Frequent flooding events and intensive agricultural, economic and industrial activities in Singapore-Malaysia Catchment have made Singapore coastal waters come under high risk of land-based pollution. A coupled hydrologic-hydrodynamic model is employed to perform three-dimensional land-based pollutant transport simulations in Singapore coastal waters. The hydrologic model component (2012 version of Soil and Water Assessment Tool) is found to be able to predict streamflow accurately, with correlation coefficients larger than 0.9 and a high Nash-Sutcliffe efficiency of more than 0.8. The hydrologic model (SWAT) is coupled with the hydrodynamic model (SUNTANS) by transferring streamflow and pollutant concentrations at nine rivers along common boundaries. The coupled hydrologic-hydrodynamic model is validated with observed sea surface elevations and velocities. A low Root-Mean-Square-Error (RMSE) of 0.10m and a high correlation coefficient of 0.98 are observed for sea surface elevations. The coupled model predicts depth-averaged U and V velocities accurately, with low RMSEs of 0.06m/s and 0.07m/s respectively and high correlation coefficients exceed 0.95. During the Northeast monsoon, pollutants from Source 1 (Johor River), Source 2 (Tiram River), Source 3 (Layang River) and Source 4 (Layau River) are dispersed into the Singapore Strait after around 2days of release, with Johor Estuary and Tekong Island highly affected. During the Southwest monsoon, it takes around 9days for pollutants from Sources 1–4 to affect the whole Johor Estuary and Tekong Island, which almost four times of dispersion time during the Northeast monsoon. The overall mean dispersion coefficient K¯ of 47.99m2/s during the Northeast monsoon is roughly four times that of 11.33m2/s during the Southwest monsoon, due to larger amount of land-based pollutants are introduced by streamflow in December and differences in the large-scale monsoon effects. It is found that the dispersion coefficient K obeys a “4/3-law”, with the length scale L defined as the distance of the center of concentration travelled.
Read full abstract