Radium-226 in the Kuroshio Water Near Taiwan : ·

This paper presents so.me initial results of the 226Ra measurements on seawaters from the areas off northern Taiwan where the Kuroshio water en• counters the East China Sea water and off eastern Taiwan within the Kuroshio current. Samples were collected from the sea surface and also in the water col11111n during the KEEP-MASS cruise onboard the R/V Vinogradov and three other cruises onboard the R/V Ocean Researcher L · . Surface water 226Ra varies somewhat randomly between 6.1 and 12.8 dpm/100 kg in the study area with a sliaht tendency to increase northward along the Kuroshio path off eastern Taiwan. Similar but less random varia. . . tions were observed by Nozaki in the northern East China Sea surface water. The shallow water 226Ra profiles in the shelf and upper slope areas show fairly uniform values at about 10 dpm/100 kg. The inter1nediate and deep water profiles in the lower slope and deep basin off the east coast of Taiwan show a monotonous increase from south to north in the layer between the depths of 300 m and 1000 m. The deep profiles at the south display a pronounced 226Ra increase from about 12 dpm/100 kg at 1000 m to about 30 dpm/100 kg at 2500 m and remain relatively constant below. Althouah the northward 226Ra increase in the inter1nediate depth along the Kuroshio is contradictory to the conventional notion predicting no signifi· cant change along the ftow, it is nonetheless consistent with the hydrographic data which show clear changes from south to north. Published 210Pb profiles in the study· area also show significant changes both in shape and activity level. Comparisons between these 210Pb profiles with the 226Ra profiles measured in adjacent areas in this study indicate that the 210Pb excess due to ·the atmospheric ftux may penetrate down to about the depths of 300 m to 1000 m. As the decay rate of the excess 210Pb is small compared to the atmospheric ftu:x, the scavenging residence time for 210Pb is controlled mainly by the atmospheric flu:x and the inventory of exce• 210Pb. The 210Pb scavenging residence time calculated for the excess 210Pb based 1 Institute of Marine Geology. National Sun Vat-Sen University, Kaohsiung, Taiwan, R.O.C.

Surface water 22 6 Ra varies somewhat randomly between 6. 1 and 12. 8 dpm/100 kg in the study area with a sliaht tendency to increase northward along the Kuroshio path off eastern Taiwan.Similar but less random varia-.
tions were observed by Nozaki in the northern East China Sea surface water.
The shallow water 22 6 Ra profiles in the shelf and upper slope areas show fairly uniform values at about 10 dpm/100 kg.The inter1nediate and deep water profiles in the lower slope and deep basin off the east coast of Taiwan show a monotonous increase from south to north in the layer between the depths of 300 m and 1000 m.The deep profiles at the south display a pronounced 22 6 Ra increase from about 12 dpm/100 kg at 1000 m to about 30 dpm/100 kg at 2500 m and remain relatively constant below.
Althouah the northward 22 6 Ra increase in the inter1nediate depth along the Kuroshio is contradictory to the conventional notion predicting no signifi• cant change along the ftow, it is nonetheless consistent with the hydrographic data which show clear changes from south to north.
Published 210Pb profiles in the study • area also show significant changes both in shape and activity level.Comparisons between these 210Pb profiles with the 22 6 Ra profiles measured in adjacent areas in this study indicate that the 210Pb excess due to • the atmospheric ftux may penetrate down to about the depths of 300 m to 1000 m.As the decay rate of the excess 210Pb is small compared to the atmospheric ftu:x, the scavenging residence time for 210Pb is controlled mainly by the atmospheric flu:x and the inventory of exce• 210Pb.
The 210Pb scavenging residence time calculated for the excess 210Pb based 1 Institute of Marine Geology.National Sun Vat-Sen University, Kaohsiung, Taiwan, R.O.C.
48 TAO, Vol.6, No.l, March 1995   on a box model ranges from 2.3 to 8. 8 yrs, which is proportional to the depth of the excess 210Pb penetration.• The 210Pb/22 .
6 Ra activity ratio is about 0.25 below the 1000 m depth in the area off northeastern Taiwan, giving a scavenging l'esidence time of about 10 yrs for Pb.This value is quite comparable to that obtained from the botto m water in the deep .• open Oceans.

