocean (AODW) (Swift et a1.198). The locations of our 14C profiles were chosen to give an optimal view of transfer of carbon in this area (Fig. 4). The profiles are located in East Greenland Current (A,C,E) at central positions in main basins (D,F,G,I) in Norwegian Atlantic Current and West Spitsbergen Current (B,H,J,K) and Atlantic Ocean south of Iceland (L,M,N). The 14C deep-sea profiles monitored 2 yr earlier during GEOSECS expedition (Ostlund, Dorsey and Brecher 1976), and some TTO profiles taken in 1981 (Ostlund and Rooth 1981) have provided an important comparison and allowed a study based on changes that have occurred over that period. TABLE 3.14C Measurements in Surface of Nordic Seas Arizona/ Trondheim Trondheim ref. (T) (m) LA1-1 T T 'N 29 5'E LA3-1 T 'N 3 2'E ± 4.2 LA4-1 T 'N 41 54'E ± 3.8 LA5-2A T 'N 29 12'E LA7-1 T 'N 5 52'E ± 4.6 LA8-2A T 'N 4 6'W ± 6.1 LA1-Ol T 'N 2 29'W ± 3.6 GS14-1 AA 'N 5 'W ± 4. GS14-2 AA 'N 5 'W ± 4. GS 16-2 T 'N 1'E ± 3.8 GS17-2 T 'N 14 5'W GS18-1B T 'N 7 29'W ± 4.5 G519-1A T 'N 9 36'E LA15-2 AA 'N 11 31'W ± 4.7 M16-1 AA 'N 15 31'W ± 4.3 JH5-212 AA 'N 4 6'W ± 4.1 JH9-212 AA 'N 16 2'W ± 4. 2 JH1-212 AA 'N 32 3'W ± 4.7 The cyclonic Greenland gyre is supported by West Spitsbergen Current in east and East Greenland Current in west. It is constrained between Fram Strait in north and Norwegian Sea in south. Four deep-sea profiles of L14C were obtained in this area (A,B,C,D, in Fig. 5a). Profiles A and B show typical differences in water masses and exchange on each side of Fram Strait.1 The East Greenland Current profile (LA8, LA17) shows a linear gradient from surface to a depth of 1 m. Between 1- and 2-m depth, curve is more irregular and certainly due to influence of or water masses. This deeper part of profile also has a slightly higher salinity than in middle of Greenland Sea (Nydal et a1.1991) and probably reflects influence of saltier EBDW. The apparent i 4C inversion between ca. 12- and 18-m depth also indicates influence of water from relatively young GSDW. 1Note that alphabetic indices in Figure 4 correspond to profiles shown in Figure 5a, b, c and d.

10 398 R. Nydal and J. S. Gislefoss w 2 1 E Fig. 4. Map of 14C depth profiles in Nordic Seas The various letters A, B,...N indicate approximate locations of 14C profiles seen in Fig. 5a,b,c,d. The West Spitsbergen Current is a branch of salty and warm Norwegian Atlantic Current. This is reflected in e14c values of upper ca. 5 m of profile B (LA7), which records a small vertical gradient. The steepest gradient for vertical exchange occurs between 5 and 1 m, where e14c values change by ca. 1%o. There is a furr small decrease in e14c down to 2 m depth. The few neighboring TTO data from 1981(Sta.154 and 156) were collected closer to Svalbard, but y seem to support shape of our profile. Profile C (LA15), taken in East Greenland Current, has a similar pattern to profile A. However, in this location, linear trend extends to a depth below 2 m. The curve is supplemented by three measurements from greater depth, taken just outside shelf (LA14), which seem to fit well with deepwater data for norrn profile (A). This feature indicates that younger GSDW is affecting profile at greater depth, i.e., ca. 2-3 m. The decrease in 14C concentrations below 3 m suggests influence of a deep current that may be connected to NSDW.

