Scavenging Phenomenon Elucidated from 234 Th / 238 U Disequilibrium in Surface Water of the Taiwan Strait

Concentrations of dissolved (Thd) and particulate (Thp) 234Th in surface water at 38 stations in the Taiwan Strait were determined for samples collected in May 2006. The spatial distribution of 234Th in the Taiwan Strait is controlled by advective input of Kuroshio Branch Water via the Peng-Hu Channel and fast removal due to the high input of riverine particulates from the Cho-Sui River. A scavenging model involving physical transport was applied to the Thd and Thp data to estimate scavenging and removal rates of 234Th. Estimated scavenging rate ranges from 32 to 703 dpm m-3 d-1 and the removal rate ranges from 24 to 560 dpm m-3 d-1. Using 234Th as a proxy of particulate organic carbon, we estimate that the removal rate of POC from surface water of the Taiwan Strait ranges from 0.3 ± 1.5 mmol-C m-3 d-1 off southwestern Taiwan to 10.2 ± 3.5 mmol-C m-3 d-1 in the central Taiwan Strait.


INTRODUCTION
The complexity of the coastal system is mainly due to strong and variable physical forcing and significant input of allochthonous materials from the landmass.The Taiwan Strait is a shallow shelf with an average depth of 60 m and serves as an important conduit for water exchange between the South China Sea and the East China Sea (Fig. 1).A large northward volume transport of 2.0 ~ 2.2 Sv was measured by Chung et al. (2001) and the transport is mainly through the Peng-Hu Channel (Jan and Chao 2003).Because this shallow and narrow water mass is bordered by the heavily populated China and Taiwan coastline, the Taiwan Strait receives large amounts of anthropogenic contaminants discharged from municipal areas through river inputs.
The combination of seismic activity and frequent typhoon results in extremely high sediment input from mountainous rivers of Taiwan.A total of 194 Mt y -1 fluvial sediments are carried into the Taiwan Strait by the mountainous rivers of western Taiwan (Dadson et al. 2003).While seawater is fed from the south, it is conceivable that the fluvial sediments contributed from the rivers on both sides of the Taiwan Strait not only serve as an important source of terrestrial materials but also as an efficient scavenger and carrier for the particles and their associated trace elements and nutrients.To understand the fate of trace metals either carried from the south by thecurrent or discharged from the land into the Taiwan Strait, it is essential to obtain more information on the rate of processes controlling the budgets of elements of interest.By using suitable naturally-occurring radionuclides, better knowledge regarding rates of processes governing trace element cycling in the Taiwan Strait can be greatly improved.Among those radionuclides, the shortlived 234 Th (t 1/2 = 24.1 d) is an ideal tracer for this purpose.

234
Th is constantly produced from 238 U in seawater.Once produced, 234 Th is quickly attached to particle surfaces and removed from the water column.Depending on how fast 234 Th is removed from a given parcel of seawater, the degree of deviation from secular equilibrium with 238 U Terr. Atmos.Ocean. Sci., Vol. 21, No. 4, 713-726, August 2010 varies in different regimes of the ocean. 234Th/ 238 U disequilibrium is suitable for the investigation of scavenging and removal processes on timescales from days to weeks. 234Th has been widely used as a proxy of particulate organic carbon (Buesseler et al. 2006 and references therein), particulate inorganic carbon (Bacon et al. 1996), organic pollutants (Gustafsson et al. 1997), and trace metals (Weinstein and Moran 2005) to investigate elemental removal processes in the upper water column of the open oceans.However, due to the dynamic nature of coastal environments, there have been few studies applying the 234 Th/ 238 U disequilibrium in coastal oceans (Santschi et al. 1979;Baskaran and Santschi 1993;Gustafsson et al. 1998;Benitez-Nelson et al. 2000;Radakovitch et al. 2003).To the best of our knowledge, no investigation of scavenging processes is available for the Taiwan Strait.
In May 2006, we took the opportunity of the Joint Hydrographic Survey launched by the National Center for Oceanographic Research of Taiwan (http://www.ncor.ntu.edu.tw/odbs/JHS/index.html,NCOR Report, 2006) to con-duct seawater sampling in the Taiwan Strait for 234 Th analyses.This paper aims to provide a quantitative estimate of scavenging and removal rates of 234 Th by using 234 Th/ 238 U disequilibrium.

