December 1994 Material Exchange Between the East China Sea and the Kuroshio Current

The water exchange rate between the continental shelf water of the East China Sea and the Kuroshio Current waters is estimated to be about 22000±9000 km3/yr or 0.7±0.3 Sverdrups which is about 25 times the value of the major river runoffs in the region. The implication is that the chemistry of the continental shelf water is mostly influenced by the Kuroshio water in­ puts. Inputs of dissolved particle-reactive trace elements into the continental shelf water from the major rivers and the upwelled subsurface Kuroshio water are efficiently sequestered by suspended particles and subsequently removed to the bottom sediments. The Okinawa Trough sediments are important sites for receiving the dispersed fine suspended particles and associated particle­ reactive trace elements from the continental shelf. (

of the major river runoffs in the region.The implication is that the chemistry of the continental shelf water is mostly influenced by the Kuroshio water in puts.Inputs of dissolved particle-reactive trace elements into the continental shelf water from the major rivers and the upwelled subsurface Kuroshio water are efficiently sequestered by suspended particles and subsequently removed to the bottom sediments.The Okinawa Trough sediments are important sites for receiving the dispersed fine suspended particles and associated particle reactive trace elements from the continental shelf.
(Key words: Coastal exchange, East China Sea, Kuroshio Current) The continental shelf of Yellow Sea and East China Sea has a total area of about 0.9 x 10 6 km 2 (water depth of less than 100 fathoms) and is one of the largest continental shelves in the world.The East China Sea is bounded by the Kuroshio Current in the east.Two of the world's major rivers i.e.Yangtze (Changjiang) River and Yellow (Huanghe) River also discharge into the East China Sea and the Pohai Gulf of the Yellow Sea respectively.
The mean water and suspended sediment discharge rates are respectively, 850km 3 /yr and 490 • X 10 6 tons/yr for Yangtze River, and 35 km 3 /yr and 900 x 10 6 tons/yr for Yellow River (Milliman and Meade, 1983).The average compositions of river water and suspended particles from these two rivers are summarized in Table 1.How much of river inputs have ended up in the continental shelf and the Okinawa Trough sediments, and how much have been exported out to the open ocean need to be addressed.
The general mixing patterns between the surface waters of the Kuroshio Current and the continental shelf during summer and winter seasons are given by Niino and Emery (1971).The intrusion of high temperature, high salinity water of Kuroshio Current along   where Sr, Sk and S8 are respectively salinity of river water, Kuroshio water, and the conti nental shelf water.The next question is what is the representative S8 of the shelf water which exchange with the Kuroshio waters?The salinity transects across the East China Sea (Chen and Bychk o v, 1992) indicate that th e salinity of 33±0.2 per mil is a reasonable boundary to separate between the continental shelf water with steep salinity gradient and the Kuroshio waters with low salinity gradient.Therefore, 33±0.2 per mil is chosen here for S 8• Similarly, the main Kuroshio waters which exchange with the shelf water are the Kuroshio Surface Water with typical salinity of 34.4 per mil, and the Upwelled Subsurface Kuroshio Water with typical salinity of 34.6 per mil.Therefore, Sk of 34.5±0.1 per mil is chosen here.Whereas, Sr is about 0.2 per mil.Substituting these numbers into the above equation, one obtains Om to be about 22000±9000 km 3 /yr or 0.7±0.3Sverdmps, which is about 25 times the value of Qr.Interestingly, the mixing rate between the shelf water and the Tsushima Current water was estimated to be about 0.6±0.2Sverdrups, based on 228 Ra! 226 Ra mass balance (Nozaki, 1989).The estimated volume of the shelf water (V 8 ) with salinity less than 33±0.2 per mil is about 2 to 3 10 4 km 3 • Therefore, the mean residence time of the dissolved non-reactive salts in the shelf water is about 1 year ( = V 8 /Q8 ), which is relatively long compared to that of the New York Bight water (3 to 4 months; Li et al. 1979).Qm of about 25• Q r implies that the chemistry of the continental shelf water is mostly influenced by the Kuroshio water inputs except near the river mouths or when the concentration of a given dissolved element in the rivers is more than 25 times of that in the Kuroshio water.Also the exchanged Kuroshio waters are most likely dominated by the Upwelled Subsurface Kuroshio Water and to a lesser extent by the Surface Kuroshio Water (Chen and Bychkov, 1992; Gong, 1992; Lin et al.,  1992).