11 Bomb 14C as a Tracer in Atmosphere and Ocean 399 In Profile D, data collected in center of Greenland gyre (LA1O) are compared to earlier GEOSECS (Sta. 17) and TTO (Sta. 148) profiles collected 18 and 9 yr earlier at approximately same location. LA1 shows a A14C range between +5%o in ocean surface to a mean value of -39 ± 2%o (4 samples) below 2 m depth. The lack of a AC depth gradient between 2 and ca. 35 m may indicate a well-mixed deep reservoir with a rapid internal circulation. We calculate that A14C values in deepest part of profile have increased by 12 ± 2%o relative to GEOSECS profile recorded in An extrapolation back to pre-bomb level (ca. 196) in deepwater is difficult to perform, because of little data and later change in deepwater formation (Schlosser et al. 1991). A linear increase in sequestration rate of tracer indicates, however, a pre-bomb A14C level of ca. -6%o for GDSW, a value close to that obtained for surface water. The A14C value of -59 ± 3%o in surface water is based on measurements of marine shells from Norrn Norway and Spitsbergen (Mangerud and Gulliksen 1975). All of TTO data from 1981 have A14C values intermediate between GEOSECS data and our 199 values. The area immediately south of Greenland Sea, designated as Norwegian and Icelandic Seas, is covered by four profiles (E,F,G,H; Fig. 5b), taken at virtually same latitude, i.e., 69 to 71 N. The upper parts of se profiles show a gradual change from East Greenland Current (E) to Lofoten Basin (H). Profile E, which includes two neighboring stations (GS17 and M14), records a linear decrease in 14C concentration from surface to ca. 15-m depth. The or profiles show a gradual eastward influence from Norwegian Atlantic Current in tendency to more uniform A14C values in ir upper depth ranges. In Profile F (GS18), our measurements are compared with TTO profile (Sta. 159) taken in 1981 at a slightly different location. If we assume that two sampling locations represent same water mass, n comparison shows that A14C value below 5 m has increased by 15 2%, between 1981 and 199. In Profile G (GS16), our measurements are compared with GEOSECS (Sta. 18) and TTO (Sta. 144) profiles. The earlier profiles show no significant input of bomb 14C below 2 m depth between 1972 and An increase of 7-8%o was observed, however, in 199. This is taken to indicate that deep convection only reached this deepwater between 1981 and 199. Compared to Profile D taken in central Greenland Sea, deepwater at location G is older and more in agreement with water found at 2 m depth in periphery of Greenland Gyre (C). Farr south and into more central part of Norwegian Sea (Fig. 5c), we find that deepwater becomes progressively older still. A comparison of time-transient data in Profile I (GS14, M1, GEOSECS (Sta. 19) and TTO (Sta. 144)) shows no significant differences in A14C for water collected below 2 m. The mean A14C value of -73 ± 3%o (6 samples) from measurements should not be much different from pre-bomb level. The pre-bomb A14C value in NSDW is at least 1%o lower than that of surface water of Greenland Sea. This corresponds to a decay of 14C during a period of 1 yr from surface to deep Norwegian Sea. If NSDW was mainly fed from Greenland Sea, this period should be identical with mean age of NSDW. This water is, however, also in exchange with EBDW, with or A14C values that may modify calculated age of NSDW (Bonisch and Schlosser 1995). The two profiles (J,K, Fig. Sc) measured in sourn Norwegian Sea indicate that vertical mixing is faster at periphery of basin than in central part. This is demonstrated clearly in Profile K, where our data (JH5) can be compared directly with that recorded at GEOSECS station (19) in The or Profile J (wear ship station) over slope of Norwegian shelf is our only winter profile in Nordic Seas. This profile has an identical pattern to a depth of ca.