INTRODUCTION
226 Ra, with a half-life of 1622 years, has been widely used as a tracer in the oceans for mixing and circulation studies (e.g., Koczy, 1958; Broecker et al., 1967; Ku et aL, 1970;   Chung and Craig, 1973).Large-scale systematic measurements of 226 Ra were undertaken in the GEOSECS (Geochemical Ocean Sections Study) program which encompassed the Atlantic, Indian, Pacific and Circumpolar Oceans (Broecker et al., 1976; Ku and Lin, 1976;  Chung and Craig, 1980; Ku et aL, 1980•; Chung, 1984; Ku and Luo, 1994).
The Kuroshio is a warm, saline western boundary current flowing northward along the east coast of Taiwan ' (Nitani, 1972).Upon reaching the continental margin off northeastern Taiwan, it turns northeastward and also intrudes into the East China Sea, creating a frontal zone mixing and upwelling of the subsurface water in the area (Fan, 1980;Liu, 1983;Liu and Pai, 1987;Chem and Wang, 1989;Wong et al., 1991;Liu et al., 1992).It should Ra data in this study and the published 210 Pb data (Lin and Chung, 1991;Yang and Lin, 1992).• •

SAMPLE COLLECTIONS AND MEASUREMENTS
Seawater samples of 20-liter size were collected during the KEEP-MASS cruise in July August, 1992 and three other Ocean Researcher I (ORI) cruises in 1992and 1993(Cruise 314 in April and Cruise 338 in December, 1992, and Cruise 348 in March, 1993).The   The surface water samples were collected using either a plastic bucket (OR 1 Cruise 314 and Cruise 338, nominal depth = •o m) or a 20-1 GoFlo bottle mounted on a hydrowire and tripped at 4 m (KEEP-MASS cruise).Each water column sample from station 14 (Cruise 314) was collected using 10 Niskin bottles of 2.5-1 size mounted on a CTD rosette; the KEEP MASS profile samples were collected with 20-1 GoFlo bottles mounted on a hydrowire and tripped at the nominal depth.The remaining water column samples (C15, C16, Tl 1 ') were collected with 20-1 GoFlo bottles on a CTD rosette.The water depth of these stations ranges from about 100 m to over 4500 m.
The collected samples were transferred on board to acid-cleaned 20-1 plastic containers to be shipped back to the laboratory where they were transferred into pre-cleaned 20-1 thick wall plastic bottles.Each of the bottles had an air-tight cap with a gas inlet and outlet.The inlet tube was attached to an air bubbler.After being transferred into the bottle, the sample was fi rst bubbled with helium as a carrier gas to remove all the radon ( 222 Rn) and other dissolved gases and then sealed for ingrowth of 222 Rn from 226 Ra decay.so TAO, Vol.6,No . . 1, March 1995 About a month later when 222 Rn and 226 Ra were in secular equilibrium, the 222 Rn in the bottle was stripped out again for measurement of the activities of radon and its two short lived alpha daughters using an alpha scintillation counter.Radon extraction arid purification were both performed using a radon stripping board designed at Scri��s for the GEOSECS 22 6 Ra program (Chung and Craig, 1980).The technique used for the 6 Ra measurement by the regenerated radon method and associated data reduction, etc. were previously described in detail (Chung, 1971).The 226 Ra standards used for calibration are those of NIST (NBS) used in the GEOSECS program.The overall precisions of the data presented here are better than ± 10% based on duplicate measurements.RN Vino g radov KEEP-MASS Cruise (7/10/-8/5/1992) showing neither a coherent relationship nor a systematic trend.The 226 Ra concentrations vary between 6. 1 and 12.8 dpm/100 kg; large differences relative to the measurement errors among adjacent stations are often observed.In the upwelling cold dome north of Keelung (Fan, 1980;Liu, 1983;Liu and Pai, 1987;Liu et al., 1992), the 226 Ra values are somewhat lower rather than higher compared to those near the mainland China and on the Kuroshio.