The net material exchange between the continental shelf water and the Kuroshio water, X, can be approximated by X = Om (C8 -Ck), where C8 and Ck are respectively the concentrations of a given material in the continental shelf and the Kuroshio waters.For example, the average nitrate concentration of the shelf water of the East China Sea is about 1 µM and of the Upwelled Subsurface Kuroshio Water about 15 µM (Gong, 1992;Liu et al., 1992), therefore; the net flux of nitrate from the Kuroshio water to the shelf water is at most about 310 x 10 9 moles N/yr [=22 x 10 15 l/yr•(l5-1) µMoles Nil].For comparison, the river nitrate flux is only 26 x 10 9 mole N/yr.These nitrate can support an average new production of about 80 mg C/m 2 /d over the continental shelf of the East China Sea {assuming Redf i eld ratio of C/N = 106/16).At a steady state, this new production should also be removed from the surface shelf water back to the Kuroshio waters and/or bottom sediments.If the average total primary production by phytoplankton over the East China Sea continental shelf is about 500 mg C/m 2 /d or greater (Koblenz-Mishke et al., 1970), then the f-factor (=new production/total primary production) would be about 0.16 or less.The f-factor of 0.66 given by Gong (1994) is probably an overestimate.Certainly, new measurements of the primary production and independent ctstimate of the f-factor in the East China Sea are badly needed.
According to_ McKee et al., (1984), the residence time of 234 Th in the water column with respect to its removal to the seabed ranges from a half day near the Yangtze River mouth to about 11 days in midshelf water.This range is much smaller than those for coastal environments without a major sediment source such as New York Bight (about 10 to 30 days; Kaufman et al. 1981).This range of 0.5 to 11 days is also much shorter than the residence time of dissolved non-reactive salts in the continental shelf water (about one year).Therefore, the inputs of dissolved particle-reactive elerr.. ents into the continental shelf water from the Yangtze River, Yellow River and the Upwelled Subsurface Kuroshio Water should be efficiently sequestered by suspended particles and removed to the bottom sediments.For comparis 9 n, the residence time of 234 Th in the surface 100 meters of the Kuroshio Current, northeast_ of Taiwan, is about 70 days (Wei, 1991), which is similar to that of the Gulf Stream surface water in the New York Bight (Kaufman et al. 1981, Li et al. 1979).
Once the particle-reactive elements in the water column are sequestered by suspended particles, they end up mostly in the fine-grained bottom sediments in the region.The distribu tion of fine-grained mud patches on the continental shelf of the East China Sea and the Yellow Sea is closely related to the cyclonic gyre systems in the area (Hu, 1984).Usually a cyclonic gyre induces upwelling within the gyre and accumulates fine-grained sediments underneath.In contrast, an anticyclonic gyre induces downwelling within the gyre and flashes out fine sediment particles (Hu, 1984 ).Another important mud patch is on the lower continental slope and the Okinawa Trough.The mud patch sediments are also high in organic carbon content (0.5 to 1 %; Niino and Emery, 1971).From the 210 Pb profiles in the sediments, the typical sediment accumulation rate for the inner shelf mud patch next to Yangtze River mouth is estimated to be about 2.1 g/cm 2 yr (or 3 cm/yr for density of 0.7g/cm 3 ), and for the outershelf mud patch about 0.15 g/cm 2 yr (DeMaster et al. 1985).For comparison, if the suspended particle inputs from the Yangtze and Yellow Rivers are evenly distributed on the whole conti nental shelf, the average sediment accumulation would be about 0.15 g/cm2 yr.The sediment accumulation rates in the southern end of the Okinawa Trough range from 0.5 g/cm2yr at depth of 900m to 0.08 g/cm2yr at 1600 m, whereas there are no sediment accumulation at the depth interval between 400 to 800 meters (Y.Chung, personal communication).
Assuming an average sedimentation rate of 0.3g/cm2yr, the total flux of sediments in the Okinawa Trough (total area of about 0.09 x 106km2 for depth greater than 900m) would be about 300 x 106tons/yr.If one half of these sediments were ultimately originated from the Yellow and Yangtze rivers and other half from local sources such as Ryukyu islands, then about 10% of the suspended particle loadings from the major rivers may have been depositing in the Okinawa Trough.Additional data on the accumulation rate and compositions of sediments in other areas of the Okinawa Trough are needed to improve this kind of estimation.