12 4 R. Nydal andj. S. Gislefoss &4C (/) -5 5 p14c ( /oo) , 2 a a) d 3 East Greenland Current 4 V Cony. Trondheim, LA8, 6.8.9, 7852'N 4'6'W V AMS. Arizona, LA17, , 79'4'N 346'W C &4C (/) West Spitsberge Current TTONAS, sta154, , 78 34'N 9'28'E TTONAS, sta156, , 78 27'N 8 19'E V Cony. Trondheim, LA7, 3.8.9, 79 27'N 5 52'E V AMS. Arizona, LA7, 3.8.9, 79 27'N 5 52'E &4C (1) E, 2 a a) b 3 4 :East Greenland Current V AMS. Arizona, LA14, , 75 'N 1'W v AMS. Arizona, LA15, , 7459'N 1131'W b 3 4- Central Greenland Sea GEOSECS, sta17, , 74 56'N 1'7'W o TTONAS, sta148, , 74'56'N 1 8'W V Cony. Trondheim, LA1, , 7459'N 229'1 V AMS. Arizona, LA1, , 74'S9'N 2'29'W Fig. 5a. 14C depth profiles in Greenland Sea

13 Bomb 14C as a Tracer in Atmosphere and Ocean 41 &4C ( /o) -5 5 D14C ( /oo) -5 5 l p 3 4 V Cony. Trondheim, GS17, 1.8.9, 69 31'N 14 5'W v AMS. Arizona, M14, , 7 22'N 18 'W Q14C ( /) West Norwegian Sea 4 TTONAS, sta159, , 68'43'N 1.34'W x Cony. Trondheim S 'N 7.3'W V Cony. Trondheim GS '8'N 7'29W V AMS. Arizona GS N 729W i 14C ( /o) a 4) b ! ' ' 4 GEOSECS, sta18, , 7''N O1'W Cony. Trondheim, GS19, , 69 57'N 9 38'E TTONAS, stal45, , 7 'N 2 29'E AMS. Arizona, GS19, , 69 57'N 9 36'E V Cony. Trondheim, GS 16, 7.8.9, 7 'N 1'E AMS. Arizona, GS16, 7.8.9, 7 'N 1'E Fig. 5b.14C depth profiles in Norwegian and Icelandic Seas

14 42 R. Nydal and J. S. Gisle foss &4c (/) -SO 5 &4C (/x) b D GEOSECS, sta19, , 6412'N 5 34'1 o TTONAS, sta144, , 67 41'N 3'2'W V AMS. Arizona, GS14, 3.7.9, 67''N 5 'W AMS. Arizona, MO1, , 67.'N 4.'1 d14c (/) -5 5 I 4 AMS. Arizona, PFI1, , 66 N 2 E &4C ( /oo South 'Norwegian 'S 3 North Atlantic Ocean - I 4aoo 4- GEOSECS, sta19, , 64 12'N 5 34'W GEOSECS, sta23, , 6 25'N 18 37'1Y v AMS. Arizona. JH5, , 64 'N 5 'W V AMS. Arizona, M16, , 62'35'N 15 31'W Fig. 5c. 14( depth profiles from Norwegian Sea to south of Iceland

15 Bomb 14C as a Tracer in Atmosphere and Ocean 43 2 m where recorded O14C value is ca. -5%o. A similar value is recorded in all profiles (B,H,J,K) taken along shelf. The most important changes in 14C depth profiles appear across ridge and down to North Atlantic Ocean. In profiles L (M16) and M (JH9) our data are compared with GEOSECS profile (Sta. 23) 1-2 furr south; JH9 is also compared with TTO profile (Sta. 142) some 8 furr east (Fig. 5c,d). The two recent profiles, which are slightly apart, have same trend as GEOSECS profile. The comparison shows that a rapid downwelling occurs south of ridge. The i14c value below 1 m is variable between +1 and -1%o, and this feature indicates that NSDW makes a very small contribution to overflow of water into formation of AODW. The downwelling water consists of mainly surface- and intermediate water, a result which is in accordance with that earlier pointed out by Heinze et al. (199). The TTO profile from 1981 represents shallower water collected on ridge furr east. Profile N, for North Atlantic Ocean, compares our recent data (JH1) with GEOSECS record (Sta. 11) obtained 3 furr west (below Denmark Strait) in 1972, and five TTO stations (164, 169, 17, 171) in same general area sampled in The deepwater reflects surface and intermediate water from north of ridge, in agreement with Strass et al. (1993) The TTO and GEOSECS profiles agree fairly well, but show a marked deviation from our profile (JH1). This raises question as to wher this is caused by an unknown accident in sample treatment at this location, or by a temporary aberration due to local circumstances. During a later cruise (Nordic WOCE 1994) we were not able to reproduce this curve, but obtained data more in agreement with TTO result. 14 ( /) -SO SO 14C (/) ' 1 1 a 'd D 4.yis FVM a b,c 2 3.._.. _...,...., _.... North Atlantic ;Ocean D GEOSECS, sta23, , 6 25'N 18 37'W TTONAS, stai42, , 61 2'N 8 1'W AMS. Arizona, JH9, '31'N 16 2'W 3 4 o GEOSECS, stall, , 63 32'N 35 14'W TTONAS, stal64, , 623'N 3'24'W V TTONAS, stal67, , 64.5'N 3319'W 8 TTONAS, stal69, , 64 39'N 34'5'W TTONAS, sta.17, , 65'6'N 34'31'W a TTONAS, sta171, , 63 4'N 33'19'W AMS. Arizona, JH1, 'N 32 3'11 Fig. Sd.14C depth profiles in Atlantic Ocean south of Iceland

