A Study of Transport and Impact Strength of Fukushima Nuclear Pollutants in the North Pacific Surface

FU Hongli, LI Wei, ZHANG Xuefeng, HAN Guijun, WANG Xidong, WU Xinrong, and ZHANG Lianxin

Key Laboratory of Marine Environmental Information Technology, State Oceanic Administration, National Marine Data and Information Service, Tianjin 300171, P. R. China

A Study of Transport and Impact Strength of Fukushima Nuclear Pollutants in the North Pacific Surface

FU Hongli, LI Wei, ZHANG Xuefeng, HAN Guijun*, WANG Xidong, WU Xinrong, and ZHANG Lianxin

Key Laboratory of Marine Environmental Information Technology, State Oceanic Administration, National Marine Data and Information Service, Tianjin 300171, P. R. China

Based on the statistics of surface drifter data of 1979–2011 and the simulation of nuclear pollutant particulate movements simulated using high quality ocean reanalysis surface current dataset, the transport pathways and impact strength of Fukushima nuclear pollutants in the North Pacific have been estimated. The particulates are used to increase the sampling size and enhance the representativeness of statistical results. The trajectories of the drifters and particulates are first examined to identify typical drifting pathways. The results show that there are three types of transport paths for nuclear pollutants at the surface: 1) most pollutant particles move eastward and are carried by the Kuroshio and Kuroshio-extension currents and reach the east side of the North Pacific after about 3.2–3.9 years; 2) some particles travel with the subtropical circulation branch and reach the east coast of China after about 1.6 years according to one drifter trajectory and about 3.6 years according to particulate trajectories; 3) a little of them travel with local, small scale circulations and reach the east coast of China after about 1.3–1.8 years. Based on the particulates, the impact strength of nuclear pollutants at these time scales can be estimated according to the temporal variations of relative concentration combined with the radioactive decay rate. For example, Cesium-137, carried by the strong North Pacific current, mainly accumulates in the eastern North Pacific and its impact strength is 4% of the initial level at the originating Fukushima area after 4 years. Due to local eddies, Cesium-137 in the western North Pacific is 1% of the initial pollutant level after 1.5 years and continuously increases to 3% after 4 years. The vertical movement of radioactive pollutants is not taken into account in the present study, and the estimation accuracy would be improved by considering three-dimensional flows.

Fukushima nuclear pollution; ensemble estimation; surface drifting buoy; ocean reanalysis

1 Introduction

On March 11, 2011, the massive Tohoku earthquake happened (9 degrees on the Richter Scale) on the east coast of Japan in the Pacific Ocean (epicenter 38.32°N, 142.37°E), and the epicenter was 130 km from Sendai (capital city of Miyagi Prefecture, Japan). The earthquake caused a huge tsunami hiting the east coast of Japan.

There are 55 nuclear reactors operating in Japan, the majority of which are located close to the coast. This earthquake and tsunami caused extensive damage to the nuclear plant of Fukushima, and resulted in a great amount of radioactive matter leaking into the sea. The incident has been listed as the biggest ever release of radioactive materials into the oceans (Quirin, 2011). High levels of radioactive iodine-131 (with a half-life of about 8 d), cesium-137 (with a half-life of about 30 years), and cesium-134 (with a half-life of about 2 years) have been been measured in waters adjacent to Fukushima. The highest concentrations, found closest to the coast, were about 38 Bq L−1for iodine-131 and 4.5 Bq L−1for cesium-137. Concerns have arisen about the potential effects of the released radiation on the marine environment and resources.

The transport of nuclear pollutants is mainly controlled by surface ocean current. Generally speaking, the most prominent feature of the surface ocean currents in the North Pacific is the strong subtropical gyre (Fig.1) (Tomczak and Godfrey, 2001), consisting of the North Equatorial Current, the Philippines Current, the Kuroshio, the North Pacific Current, and the California Current. These currents have the potential of bringing radioactive materials from the Fukushima nuclear accident site to different areas of the Pacific Ocean. Since the current speed is small, no radioactive materials have yet been detected along the coast of circum-Pacific countries. However, due to the long half-life of some radioactive isotopes, radioactive contaminants could remain in the ocean for years (Buck et al., 2011).