MATERIALS AND METHODS
At the stations shown in Fig. 1, large-volume seawaters were sampled between 21 and 27 May 2006 as part of the Joint Hydrographic Survey project aboard three research vessels: the R/V Ocean Researcher I (cruise #796), the R/V Ocean Researcher II (cruise #1353), and the R/V Ocean Researcher III (cruise #1153).Seawater was collected at a depth of 2 meters using 10-or 20-L Teflon-coated Go-Flo bottles mounted on a Sea-Bird CTD (SBE 9/11) rosette assembly.A total of 38 large-volume (20 L) surface seawater samples were collected for the determination of dissolved ( 234 Th d ) and particulate ( 234 Th p ) 234 Th, and related hydrographic parameters (nutrients, particulate organic carbon and nitrogen).

Fluorescence and Light Transmission
Along the cruise track, continuous measurements of fluorescence and transmission were made by attaching a fluorometer and a transmissometer with 25 cm light path length on the underway system (Wetlabs C-Star) installed onboard.The intake of the underway system is about 2 meters below the surface.

Dissolved 234 Th
Immediately after the seawater was transferred to the pressure drums, compressed air (at 12 p.s.i.) was used to pressure filter seawater through preweighed 142-mm Nuclepore filters (0.45 μm pore size) mounted in a Plexiglas filter holder.The filters were rinsed with approximately 15 ml deionized distilled water to remove sea salt and stored in a petri dish to determine the total suspended matter (TSM) concentration and particulate 234 Th activities.
The filtrate was acidified with about 20 ml concentrated HCl and spiked with 30 dpm of 230 Th and 50 mg of Fe carrier.The samples were bubbled vigorously for at least 3 hours to help attain isotopic equilibration.Concentrated NaOH was then added to raise the pH to ~8 to precipitate Fe(OH) 3 .The Fe(OH) 3 precipitates, with adsorbed thorium, were collected by siphoning and centrifugation, and then dissolved in concentrated HCl to a final concentration of 9 N HCl.These samples were then passed through an anion exchange column (Dowex1X-8) preconditioned with 9 N HCl to separate uranium from thorium.Thorium samples were purified by passing the sample through three anion exchange columns pre-conditioned with 8 N HNO 3 .The sample was evaporated to incipient dryness and converted to a pH~1 solution ready for Th extraction. 234Th and the yield tracer, 230 Th, were extracted from the aqueous phase into a 0.4 M TTA (thenoyltrifluoroacetone)-benzene solution and the organic solvent was stippled on a stainless-steel disc.Preconcentration and separation of uranium and thorium from the filtered seawater samples were completed in three days after samples were collected.

TSM and Particulate 234 Th
After weighing for the determination of TSM concentration, the filters were decomposed by soaking in ~10 ml of concentrated NH 4 OH.The samples were gently heated to evaporate the NH 4 OH and then fluxed in HClO 4 /HF to thoroughly digest organic and inorganic materials.After digestion, the samples were purified and mounted on stainlesssteel discs following the same procedures as with dissolved 234 Th samples.
The activities of 234 Th were counted by a low background (< 0.15 cpm) anticoincidence counter (Riso GM25-5) via its β-emitting daughter 234 Pa.The counting efficiency of the beta counter was regularly monitored by counting a 238 U standard source.Chemical yield of thorium was estimated by counting spiked 230 Th using silicon surface-barrier detectors (EG&G Ortec 576).The counting efficiencies of the α detectors were calibrated against NIST traceable 230 Th (Isotope Products Laboratory 387-67-3) standard plates.Activities of 234 Th reported here were corrected back to the sampling time after the 234 Th ingrown from 238 U was subtracted.