16 44 R. Nydal and J. S. Gislefoss SUMMARY AND CONCLUSION In addition to those progressive changes recorded in concentration of CO2 in atmosphere, time-dependent distribution pattern for "bomb 14C" introduced into upper atmosphere during nuclear weapons test programs is an important tool for testing models that describe carbon exchange in nature. Here we have presented and discussed several such 14C data sets recorded via samples collected over past 3 yr from lower atmosphere (troposphere) and oceans. The trend in atmospheric 14C concentrations recorded from norrn Norway and Canary Islands show large seasonal variations during 196s due to net downward transfer of major excess of "bomb 14C" that had been injected directly into stratosphere. Both curves indicate that, after 1972, troposphere can be considered in general as a single well-mixed reservoir of "bomb 14C", but with some very small localized disturbances still evident. Both sampling stations record an almost exponential decrease in concentration of excess 14C, with a rate constant of.55 yr-1. Where small seasonal variations still occur, se are in main due to localized dilution by 14C- free CO2 produced by increased combustion of fossil fuels in winter. By 1992, e14c level in lower atmosphere was ca. +1%o above pre-bomb level, and this was equal to 14C enrichment recorded in surface ocean water at equatorial latitudes (45 N to 45 S). For past 25 yr, trend in e14c in equatorial surface ocean can be approximated by a horizontal line, i.e., no significant temporal variation has occurred during that time. This feature reflects role of water mass as an effective buffer to 14C exchange between atmosphere and intermediate and deep ocean carbon reservoirs. The e14c profile of Atlantic Ocean surface water shows an approximate latitudinal symmetry around equator poleward to 6 N/S, with most stable regions coincident with atmospheric high pressure cells at 3 N and 3 S. A slight decrease in 14C concentration occurs at equator, but a more dramatic lowering is evident toward higher latitudes, where more stable surface layer vanishes. The rate of furr decrease in amount of "bomb 14C" in atmosphere will be governed mainly by ongoing exchange of CO2 with deep ocean. For past five years, we have attempted to use 14C as a tracer to study transfer of carbon from Nordic Seas to deepwater reservoir of Atlantic Ocean. Several deep-sea profiles have been produced to cover North Atlantic Ocean from south of Iceland northward to Fram Strait. These data have been compared with similar profiles obtained during GEOSECS expedition in 1972 and TTO expedition in Our measurements confirm that water that is moving southward over Greenland-Scotland Ridge into deep Atlantic derives mainly from surface and intermediate depths in Nordic Seas. The deepwater in central region of Norwegian Sea is too dense to have an important role in mass transfer over ridge. The obtained age of NSDW is ca. 1 yr in case that main water derives from surface of Greenland Sea. The additional exchange of water between NSDW and EBDW may, however, modify this result. ACKNOWLEDGMENTS Atmospheric 14C samples on Izana, Tenerife, and Fruholmen, Nordkapp were collected by several people, most of m mentioned in a previous paper (Nydal and Lovseth 1983). Sampling during last ten years was kindly performed by Ramon Juega Buide, Observatorio Especial de Izana (Tenerife), and Odd Salomonsen, Fruholmen Lighthouse, Ingoy. We are furr greatly indebted to those earlier-mentioned people in Wilhelmsen and Fred Olsen shipping companies for collecting samples on surface of Atlantic, Pacific and Indian Oceans (Nydal et a1.1984). Sampling and processing during last ten years was performed mainly on a single ship, MS Tourcoing in Barber