Li et al. (2011) analyzed the drifting pathways of Fukushima Nuclear Leakage Pollutant based on Argo data.Tsumune et al. (2011) then used a regional ocean model to simulate the cesium-137 concentrations resulting from the direct release to the ocean off Fukushima and found that 1) the released cesium-137 advected southward along the coast during the simulation period; 2) the eastwardflowing Kuroshio and its extension transported cesium-137 during May 2011; and 3) as a result of oceanic advection and diffusion the cesium-137 concentrations decreased to less than 10 Bq L−1in the simulation domain by the end of May 2011. The spread of nuclear radiation from the power plant through the atmosphere and ocean was predicted by a climate forecasting model and an ocean circulation model under some idealized assumptions; the result indicated that the nuclear material in the ocean would be slowly transported northeast of Fukushima and reach 150°E in 50 d, and the nuclear debris in the ocean would be confined to a narrow band (Qiao et al., 2011).

Fig.1 Surface currents of the North Pacific Ocean. Abbreviations stand for Philippines Current (PC), North Pacific Current (NPC) (Tomczak and Godfrey, 2001).

This study focuses on the impact of Fukushima nuclear pollutants on different areas of the North Pacific through surface ocean currents from a viewpoint of ensemble estimation. In Section 2, the surface drifters passing by the area adjacent to the Fukushima nuclear pollutant source are identified and the CORA (China Ocean Reanalysis) current dataset is used to calculate the trajectories of the nuclear pollutant particulates. Section 3 gives the methods of tracing the nuclear pollutant particulates and estimates their impact strength at different timescales. In Section 4, the trajectory types of all surface drifters and simulated nuclear pollutant particulates are statistically examined, and the impact strength of the nuclear pollutants at different timescales is estimated. Discussion and conclusions are given in Section 5.

2 Data

In the study, the radioactive pollutant in the ocean is treated as a mixture of multiple Lagrangian particulates, and each particulate represents a radioactive element. The particulates can move in both horizontal and vertical directions, but cannot diffuse and mix with surrounding sea water (Liu and Kuai, 2011). Since vertical motion in the ocean is mostly very small compared to horizontal advection and very difficult to measure and simulate, the vertical movement of radioactive pollutants is not taken into account in the present study, and the Lagrangian particulates are assumed to drift at the ocean surface. Based on the surface drifters and nuclear pollutant particulate simulations, the drifting trajectories and time scales of transport of the radioactive pollutant particulates are estimated.

2.1 Surface Drifter Data

Fig.2 The pink semi-circle area with a radius of 300 km designates the source region of Fukushima (star) nuclear pollutants. The black lines and red dots represent the trajectories and terminal positions of the surface drifters originated from the source region with the lives of less than 1 year (a), 1–2 years (b), 2–3 years (c), more than 3 years (d), respectively.

The location information of surface drifters can be accessed from the website of the Marine Environment Data Services (MEDS) in Canada. According to the public news reports on Fukushima nuclear pollutants during March 12–15, 2011, a small semi-circle region centered at Fukushima with a 300 km radius is used as the source region of nuclear pollutants (the pink area of Fig.2). The spatial scale of the pollutant source is consistent with thescale of rapid mixing related to the confluence of the Oyashio and Kuroshio waters in this region (Shimizu et al., 2001). Here all surface drifters of 1979–2011, which passed the source region (176 buoys, Fig.2), are used to analyze possible transport paths of nuclear pollutants and estimate the transport time scales for the pollutants to reach the easternmost and westernmost coast areas of the North Pacific. Fig.2 shows initial positions (pink area), terminal positions (red points), and trajectories (black line) of surface drifters with the drifting lives of less than 1 year (a), 1–2 years (b), 2–3 years, and greater than 3 years, respectively.

As shown in Fig.2, the lives of all selected surface drifters are completely different, and most of them cannot be used to estimate the long-term transport paths of nuclear pollutants because of instrument malfunction within one year of tracking. As a result, the number of surface drifters actually used is much less than 176. Due to the limitation of sample size (the number of surface drifters), the results from surface drifters only represents a statistical trend at low significance levels. In addition, the selected surface drifters not only passed the source region of nuclear pollutants in March but also in other months, therefore the estimated drifting paths and time scales of the nuclear pollutants transport represent annual mean results.