Particulate Organic Carbon
Particulate organic carbon (POC) concentrations were determined by filtering 2 L seawater through precombusted (450°C) Whatman 25 mm GF/F filters.Filters were wrapped in aluminum foil and stored at -4°C.In the laboratory the filters were acid-fumed to remove carbonates, and then analyzed for carbon and nitrogen with a Fisons elemental analyzer (NA1500).A calibration curve was obtained using acetanilide (C 8 H 9 NO) as a standard.The overall procedural errors estimated from duplicate runs were better than ±2% for carbon and nitrogen.

RESULTS
All data obtained from the bottle samples are given in Table 1 and shown as contour maps (Figs. 2 -9).The distributions of surface temperature and salinity are shown in Figs. 2 and 3, respectively.They reveal the contribution of three water masses, the China Coastal Water (CCW), the Kuroshio Branch Water (KBW), and the South China Sea Surface Water (SCSSW).Generally, the hydrography in the Taiwan Strait is controlled by the interaction of the CCW originating from the north with either KBW or the SCSSW from the south.Due to persistent northward currents in the Taiwan Strait, the CCW, which is characterized by low temperature and low salinity, is limited to the northwestern Taiwan Strait.Significant freshening shown in the salinity distribution in the northwestern Taiwan Strait is evidently caused by the influence of riverine input from China coast.The KBW, with high temperature and high salinity, dominates the southeastern part of the Taiwan Strait and reaches as far north as to 24°30'N in the coastal region of western Taiwan.With intermediate temperature and salinity values, the SCSSW occupies most part of the Taiwan Strait.As shown in Fig. 1, a strong and persistent current from the northeastern South China Sea through the Peng-Hu Channel can be seen (Jan and Chao 2003).Chung et al. (2001) concluded that the water mass in this northward current originated from the KBW in May and from the South China Sea Warm Water in August.
Distributions in surface water of fluorescence expressed as relative fluorescence units (RFU) and light beam attenuation coefficient (BAC) are shown in Figs. 4 and 5,   respectively.Generally, the distribution of fluorescence reflects the geographic distribution of chlorophyll concentration.We found higher RFU values in the western and northern portions of the Taiwan Strait.Higher fluorescence water was associated with higher silicate concentration (data not shown).The distribution of fluorescence followed the pattern in concentration of total pigments measured using HPLC in samples collected from the same stations as this study (Tung 2007).The highest and lowest biomasses were found in the Ming River mouth and Cho-Sui River mouth, respectively.According to Tung (2007), the chlorophyll biomass is mostly (> 80%) contributed by diatoms in the region of the Ming River and Cho-Lung River mouths.
The distribution of BAC essentially reflected the geographical variability of TSM, as shown in Fig. 6, and showed a positive relationship between TSM and BAC: TSM (mg L -1 ) = 2.40 × BAC (m -1 ) -0.75, R 2 = 0.88, n = 64.In the vicinity of the Cho-Sui River mouth, the BAC was the highest, indicating fluvial sediment input.The BAC in the mouth of the Ming River mouth showed elevated values, whereas the TSM distribution in the general area remained low.Because the beam attenuation coefficient is also affected by colored dissolved substances (Zaneveld 1994), elevated BAC in the vicinity of Ming River mouth may have been caused by fluvial input of dissolved substances rather than by absorption and scattering by particles.Indeed, Hung et al. (2000;2003) observed a water mass of high dissolved organic carbon (> 90 μm) in the vicinity of the Ming River mouth.Not surprisingly, TSM concentration was higher in the vicinity of Cho-Sui River mouth and in the coastal water of China.In contrast, clearer water from the South China Sea dominated the southern Taiwan Strait.
The distribution of POC in the study region is shown in Fig. 7. Generally, the POC concentration was higher in the western than in the eastern Taiwan Strait.It is evident that the POC concentration increases in areas under the influence of riverine inputs.No POC data were measured south of 23°N, but POC concentration ranging between 8.8 and 16.3 μm were observed by Kao et al. (2006)