SURFACE WATER 226 Ra
A vague trend of a northward increase is discernible along the east coast of Ta iwan.:z, • .?• • . •h.

C )
if .
---. • Both features appear to be contradictory to the conventi.onal• wisdom that the 226 Ra concentration gradient, if any, should be in the west to east rather than in the north to south direction in the Kuroshio east of Ta iwan and that the 226 Ra should be higher rather than lower in the upwelling area.Thus, a more closely-spaced sampling within a cruise or short period is needed to •resolve the details.
The sutface water 226 Ra in the study area has a large variation th�t is not observed in the surface water of any open oceans where 226 Ra usually varies between 6 and 8 dpm/100 kg (Broecker et al., 1976; Chung and Craig, 1980; Ku et al., 1980).However, the observed variation, albeit random and large, is somewhat similar to that observed in the northern East China Sea by Nozaki (1989).-Marginalseas have a higher surface water 226 Ra variation probably due to localized river fluxes .and diffusion from different bottom sediments as well as from coastal zones of various landmas• s.A large random spatial variation in surface water 226 Ra content warrants a further detailed investigation.

HYDROGRAPHY
Figure 3 shows composite potential temperature, salinity, dissolved oxygen, and silicate profi les from stations Tl l, Tl4, T17 and 14 roughly along the path of Kuroshio from south to north (Figure 1 ).Although quite scattered, these profiles basically show some common gross hydrographic features, i.e. a sharp subsurface salinity maximum ( rv200 m), a salinity minimum near 500 m depth (Figure 3B), and an oxygen minimum around 800 m depth (Figure 3C).Dissolved .oxygen and silicate profiles (Figures 3C and 30) may be grouped into two types: the stations at the south and the center (Tl l, Tl4) show higher dissolved oxygen and lower silicate, while the stations farther north (Tl 7, 14) show lower dissolved oxygen and higher silicate above the depth of the salinity minimum.It is not clear why these changes take place within a short distance of 90 km along and within the Kuroshio (Figure 1) between Tl4 and T17, since about 150 km to the south of Tl4 and away from the Kuroshio path, Tl I shows the same profile characteristics as T14.
The secondary features revealed in Figure 3 are: the potential temperature profiles of Tl 1 and T14 are identical while that of Tl 7 has a higher temperature below the depth of 500 m and a lower temperature above this depth (Figure 3A); the salinity profiles of Tl 1 and T14match fairly well above the salinity minimum but below this, T14 has higher values (Figure 3B ); the salinity profiles of T 17 and 14 are identical and show a lower salinity• maximum and higher salinity minimum, suggesting a northward ''erosion'' of the extrema; the dissolved oxygen minima at Tl l and T14 are lower than those at T17 and 14, also showing a northward erosion or the effect of vertical mixing.
These hydrographic profiles appear to be inconsistent with the conventional wisdom which would predict no significant hydrographic variations within a relatively short distance along the advective path of the Kuroshio.The changes of all the hydrographic parameters take place between T14 and T17.A detailed CID profiling between these two stations is required to verify there changes.However, these features are not inconsistent with the 226 Ra profiles which show a northward increase from Tl l, via T14 to Tl7 and 14 at the corresponding depth range (see the next sectiop.) .• Cl.