2.2 Ocean Reanalysis Data

A global ocean reanalysis system, parallel to the regional ocean reanalysis system for the coastal waters of China and adjacent seas under the project of CORA (http://www.cora.net.cn; Han et al., 2011), was developed by the Key Laboratory of Marine Environmental Information Technology (MEIT) of China National Marine Data and Information Service (NMDIS). The system domain is from 0° to 360° in longitude and from 75.25°S to 84.75°N in latitude. Global ocean reanalysis datasets (here after called CORA) of sea surface height (SSH), temperature, salinity, and current have been generated, which refer to daily-mean fields of ocean states with varying horizontal resolution from 0.5°(longitude)×0.5°(latitude) in high latitudes to 0.5°(longitude)×0.25°(latitude) near the equator for 35 vertical levels, and covers a period of 25 years from January 1985 to April 2009.

The ocean model employed in the global ocean reanalysis system is the Massachusetts Institute of Technology’s General Circulation (MITgcm) Model (Marshall et al., 1997). A sequential three-dimensional variational (3DVAR) analysis scheme has been implemented using a multi-grid framework (Li et al., 2008, 2010; Xie et al., 2011). Such a scheme can be performed in three dimensional space, which is different from the traditional 3DVAR performed on each model level by ignoring the vertical correlation (Derber and Rosati, 1989). This sequential 3DVAR analysis scheme can retrieve resolvable information from longer to shorter wavelengths for a given observation network and yield multi-scale, inhomogeneous analysis. The ocean model is forced by the wind field from the US National Centers of Environmental Prediction (NCEP) reanalysis from 1 January 1985 to 30 June 1987 and a new Cross-Calibrated, Multi-Platform (CCMP) ocean surface wind field from the Physical Oceanography Distributed Active Archive Center (PO.DAAC) from 1 July 1987 to 31 April 2009. The heat flux and fresh water flux are also taken from NCEP. River runoffs were not considered in this study. The ocean observations include sea surface temperature (SST) and sea surface height anomaly (SSHa) from satellites, and temperature and salinity profiles from the Argo and World Ocean Database 2009 (WOD09) maintained by the National Oceanographic Data Center (NODC). More details in the evaluation of CORA dataset can be found in Han et al.’s study (2013).

To accurately calculate the Lagrangian trajectory of nuclear pollutants, the CORA surface current dataset in the North Pacific was examined by using surface drifters and observational daily-mean current provided by TAO (Tropical Atmosphere-Ocean) Project Office of PMEL (Pacific Marine Environmental Laboratory). The averaged currents in the top 30 m of the water column were used to represent surface currents. The CORA surface currents were interpolated in space and time to properly compare with validated datasets of surface drifters and TAO, and statistics were obtained from these datasets.

The time-averaged zonal velocities from drifters and CORA agree at all latitudes in both longitudinal ranges as shown in Fig.3. Positive and negative zonal currents indicate eastward and westward flows, which correspond to the North Equatorial Counter Current and the North Equatorial Current at 0°–30°N, respectively. Within 30°–50°N, all zonal currents flow to the east under the influence of the Kuroshio, Kuroshio-extension Current and North Pacific Current. The difference of meridional currents between drifter and CORA data is modest but larger than that of zonal currents (Fig.3). The meridional current direction from the CORA dataset is well reproduced in general although the meridional current speed in the equator is underestimated.

Fig.3 The time-averaged zonal (a and b) and meridional (c and d) speeds (m s−1) of surface drifters (blue line) and CORA surface currents (green line). Data have been further averaged in 2° latitude by 50° longitude bins.

A moderate correlation (Fig.4) is found between the drifter and CORA zonal currents, peaking around the North Equatorial Counter Current at 5°–8°N and diminishing poleward. The equator is a nadir for the meridional current correlation, which increases to more than 0.3 at middle latitudes and decreases to less than 0.3 at high latitudes.

Fig.4 The correlations between drifters and CORA for zonal (a and b) and meridional speeds (c and d).