DISCUSSION
Due to the highly particle-reactive characteristics, 234 Th is rapidly removed from seawater to the extent that a large deficiency from secular equilibrium with 238 U, especially in the shallow shelf region of the Taiwan Strait.The 238 U activity in the coastal water off western Taiwan follows the salinity-238 U relationship of Ku et al. (1977), 238 U (dpm L -1 ) = 0.07081 × Salinity (unpublished results).Accordingly, this relationship was used to calculate 238 U activity from the salinity data.The ratio of total (dissolved + particulate) 234 Th to 238 U at most of our sampling stations were lower than 0.4, indicating that removal of 234 Th was faster than supply by mixing of different water mass in the Taiwan Strait.
Where, U i = 238 U activity in the i-th box (dpm m -3 ); Thd ).
In the model, the particulate phase is assumed to be transported by current because fine particles remain in suspension in this dynamic system of strong advection.The residence times of dissolved (τ d ) and particulate (τ p ) 234 Th with respect to scavenging and particle removal, respectively, can be calculated by: J Th  The results of the model calculation on dissolved and particulate 234 Th for the five areas are shown in Table 2. Since both ADCP and current meter data collected from the Luzon Strait showed a persistent and strong flow of Kuroshio Branch Water through the northern Luzon Strait into the region off southwestern Taiwan (Liang et al. 2003), we used 1.5 and 0.25 dpm L -1 , respectively, as the 234 Th d and 234 Th p of influx water into Box I (Wei and Hung 1992).The average 234 Th d and 234 Th p activities of the influx water to Box II were 1.0 ± 0.2 and 0.2 ± 0.04 dpm L -1 , respectively, which were measured at the station (118°00'E, 22°18'N) in the northern South China Sea (unpublished results).
Depending on the average current velocity and the differences of 234 Th d and 234 Th p between adjacent boxes, the contributions of physical transport to the 234 Th d and 234 Th p mass balance vary from -40 to 643 dpm m -3 d -1 for 234 Th d and from -128 to 86 dpm m -3 d -1 for 234 Th p .Positive values indicate that more 234 Th was imported than exported from the box, whereas negative values indicate the amount of 234 Th imported to the box was less than the amount exported.The percentage of advection to the P value ranges from 21% in Box II to 36% in Box IV.Since the advection term in Eq. ( 2) was significantly smaller than the particle removal term, it is reasonable to conclude that vertical processes for 234 Th removal were more important than horizontal transport.Not surprisingly, physical transport causes the largest effect on the 234 Th budget in the central part of the Taiwan Strait (Boxes III and IV) because a large amount of dissolved 234 Th was  carried into the Taiwan Strait by the strong current through the Peng-Hu Channel (Fig. 1).Note that the advection term for 234 Th d in Box III exceeds the in situ 234 Th production from 238 U decay (~70 dpm m -3 d -1 ) by a factor of 9.
Although the advection terms were significant in each box, they were generally smaller than the scavenging [J in Eq. (1)] and particle removal [P in Eq. ( 2)] terms.Since only 234 Th d and 234 Th p in the surface water were determined, removal rates in dpm m -3 d -1 rather than fluxes in dpm m -2 d -1 , which requires knowledge of vertical profiles of 234 Th, are reported here.Relating to the biochemical settings of each domain, the J ranges from 32 dpm m -3 d -1 in Box I to 703 dpm m -3 d -1 in Box III and the P ranges from 24 dpm m -3 d -1 in Box II to 560 in Box III.
Box I represents the only domain deeper than 200 m in the Taiwan Strait and can illustrate the hydrographic characteristics of Kuroshio Branch Current.A scavenging rate of 32 dpm m -3 d -1 and particle removal rate of 24 dpm m -3 d -1 were obtained from the model calculations, which result in residence times of 50 and 11 days for dissolved 234 Th and particulate 234 Th, respectively.The model results imply that 234 Th cycling in this region is limited by the scavenging rate and, once scavenged onto particles, 234 Th is removed vertically from the surface water in a short period of time.
Box II encloses the Taiwan Bank, a shallow area of only 20 m average depth.An eddy covering the Taiwan Bank (Fig. 1) may enhance the suspension of particles.The 234 Th d and 234 Th p data in Box II also indicated relatively homogeneous distribution, supporting enhanced mixing due to the eddy.In contrast to Box I, the J term is larger than the P term in Box II, implying that particle settling was inhibited by strong tidal forcing.Thus our data suggest a short particulate 234 Th residence time of only 1 day.Both the highest scavenging and removal rates derived from the model were found in Box III.As shown in the 234 Th d contour map, northward transport of South China Sea water brings in large amount of 234 Th through the Peng-Hu Channel.The role of Cho-Sui River can be seen from the distribution of 234 Th p , which was dramatically elevated around 23 ~ 24°N in the eastern Taiwan Strait.Terrestrial materials brought in by the Cho-Sui River scavenge thorium from seawater to cause the distinct patch of elevated 234 Th p in the map.Among many rivers of the western Taiwan, the Cho-Sui River is the largest source of fluvial sediments to the Taiwan Strait.The annual sediment input of Cho-Sui River is 64 Mt, accounting for 70% of total sediment transportation by western Taiwanese rivers (Dadson et al. 2003).This large fluvial input serves as an efficient interceptor for particle-reactive elements transported from the south.The high P value found in Box III is conceivable based on the fact that fast settling coarse sediments from the Cho-Sui River quickly remove 234 Th from the water column.Residence times of dissolved and particulate 234 Th with respect to scavenging and particle removal are short, one day.Box V covers the region affected by Ming River input, in which both 234 Th d and 234 Th p are low (Figs. 8 and 9).Among the five boxes, the scavenging rate in this domain is comparable with that in Box I, 53 dpm m -3 d -1 , and the particle removal rate is relatively low, 64 dpm m -3 d -1 .The relatively low J is attributed to the high concentration of dissolved organic matter in the region.Hung et al. (2000) observed a high concentration (80 ~ 119 μm) of DOC, of which ~27% is in the colloidal form (> 1 kDa), in the region near the Ming River.High dissolved organic materials provide more complexing ligands for retaining thorium in the dissolved phase (Santschi et al. 2003).