17 Bomb 14C as a Tracer in Atmosphere and Ocean 45 Line (Wilh. Wilhelmsen), and we are especially thankful to chiefs, Harald Jorgensen, Kristian Larsen, Frank Christoffersen, Per Strandklev and Tor Olsen. Thanks are also due to Chief Thorvald Benjamensen on board RV Andenes for collecting and processing surface samples on a return trip to Antarctic in winter We are also thankful to captains and crews on vessels Lance, Mosby, G.. Sars, Johan Hjort and Polarfront during cruises in Nordic Seas from 1989 to We give special thanks for helpful assistance and discussions with Johan Blindheim, Institute of Marine Research Bergen; Torgny Vinje, The Norwegian Polar Research Institute, Arne Foldvik and Svein Qsterhus, Geophysical Institute, University of Bergen, and KAre Misje, Misje Offshore Marine, Bergen. The latter is in charge of wearship Polarfront in Norwegian Sea. We are furr indebted to Indunn Skjelvan (previously on our staff), now at Center for Studies of Environment and Resources (CARDEEP), Bergen, for processing CO2 samples on cruises with Lance in 199, and Hdkon Mosby in Thanks also to staff of Radiological Dating Laboratory, especially Fred H. Skogseth, for careful work in processing and counting large number of conventional 14C samples. Most of deep-sea 14C measurements were performed by accelerator mass spectrometry at Arizona AMS Facility, Tucson; and we are especially grateful to A. J. Timothy Jull, Douglas J. Donahue and Laurence J. Toolin. These AMS measurements were partly supported by grant EAR from U.S. National Science Foundation. The main financial support during many years of this work has been provided by Research Council of Norway (earlier NAVF). REFERENCES Bonisch, G. and Schlosser, P Deep water formation and exchange rates in Greenland/Norwegian Seas and Eurasian Basin of Arctic Ocean as derived from tracer balances. Progress in Oceanography 35: Bourke, R. H., Paquette, R. G., Bly, R. F. and Stone, M. D On deep and bottom waters of Greenland Sea from summer 1989 to 199 data. Journal of Geophysical Research 98(C3): Broecker, W. S. and Peng, T.-H Tracers in Sea. Palisades, New York, Eldigio Press: 69 p. Broecker, W. S., Peng, T.-H., Ostlund, G. and Stuiver, M The distribution of bomb radiocarbon in ocean. Journal of Geophysical Research 9(C4): Donahue, D. J., Linick, T. W. and Jull, A. J. T. 199 Isotope-ratio and background corrections for accelerator mass spectrometry radiocarbon measurements. Radiocarbon 32(2): Feely, H. W., Katzman, D. and Tucek, C. S th Progress Report Project Stardust, DASA. Foldvik, A. and GammelsrOd, T.1988 Notes on sourn ocean hydrography, sea-ice and bottom water formation. Paleogeography, Paleoclimatology and Paleoecology 67: Gislefoss, J. S Carbon profiles in Nordic Seas. PhD. dissertation, NTH-Trondheim:17 p. Gislefoss, J., Nydal, R., Donahue, D. J., Jull, A. J. T. and Toolin, L. J Tracer studies of 14C in Nordic Seas by AMS measurements. Nuclear Instruments and Methods in Physics Research B92: Gislefoss, J. S., Nydal, R., Skjelvan, I., Nes, A., Q sterhus, S., Holm6n, K., Jull, T. and Sonninen, E Carbon profiles in Nordic Seas. Data Report 2. Radiological Dating Laboratory, Trondheim, Tapir: 61 p. Heinze, C., Schlosser, P., Koltermann, K. P. and Meincke, J. 199 A tracer study of deep water renewal in European polar seas. Deep-Sea Research 37(9): Levin, I., Graul, R. and Trivett, N. B. A Long-term observations of atmospheric CO2 and carbon isotopes at