The reanalysis surface currents are also evaluated by comparing the trajectories of the surface drifters with those of the surface drifting particles. The surface drifting particles have the same initial positions as the surface drifters, and their trajectories can be obtained using the following equation:

where u(x, y, t) and v(x, y, t) denote the eastward and northward components of surface current at the point (x, y) and the time t, respectively. The discrete differential form of Eq. (1) is

where ‘i’ indicates the i-th drifting particle; ‘n’ indicates the n-th time-step; ‘dt’ is the time step and equal to the time interval in the CORA dataset, 7.5 min; tnis the time at the n-th time-step and equal to ndt+t0; t0, xi(t0) and yi(t0) indicate the initial time and position of the i-th drifting particle, respectively, which are the same as those of the corresponding surface drifters.

Fig.5 shows the observed (red line) and calculated (green line) Lagrangian trajectories of the surface drifters 1177777, 1177877, 1177977, 1299091, 1498177, 54688, 56699 and 63098, respectively. As can be seen, most of the observed and calculated trajectories are overlapped, demonstrating that the CORA surface currents are reconstructed well to represent the characteristics of surface currents in the middle latitudes of the North Pacific.

Fig.5 The observed (red line) and simulated (blue line) Lagrangian trajectories of surface drifters (a) 1177777, (b) 1177877, (c) 1177977, (d) 1299091, (e) 1498177, (f) 54688, (g) 56699, and (h) 63098 during different time periods. The black circles indicate the initial positions of the drifters.

Time series of TAO current data at near-equatorial locations are used to validate CORA surface currents. The comparison between TAO and CORA time-averaged currents (Table 1) shows that the surface current speed and direction of CORA are well reproduced except at a few locations where the current speed is poorly calculated. In the near-equatorial region, the correlation between TAO and CORA zonal currents is higher, but the correlation between meridional currents is lower than that in other regions.

Table 1 The time-averaged surface speeds and correlations between TAO and CORA for zonal and meridional components

The comparison of CORA surface currents based on TAO dataset and drifter data indicates that the present CORA data provides, to a certain degree, reliable estimates of zonal and meridional time-mean currents. The accuracy of zonal currents is higher than that of meridional currents, especially in the equatorial areas. This may be attributed to the small time and space scales of nearequatorial meridional currents, which are closely associated with tropical instability waves and poorly resolved by the present reanalysis system. The correlations between drifters and CORA for zonal or meridional speeds diminish at the polar region, due to the decrease of the Rossby deformation radius and inertial timescale, which cannot be resolved well by the present reanalysis system either. A reanalysis system with a finer resolution is needed for further improvement of the CORA dataset, especially at low or high latitudes. In this study, the CORA surface currents at middle latitudes are mainly used to calculate the trajectories of the nuclear pollutant particulates, therefore the poorly produced CORA currents at low or high latitudes hardly affect the study results.

3 Method

To increase the sampling size and enhance the statistical representativeness, the gridded CORA surface currents were used to simulate the trajectories of nuclear pollutant particulates. Initially, 5638 particles were randomly deployed in the defined source region of nuclear pollutants (the pink area in Fig.2) on March 11 of each year for a total of 25 years (1985–2009). A Lagrangian integration with a 7.5-minute time step was conducted to track these particles for a five-year period after they had been released. The Lagrangian integration method is similar to that used to calculate the trajectories of the drifters in Section 2. Because the initial times of these drifting particles are in the March of each year, their trajectories can used to estimate the transport and impact strength of nuclear pollutants in the March.

Fig.6 shows the Lagrange trajectories (blue line) and terminal positions (red points) of 5638 nuclear pollutant particles for one year (a), two years (b), three years (c), and five years (d) after they were released in the source region of nuclear pollutants (pink area) on March 11, 2004. The trajectories of these particles look similar to those of drifters in a short period of time, but the differences become noticeable for a long period of simulation. For example, the trajectories show that drifters have a tendency to move eastward, but particles move both eastward and southwestward after two years of release.