Here we use 234 Th as a proxy of organic carbon to estimate the removal rate of POC via particle settling in the Taiwan Strait.Removal rates of POC (F POC ) were calculated by P i x(POC/ 234 Th) i , where (POC/ 234 Th) i is the ratio of particulate organic carbon to particulate 234 Th in suspended particles in box i.Before we proceed, some reflection of this application may be in order.Although 234 Th has been extensively used as a proxy of POC, large uncertainty arising from the variability of the POC/ 234 Th ratio in the oceans has been found.Buesseler et al. (2006) discussed the factors that may cause variability of POC/ 234 Th ratios in particles, including size, composition, shape, morphology, and sinking velocity of particles.Different sampling techniques have been used to collect marine particles for the determination of the POC/ 234 Th ratio.Many used largevolume in situ pump and chose the POC/ 234 Th ratio in > 53 μm particles as representative value for POC export calculation (Buesseler et al. 1995;Bacon et al. 1996;Hung et al. 2004), while others, including those involved in the JGOFS EqPac program (e.g., Murray et al. 1996), considered the ratio determined from the sinking particles directly collected by sediment traps at the euphotic depth to be the most suitable POC/ 234 Th ratio for 234 Th proxy approach of POC export.The POC/ 234 Th ratio in sinking particles was not available in this study; however, our measurements in the northern South China Sea and the Kuroshio showed that the POC/ 234 Th ratio of 0.45 -10 μm particles (15 ~ 20 μmol dpm -1 ) is fairly similar to that in sinking particles (10 ~ 15 μmol dpm -1 ) collected by floating traps.On the other hand, the POC/ 234 Th ratio in suspended particles of larger sizes (10 -63, 63 -153, > 153 μm) is 3 ~ 8 fold higher than that in sinking particles.The similar POC/ 234 Th ratio between fine (0.45 -10 μm), suspended particles and sinking particles indicates little fractionation of carbon and 234 Th during particle aggregation processes.Though we are aware of the caveats of utilizing 234 Th as POC proxy for estimating export, we argue that the results collected using conventional filtration of seawater through 0.45 μm filters can provide a first-order approximation of scavenging process in the Taiwan Strait.
The POC/ 234 Th i ratios in the Taiwan Strait do exhibit large variations, ranging between 8 to 154 μmol dpm -1 (Fig. 11).Benitez-Nelson et al. (2000) also observed a large range of POC/ 234 Th ratio, from 2 to 534 μmol dpm -1 , with most values > 10 μmol dpm -1 , in filtered particles collected from the surface water of the Gulf of Maine.Variation in biogeochemical settings may be responsible for the high variability of POC/ 234 Th ratio in the Taiwan Strait.Specific concentration of organic carbon in the suspended particles shows higher abundance of organic-rich particles along the China coast, which may be caused by enhanced growth of biological particles induced from terrestrial input of nutrients.The biomass in the domain covered by Box V is dominated by diatoms (Tung 2007).The highest POC/ 234 Th in the northwestern Taiwan Strait can reach values as high as 154 μmol dpm -1 , implying that planktons grown in the region are larger in size.These carbon-enriched meso-to macroplanktons, with smaller surface to volume ratios, provide lower surface areas for thorium scavenging.In contrast, POC/ 234 Th is much lower in the southern Taiwan Strait, suggesting that plankton residing in the South China Sea seawater are smaller in size due to the oligotrophic nature of the water mass.Tung (2007) found the dominant plankton in this region is cyanobacteria.Our results showed that the lowest POC/ 234 Th was found in Box III, the region under the influence of the Cho-Sui River, with large amounts of detrital material containing less organic carbon and hence low in POC/ 234 Th.
The average POC/ 234 Th ratio, the POC removal rate and the residence time of POC (τ POC ) in the five areas of the Taiwan Strait are shown in Table 2. Errors (±1σ) were propagated from uncertainties associated with measured and estimated parameters.The removal rate of POC from the surface water of the Taiwan Strait ranges from 0.3 ~ 10.2 mmol-C m -3 d -1 .Generally, the POC removal rate in the southern Taiwan Strait is lower than that in the northern Taiwan Strait.The Relatively low POC removal rate in the Taiwan Bank, 2.2 mmol-C m -3 d -1 , may be a result of resuspension in this shallow region with an average water depth of 20 m.In comparison with the primary productivity of 1.1 ~ 2.1 mmol-C m -3 d -1 in this region (Huang et al. 1999), most POC produced in the surface is exported to the deep layer.With respect to the removal rate, the residence time of POC in the Taiwan Strait ranges from 1 day in Boxes III and IV to 15 days in Box I.The POC residence time with respect to particle removal in Box V is 3 ± 3 days.This estimate of POC residence time is comparable with Hung et al. (2000), who found a residence time of 6 days for POC estimated from the ratio of the POC inventory in the euphotic layer and the rate of new production in the inner shelf of the southern East China Sea, the region immediately to the north of our Box V.