As shown by DeMaster el al. (1985), the high fl uxes of excess 234Th to the inner-shelf mud patch relative'to the predicted fluxes from the water column production (2.2 to 27.6 times) strongly suggest lateral transport and deposition of 234Th from offshore waters to near-shore on the time scale of about 100 days (about 4 half lives of 234Th).In contrast the reiative fl uxes (the observed over the predicted from water column production plus atmospheric inputs) of excess 210Pb in the same mud �atch are only 0.2 to 2.6 times on the time scale of about 100 years (about 4 half lives of 21 Pb).The implication is that any particle reactive elements in the continental shelf waters would be rapidly sequestered by particles and deposited in the inner-shelf mud patch on a 100 days time scale, but subsequently be dispersed out from the area on a in the Okinawa Trough sediments (900 to 1600m) northeast of Taiwan are respectively 4 to 10, and 10 to 50 times of those predicted from water column production plus atmospheric inputs for 210Pb (Y Chung, personal communication).The average content of Mn in the Okinawa Trough sediments is about 0.3% (Table 1) and the highest value can be as high as 0.7% (Zhao e� al. 1984), as compared to about 0.1% for the river suspended particles (Table 1 ).It is often suggested that manganese oxide particles or coatings are the most efficient scavengers for 210Pb, 230Th and other trace metals in the oceans (Bacon et al. 1976;Li, 1982).It wiil be instructive to see how effective the Okinawa Trough sediments also accumulate anthropogenic pollutants such as 239•240Pu, 137 Cs, pesticides, trace metals etc. ' . The low temperature, low salinity waters along the mainland coast are southward fl owing Yellow Sea Coastal Current and Zhejiang Coastal Current.The hydrographic transects across the continental shelf clearly show the intrusion of the Upwelled Subsurface Kuroshio (OSK) water across the shelf break onto the continental shelf and underneath the less saline continental shelf water(Chen and Bychkoz, 1992;Gong, 1992;Liu et al. 1992).As shown schematically in Figurel, the water budget for the continental shelf water is controlled by river inputs (Qr), evaporation (Qv) and precipitation (Qp) over the continental shelf, Kuroshio waters mixing into the shelf (Qm), and the output of the shelf water to the Kuroshio waters (Q8).By water mass balance, Q8 should be equal to Qm+Qr+Q p-Q v at a steady state.According toOberhuber's (1988) compilation , the net precipitation over evaporation in the area is about 17 cm/yr or Qp-Q� = 140±10 km 3 /yr as compared to the river input, Qr, of 885 km 3 /yr.

Fig. 1 .
Fig. 1.The schematic diagram for the continental shelf surface water budget on the East China and Yellow Seas, where Qr = runoff inputs from Yangtze and Yellow Rivers, Qp = precipitation inp u t, Qv= evaporation output, Qm= mixing rate of the Kuroshio waters into the continental shelf and Q8= output rate of the shelf water to the Kuroshio waters.Assuming a steady state, the salt mass balance in the continental shelf water requires: Or• Sr +O m • S k = (O m +Or+Op-Ov )SB or Qm=(Qr+Qp-Qv)•S8/(Sk-S8)-0r•Srl(Sk-Ss) 100 years time scale (DeMaster et al. 1985).Because the observed fluxes of both excess 234Th and excess 210Pb in the mid-shelf sandy sediments and in the outer-shelf mud patch are all below the predicted fluxes from water column production plus atmosphere inputs (DeMaster et al. 1985), the continental slope and Okinawa Trough sediments should be important sites for receiving the dispersed fine particles and associated particle reactive elements from the continental shelf.Indeed, the fluxes of excess 210Pb and excess 230Th

Table 1 .
Av erage compositions of waters and suspended particles from the Yangtze and Yellow Rivers, as well as the upwelled Kuroshio sub surface water and average Okinawa Trough sediments.and west coasts of Korean Peninsula are respectively the Tsushima Current and the Yellow Sea Wann Current.Another wann saline water intrusion northeast of Taiwan Island is the Taiwan Warm Current -the south