The great number of particles make it possible to estimate the spatial distribution and temporal variability of the impact strength of nuclear pollutants as a leaking incident happening on March 11 of each year from 1985 to 2009. For example, at any given time t after the particle releasing, the relative concentration of nuclear pollutants CR(i, j, t) can be calculated by

where Cois the initial number of particles per unit area in the originating region, and C(i, j, t) is the number of particles per unit area at point (i, j) and time t. The impact strength of the nuclear pollutants S(i, j, t) can be given by

where t1/2represents the half-life of the related radioactive element, and the CR(i, j, t) indicates the relative concentration of nuclear pollutants at point (i, j) and the t. Based on the Eq. (4), the impact strength of the nuclear pollutants can be calculated at any time after released. For the nuclear pollutant release on March 11 of each year from 1985 to 2009, the impact strength is associated with 25 determinants at any given time. The spatial distribution of the impact strength of the radioactive elements from the Fukushima accident can be estimated by the weighted average of the 25 determinants.

Fig.6 The Lagrangian trajectories (blue line) and terminal positions (red dots) of 5638 nuclear pollutant particles during (a) one year, (b) two years, (c) three years, and (d) five years tracking after released in the source region on March 11, 2004.

4 Results

Three typical trajectories of surface drifters are identified and classified, which reach the westernmost area of the North Pacific delineated by the longitudinal line of 125°E or the easternmost area of the North Pacific delineated by 130°W, respectively (Fig.7). 80% (141) of the 176 surface drifters that passed by the defined source region of nuclear pollutants were carried by the Kuroshio and Kuroshio-extension currents and the North Pacific Current and moved eastward, and 20% of the drifters moved southward (Fig.2). Among the 141 buoys moving eastward, except for signal loss of a few drifters after a short period of tracking, only 13 reached the easternmost area of North Pacific (Figs.7 and 8). Similarly, among the 35 buoys moving southward, a few stopped sending signals and could be tracked after a short period of time; only 5 buoys turned westward and eventually reached the east coast of China (Fig.8). Among the five drifters, one drifter was found to travel with the subtropical current branch at the central Pacific (Neumann, 1968) and had a relatively longer trajectory (Fig.8); the other four were transported to the east coast of China by local circulations associated with eddies (Fig.8).

Fig.9 shows the trajectories of the five buoys reaching the east coast of China. The red dashed line marks the boundary between the Pacific Ocean and the coastal region of China, and the five blue dots indicate the initial positions as the five drifters entered the east coast area of China.

Fig.7 Pathways of surface drifters and nuclear pollutant particles that originated from the source region (pink area) and moved eastward (blue), southwest (red), and first eastward and then westward (green). The numbers outside (inside) of parentheses indicate drifters (nuclear pollutant particles) that reached the westernmost (west of the dashed-red line) area and the easternmost (east of dashed-yellow line) area of the North Pacific.

Fig.8 The observed trajectories of surface drifters that originated from the source region (pink area) and reached the easternmost area of the North Pacific (13 out of 176, and 5 are shown here as examples) and the westernmost area of the North Pacific (4 out of 176 for the case influenced by local currents (red) and only 1 out of 176 for the case by the subtropical circulation (green)). The pairs of numbers represent the estimated travel time to the destination and the standard deviation evaluated by a corresponding set of observed trajectories, respectively.

Statistical analysis was conducted on the trajectories of all nuclear pollutant particles and it was found that the particle trajectories look similar to the drifters’. Carried by the Kuroshio and Kuroshio-extension currents and the North Pacific Current, most particles move eastward and 3808 among them eventually enter the easternmost area of the North Pacific (Figs.7 and 10). Some particles firstmove eastward and then towards southwest at the central Pacific under the influence of the subtropical current. Among them, 7608 turn westward and eventually move to the east coast area of China (Figs.7 and 10; note that only 1 drifter is found in the observational dataset). In addition, local eddy-associated circulation carried 765 tracers to the east coast area of China (Figs.7 and 10). The number of the particles for each path is much greater than that of surface drifters, and the proportion of the particle numbers among different drifting paths (765:3808:7608) is also different from that of surface drifters (4:1:13). In addition, among the particles moving to west, some can finally arrive at the coastlands of China. But the drifters reaching the coastlands of China have been not yet observed.