CONCLUSIONS
From the distribution of dissolved and particulate 234 Th we made a quantitative estimate of the scavenging and particle removal rates in the surface water of the Taiwan Strait.Multiplying the POC/ 234 Th ratio by the particle removal rate estimated from 234 Th/ 238 U disequilibrium, the removal rate of particulate organic carbon from the surface water of the Taiwan Strait was also estimated.Based on the environmental settings (hydrography and topography), the Taiwan Strait was divided into five regimes for our model manipulation.The calculated results are summarized as follows: (1) Occupied by the Kuroshio Branch Water, region 1 has higher temperature and salinity.This region has the highest scavenging and removal rates of 234 Th, with τ d = 50 and τ p = 11 days, similar to open ocean values.
(2) Overlying the Taiwan Bank, region 2 has the lowest scavenging and removal rates due to resuspension of bottom sediments in this shallow region.The removal rate of POC from the surface water in this region is therefore relatively low, 2.2 ± 0.8 mmol-C m -3 d -1 .(3) Covering the Cho-Sui River's mouth, region 3 has extremely high particle concentration.Therefore, 234 Th is scavenged and removed at a very fast rate, with τ d and τ p of one day only.(4) In the middle of Taiwan Strait, region 4 has relatively high scavenging and particle removal rates, hence a short τ d and τ p of ≤ 1 day.(5) Encompassing the Ming River's mouth, region 5 is characterized by low temperature, low salinity and high chlorophyll-a.In this region both τ d and τ p are ~3 days.