continental sites in Germany. Tellus 47B (Series B): Linick, T. W., Jull, A. J. T., Toolin, L. J. and Donahue, D. J Operation of NSF-Arizona Desolator Facility for Radioisotope Analysis and results from selected collaborative research projects. In Stuiver, M. and Kra, R. S., eds., Proceedings of 12th International 14C Conference. Radiocarbon 28(2A): Mangerud, J. and Gulliksen, S Apparent radiocarbon ages of recent marine shells from Norway, Spitsbergen and Arctic Canada. Quaternary Research 5: Meijer, H. A. J., van der Plicht, J., Gislefoss, J. S. and Nydal, R Comparing long-term atmospheric 14C and 3H records near Groningen, The Nerlands with Fruholmen, Norway and Izana, Canary Islands. Radiocarbon 37(1): Nakamura, T., Nakazawa, T., Nakai, N., Kitagawa, H.,

18 46 R. Nydal and J. S. Gislefoss Honda, H., Itoh, T., Machida, T. and Matsumoto, E Measurement of 14( concentrations of stratospheric CO2 by accelerator mass spectrometry. In Long, A. and Kra, R. S., eds., Proceedings of 14th International 14C Conference. Radiocarbon 34(2): Nydal, R Furr investigation on transfer of radiocarbon in nature. Journal of Geophysical Research 73(12): Nydal, R Application of bomb 14C as a tracer in global carbon cycle. Trends in Geophysical Research 2: Nydal, R., Gislefoss, J., Skjelvan, I., Blindheim, J., Foldvik, A., Vinje, T. and fsterhus, S Measurements of carbon profiles in Nordic seas. Norsk Polarinstitutt Rapportserie Datareport 75:1-43. Nydal, R., Gislefoss, J., Skjelvan, I., Skogseth, F. H., Jull, A. J. T. and Donahue, D. J C profiles in Norwegian and Greenland Seas by conventional AMS measurements: In Long, A. and Kra, R. S., eds., Proceedings of 14th International 14( Conference. Radiocarbon 34(3): Nydal, R., Gulliksen, S., Lovseth, K. and Skogseth, R H Bomb 14C in ocean surface, Radiocarbon 26(1): Nydal, R. and Lovseth, K Tracing bomb 14C in atmosphere Journal of Geophysical Research 88(C6): Ostlund, H. G., Dorsey, H. G. and Brecher, R. (ms.) 1976 GEOSECS Atlantic, radiocarbon and tritium results. Data report from Rosenstiel School of Marine and Atmospheric Sciences. Florida, University of Miami: 93 p. Ostlund, H. G. and Rooth, C. G. H. 199 The North Atlantic tritium and radiocarbon transients Journal of Geophysical Research 95(C11): 2,147-2,165. Pickard, G. L. and Emery, W. J. 199 Descriptive Physical Oceanography. Oxford, Pergamon Press: 32 p. Schiosser, P., Bonisch, G., Rhein, M. and Bayer, R Reduction of deep-water formation in Greenland Sea during 198s: Evidence from tracer data. Science 251: SIPRI Yearbook 1975 World Armament and Disarmaments. Stockholm, Almqvist & Wiksell. Slots, P. J., Jr., Jull, A. J. T., Linick, T. W. and Toolin, L. J Preparation of small samples for 14C accelerator targets by catalytic reduction of CO. Radiocarbon 29(2): Smethie, W. M., Jr., Ostlund, H. G., and Loosli, H. H Ventilation of deep Greenland and Norwegian seas: Evidence from krypton-85, tritium, carbon- 14 and argon-39. Deep-Sea Research 33: Strass, V. H., Fairbach, E., Schauer, U. and Sellmann, L Formation of Denmark Strait overflow water by missing East Greenland Current. Journal of Geophysical Research 9(C4): Stuiver, M. and Polach, H. A Discussion: Reporting of 14C data. Radiocarbon 19(3): Swift, J. H., Aagaard, K. and Malmberg, S.-A. 198 The contribution of Denmark Strait overflow to deep North Atlantic. Deep-Sea Research 27A: Swift, J. H. and Koltermann, K. P The origin of Norwegian Sea deep water. Journal of Geophysical Research 93: Tans, P A compilation of bomb 14C data for use in global carbon model calculation. In Bolin, B., ed., Carbon Cycle Modeling. SCOPE: 39 p. U.N. Report 1964 Radioactive contamination of environment by nuclear tests, effect of atomic radiation. United Nations Scientific Committee Report 14 (Al 5814):12 p.