The transport time scales of the nuclear pollutants traveling to the easternmost and the westernmost areas of the North Pacific, are estimated based on the travel time of the observed drifters and simulated nuclear pollutant particles. From a group of trajectories, the mean travel time and the standard deviation of travel time are estimated and shown in Figs.8 and 10. Generally, the transport time estimated from the simulated nuclear pollutants is longer than that from the observed drifters. Obviously, the sampling size and the reanalysis current quality influence the estimation of the travel time. If the samples of drifters are enough many in number and the drifters passing by the source region of nuclear pollutants in other months and with short lives are removed, the results from the surface drifters will be more reliable. The higher the precision of the reanalysis data is, the more accurate the results from particles are. In the study, the greatest uncertainty is in the time scale of nuclear pollutants reaching the westernmost area of the North Pacific along with the subtropical circulation branch. For example, the estimated time of one drifter is 1.6 years while the estimated average time of the 7608 simulated particles is 3.6 years. For two other subgroups of trajectories, the uncertainty in the transport time is relatively small. The time for nuclear pollutants to reach the easternmost area of the North Pacific is 3.2 years if estimated using the drifter data or 3.9 years using the simulated particles. Carried by local currents, the nuclear pollutants need 1.3 years to reach the east coast area of China if estimated from the drifter data or 1.8 years from the simulated particles. High levels of radioactive iodine-131 (with a half-life of about 8 days), cesium-137 (with a half-life of about 30 years), and cesium-134 (with a half-life of about 2 years) have been detected in the seawater adjacent near Fukushima. The long half-life of cesium-137 will produce more damage. Based on the method described in Section 3, the spatial distribution of the impact strength of radioactive element Cesium-137 is analyzed. Fig.11 shows the distribution of the impact strength of Cesium-137 at year 1.5 (panel a), year 3.5(panel b) and year 4 (panel c), respectively. Cesium-137 has a high concentration in the middle of the North Pacific at year 1.5 and the impact strength is up to 4%. Since the nuclear pollutants are in the area mostly influenced by the eastward North Pacific Currents, the center of the impact strength of Cesium-137 moves to the eastern part of the North Pacific at year 4 and the impact strength is also as high as 4%. Due to the role that local eddies and subtropical currents play, a small amount of Cesium-137 can be transported across the Kuroshio to the east coast area of China. Fig.11 shows that starting from 1% in the east coast area of China at year 1.5, the impact strength continues to increase to 3% at year 4.

Fig.9 Five drifter trajectories reaching the east coast of China after they crossed the dashed-red line at the blue dot locations.

Fig.10 The trajectories of the nuclear pollutant particles. Five trajectories are shown for each group. The mean travel time and the standard deviation are calculated by a corresponding set of trajectories.

Fig.11 The distributions of the impact strength of Cesium-137 at year 1.5 (a), year 3.5 (b), and year 4 (c), respectively.

5 Discussion and Summary

Based on news reports on nuclear pollutants during March 12–15, 2011, a small semi-circle region centered at Fukushima with a 300 km radius is defined as the source region of nuclear pollutants. All surface drifter data of 1979–2011 around the region are collected to conduct the statistical analysis of pollutant transport paths and to estimate the time scales of the pollutants reaching different parts of the Pacific Ocean from Fukushima. To enhance the representativeness of the statistics, the 25-year reanalysis gridded surface currents from CORA are used to produce a great number of nuclear pollutant particles. Based on the trajectories of these particles, the impact strength of the Fukushima nuclear pollutants is estimated corresponding to different time scales.

The results show that most nuclear pollutant particles moving eastward reach the easternmost part of the North Pacific within 3.2–3.9 years. The nuclear pollutants moving westward follow two pathways. Carried by local circulation, the nuclear pollutants reach the westernmost part of the North Pacific within 1.3 years using the drifter data and 1.8 years using the particle tracking calculations. Influenced by subtropical circulation, the time scale for reaching the westernmost part of the North Pacific is 1.6 years using the drifter data and 3.6 years using the particle tracking calculations.

The impact strength distributions at different times are given. The results show that the eastward Kuroshioextension currents and North Pacific currents are the major carriers of nuclear pollutants, and the accumulation of Cesium-137 occurs in the east part of the North Pacific and the impact strength is as high as 4% at year 4. Associated with local eddies, a small amount of Cesium-137 is transported to the east coast area of China across the Kuroshio axis, and its impact strength reaches up to 1% at year 1.5 and increases to 3% at year 4.

To improve the accuracy of nuclear pollutant transport and impact strength in future studies, larger datasets and three dimensional flow fields need to be included.