Fig. 1 .
Fig. 1.Bathymetric map with large-volume sampling stations of the Taiwan Strait.Mean surface current calculated from shipboard May ADCP data collected during 1991 ~ 2006 are shown as vectors.

Fig. 2 .
Fig. 2. Distribution of temperature in surface water of the Taiwan Strait.

Fig. 4 .
Fig. 4. Distribution of relative fluorescence unit (RFU) in surface water of the Taiwan Strait.RFU was measured by fluorometer attached to the underway system during the cruises.
in the surface water off southwestern Taiwan, indicating the POC concentrations in the southern part of the Taiwan Strait may have been similar to that in the eastern side of the strait.Contours for activities of dissolved ( 234 Th d ) and particulate ( 234 Th p ) 234 Th are shown in Figs. 8 and 9, respectively.The spatial distribution of 234 Th d corroborates the circulation pattern in the Taiwan Strait.A relatively high 234 Th d is associated with the KBC and extends northward through the Peng-Hu Channel.The 234 Th d in the shallow shelf of the study area showed very low values (< 0.5 dpm L -1 ).Unlike the 234 Th d distribution, the 234 Th p showed low values in the regions of the Ming River mouth and off southwestern Taiwan.In contrast, a patch of high 234 Th p water was indicated in the vicinity of the Cho-Sui River mouth.

Fig. 6 .
Fig. 6.Distribution of total suspended matter (TSM) concentration in surface water of the Taiwan Strait.

Fig. 8 .
Fig. 8. Distribution of dissolved 234 Th in surface water of the Taiwan Strait.

Fig. 7 .
Fig. 7. Distribution of particulate organic carbon in surface water of the Taiwan Strait.
234 i = dissolved 234 Th activity in the i-th box (dpm m -3 ); Thp 234 i = particulate 234 Th activity in the i-th box (dpm m -3 ); Vi = average current velocity in the i-th box (m d -1 ); ΔL = distance between the centers of consecutive boxes (m); λ = radioactive decay constant of 234 Th (d -1 ); Ji = net change rate of dissolved 234 Th due to scavenging process in the i-th box (dpm m -3 d -1 ); and Pi = net change rate of particulate 234 Th due to particle removal process in the i-th box (dpm m -3 d -1 environmental settings and the spatial distribution of 234 Th d and 234 Th p , the study area was divided into five boxes (Fig.10) and treated as an advection-scavengingremoval system.Following the procedures ofLiang et al. (2003), the mean current velocity in May at the boundary of each box was estimated by averaging archived shipboard ADCP data (NCOR Data Bank) collected during 1991 ~ 2006.In each box, the number of ADCP data sets for averaging current velocity ranges from 2400 to 3115, which gives a root mean square (rms) error of ~5 cm s -1(Liang et al. 2003).

Fig. 9 .
Fig. 9. Distribution of particulate 234 Th in surface water of the Taiwan Strait.

Fig. 10 .
Fig. 10.Five regions in the Taiwan Strait delineated for the model calculation.

Fig. 11 .
Fig. 11.Distribution of the ratio of particulate organic carbon to particulate 234 Th (POC/ 234 Th) in surface water of the Taiwan Strait.

Table 1 .
Temperature, salinity, concentration of total suspended matter (TSM), dissolved ( 234 Th d ) and particulate ( 234 Th p )234Th activities, and particulate organic carbon concentration obtained from the surface water of the Taiwan Strait.Standard deviations of 234 Th d and 234 Th p are based on propagated counting error (1σ).