SS
but it cannot be ruled out as a measurement error.However, the hydrographic data at this station do not support the existence of this minimum.Higher sampling density around this depth together with CTD and.other hydrographic observations is required to resolve this problem.If this minimum is real, then it may serve to indicate the spreading of the Antarctic Bottom Water which is known to have a lower 226 Ra concentration (Chung and Craig, 1980).
The general pattern of the deep water profiles is quite similar that observed from the deep North Pacific, and the concentration level is consistent with that expected from the known Pacific profiles (Chung and Craig, 1980).

PROPERTY VS PROPERTY RELATIONSHIPS
As expected from the hydrographic features and 226 Ra distributions (Figures 3 and 4), the property vs property plots between parameters can better illustrate the trend or change in their relationships from south to north.Plots of potential temperature vs. salinity (T-S • diagrams), salini� vs. dissolved oxygen (S-0 2 ), potential temperature vs. silicate (T-Si), and silicate vs. 2 6 Ra (Si-226 Ra) are shown for comparisons of stations Tl l, Tl4, T17 and 14 (Figure 5).
The T-S diagrams show that both the salinity maximum and minimum are fairly close for Tll and T14, but this maximum is higher than, and this minimum is lower thari that for T17 and 14.These extrema for T17 and 14 are also quite close to each other except for temperatures greater than 18°C (Figure 5A).Below the salinity minimum, the plots of T14 are parallel to, and higher than those of Tl 1 mainly due to the higher salinity at T14 (Figure 3).The salinity vs. oxygen plots show a loop due to the fact that the depth of the 0 2 minimum is greater than that of the S minimum.The salinity minimum at Tl 7 and 14 is clearly higher than at Tl 1 and T14 (Figure 5B) as also seen in the TS diagrams.
Tl 7 is systematically higher than other stations in the T-Si plots (Figure 5C) due to the higher T below 500 m and the higher Si above this depth (Figure 3A and 30).The high Si content is associated with the high 226 Ra concentration as shown in the Si-226 Ra plots (Figure 50).226 Ra shows a large increase from Tl 1 at the south to Tl 7 at the north in the silicate values between 25 and lOOµM.Such silicate vs.
226 Ra relations are far from linearity, suggesting that silicate is not a major carrier for 226 Ra.
7. COMPARISONS OF 22 6 Ra AND 210 Pb PROFILES The extent of 210 Pb/ 226 Ra disequilibrium is a useful indicator for the scavenging pro cess and the mean residence time of the particulate matter in the deep oceans (e.g., Craig et  al., 1973; Bacon et al., 1976; Chung, 1981; Chung and Craig, 1983; Chung, 1987).There are a few published 210 Pb profiles of the study area (Lin and Chung, 1991;Ya ng and Lin, 1992).Now that 226 Ra profiles are available, it is imperative that the 210 Pb and 22 6 Ra relations be examined.Unfortunately, these 226 Ra profiles are neither located at the 210 Pb stations nor were they collected at the same time.It is not yet clear at present whether a steady-state condition can be established for these tracers.Locations o f the 210 Pb and 226 Ra profiles available for comparison are shown in Figure 6.The 210 Pb profi les at Hl6, R19 and R23 were published (Lin and Chung, 1991) as was that at 4531 (Yang and Lin, 1992)   down to 1000 m based on the 22 6 Ra profi le at Tll, but only down to 500 m based on the profile at Tl4.As R19 is located between • T14 and Tl7, its 210 Pb profi le is compared either with the 226 Ra profile at T14 to the south or T17 to the north (Figure 7B).The 210 Pb profile shows a distinct m�imum at the depth of 100 m rather than at the surface.The two 226 Ra profiles are fairly similar above 400 m, but they display large systematic differences below this depth with T17 being higher.The excess 210 Pb inventory based on 226 Ra profi les at T14 and T17 is about 2.1 and 1.3 dpm/cm 2 , respectively.The penetration depth of the excess 210 Pb is about 300 m.
The two 210 Pb profiles (H l6 and 4531) are similar in showing surface maximum with • values signifi cantly higher than those of the other stations to the south (Figure 7).However, these two profiles cross over each other twice, and the lower values are at 4531 below 200 m depth.The three 22 6 Ra profiles shown in Figure 7C are located very close to one another .but farther away from 816 and 4531 (Figure 6); they are about identical above 500 m but scattered below.The inventory of excess 210 Pb is about 2.3 dpm/cm 2 at H16 and about 1.7 . dpm/cm 2 at 4531.The penetration depth is 360m at H16 and 300 ma t 4531.Below 1000 m the 210 Pbl 22 6 Ra activity ratio (H16 vs Tl 7) is about 0.25, corresponding to a scavenging residence time of about 10 yrs.This agrees with the previous estimate by Lin and Chung (1991).Using this equation, the '11 value for each pair of the inventories within the box can be calculated.The scavenging residence time, T w, is simply 1/\11 .Both W and rw are listed in Table 3 which also summarizes relevant data used for the calculations.The T w values range between 2.3 and 8.8 yrs.The high value of 8.8 yrs is due to the deep penetration (compared with the Tl I 226 Ra profile) resulting in a high 210 Pb inventory.The other T w values are, in general, still greater than the 210 Pb scavenging residence time of about 1.5 .' ' •.