Acknowledgements

This research is supported by the National Basic Research Program (Grant No. 2013CB430304), the National Natural Science Foundation of China (Nos. 41206178, 41030854, 41106005, 41176003 and 41306006), and the National High-Tech R&D Program of China (No. 2013 AA09A505).

Buck, E. H., Upton, H. F., and Folger, P., 2011. Effects of radiation from Fukushima Daiichi on the U.S. marine environment. Congressional Research Service (CRS) Report for Congress, R41751: 4.

Derber, J., and Rosati, A., 1989. A global oceanic data assimilation system. Journal of Physical Oceanography, 19: 1333-1347.

Han, G. J., Fu, H. L., Zhang, X. F., Li, W., Wu, X. R., Wang, X. D., and Zhang, L. X., 2013. A global ocean reanalysis product in the project of China ocean reanalysis (CORA). Advances in Atmospheric Sciences, 30 (6): 1621-1631.

Han, G. J., Li, W., Zhang, X. F., Li, D., He, Z. J., Wang, X. D., Wu, X. R., Yu, T., and Ma, J. R., 2011. A regional ocean reanalysis system for coastal waters of China and adjacent seas. Advances in Atmospheric Sciences, 28: 682-690.

Li, W., Xie, Y. F., He, Z. J., Liu, K. X., Han, G. J., Ma, J. R., and Li, D., 2008. Application of the multi-grid data assimilation scheme to the China Seas’ temperature forecast. Journal of Atmospheric and Oceanic Technology, 25 (11): 2106-2116.

Li, W., Xie, Y., Deng, S.-M., and Wang, Q., 2010. Application of the multigrid method to the two-dimensional doppler radar radial velocity data assimilation. Journal of Atmospheric and Oceanic Technology, 27: 319-332.

Li, Z. Q., Wu, B. Q., Yuan, Y. M., Bo, W. B., 2011. Drift-path Analysis of the Fukushima Nuclear Leakage Pollutant. Hydrographic Surveying and Charting, 31 (4): 47-49.

Liu, A. H., and Kuai, L. P., 2011. A review on radionuclides atmospheric dispersion modes. Journal of Meteorology and Environment, 27 (4): 59-65

Marshall, J., Hill, C., Perelman, L., and Adcroft, A., 1997. Hydrostatic, quasi-hydrostatic and nonhydrostatic ocean modeling. Journal of Geophysical Research, 102: 5733-5752.

Neumann, G., 1968. Ocean Currents. Elsevier Oceanographer Series, Elsevier Co., New York, 351pp.

Shimizu, Y., Yasuda, I., and Ito, S.-I., 2001. Distribution and circulation of the coastal Oyaishio intrusion. Journal of Physical Oceanography, 31: 1561-1578.

Quirin, S., 2011. Radiation release will hit marine life. Nature, 472: 145-146.

Qiao, F. L., Wang, G. S., Zhao, W., Zhao, J. C., Dai, D. J., Song, Y. J., and Song, Z. Y., 2011. Predicting the spread of nuclear radiation from the damaged Fukushima Nuclear Power Plant. Chinese Science Bulletin, 56 (18): 1890-1896, DOI: 10.1007/ s11434-011-4513-0.

Tomczak, M., and Godfrey, J. S., 2001. Regional Oceanography: An Introduction. Elsevier Science Ltd., Oxford, UK, 110-120.

Tsumune, D., Tsubono, T., Aoyama, M., and Hirose, K., 2011. Distribution of oceanic 137Cs from the Fukushima Dai-ichi Nuclear Power Plant simulated numerically by a regional ocean model. Journal of Environmental Radioactivity, DOI: 10.1016/j.jenvrad.2011.10.007.

Xie, Y., Koch, S., Mcginley, J., Albers, S., Bieringer, P. E., Wolfson, M., and Chan, M., 2011. A space-time multiscale analysis system: A sequential variational analysis approach. Monthly Weather Review, 139: 1224-1240.

(Edited by Xie Jun)

(Received February 27, 2012; revised July 23, 2012; accepted May 30, 2013)

© Ocean University of China, Science Press and Springer-Verlag Berlin Heidelberg 2014

* Corresponding author. Tel: 0086-22-24010846

E-mail: gjhan@mail.nmdis.gov.cn