.
be of great interest to see • how 226 Ra behaves and distributes under the influence of the Kuroshio and its exchan e with the ambient water.Furthermore, the extent of radioactive into the particulate scavenging mechanism as influenced.by the advective and upwelling processes.However, until now no 226 Ra data have been collected in the southern Kuroshio water near Taiwan. . .The purpose of this paper is to present for the first time the results of 226 Ra measure ments made on water samples collected from the Kuroshio off eastern and northeastern Taiwan on the KEEP-MASS (Kuroshio Edge Exchange Processes-Marginal Sea Study) cruise of the Russian R/V Vinogradov in July-August, 1992, and on three cruises of the R/V Ocean Researcher I in 1992 and 1993.Besides discussing the 226 Ra results in relation to hydrog ra :&hy, the authors also re-evaluate the status of radioactive disequilibrium between 210 Pb and 2 Ra at several stations with special attention to the 210 Pb excess in surface and subsurface waters bCl:Sed on the 226 .

Fig. 1 .
Fig. 1.Station locations for Ra-226 measurements.Station number prefixed with a T (except Tl 1 ') or an E was occupied by the R/V Vinogradov for the KEEP-MASS program.The Kuroshio path and its direction are indicated.
Fig. 2. Surface water temperature and Ra-226 distributions off eastern and north eastern Ta iwan.The temperature values are underlined.

Fig. 4 .
Fig. 4. Ra-226 profiles shown in three different depth-scales.A northward increase in Ra-.226 in the intermediate depth between 200 m and 1000 m is illustrated in B.
Yu-Chia Chung & Hung-Chou Yin St mean of the duplicate measurements; the error quoted for the mean is one standard deviation of the repeated measurements or of the overall counting statistics, whichever is greater.No hydrographic data are available for several stations of the KEEP-MASS cruise and all the stations of the Cruises 338 and 348.The KEEP-MASS temperature data were reported to one hundredth of a degree centigrade(Chen, 1992).Station 14 of Cruise 314 is also a profile station ; E24a is a profile station from the East China Sea.

Table 2 .
(Continued) (Lin and Chung, 1991) over 226 Ra from the surface water down to a few hundred meters due to the atmospheric flux, and it becomes deficient below _this depth due to scavenging by Yu-Chia Chung & Hung-Chou Yin 57 settling particles.Based on a 22 6 Ra profile calculated from a silicate profile in the Okinawa trough through a linear Ra-Si relationship(Lin and Chung, 1991), the depth of excess 210 Pb penetration was estimated at 500 m.These features are quite similar to those observed in the open oceans except for the penetration •depth which is usually about 800 m-1000 m in the open oceans.