Model development for nutrient loading from paddy rice fields

The objective of this study was to develop a model groundwater loading effects of agricultural management systems in paddy fields (GLEAMS-PADDY) to predict nutrient loading to surface waters from paddy rice fields. This model was developed by modifying the GLEAMS model that is used for uplands. The GLEAMS-PADDY model is composed of hydrology and chemical sub-models. In the hydrology sub-model of the GLEAMS-PADDY model, the ponded depth routing method was used to handle ponded water condition of paddy fields. With the chemical sub-model, the soil was assumed to be saturated, and the soil profile in the root zone was divided into oxidized and reduced zones. Field experiments were performed at the Soro region, Chungbuk Province of South Korea, from May to September 1999 and 2000. Field data included rainfall amount, irrigation water input, drainage water output, and percolation. Concentrations of total nitrogen (T-N) and total phosphorus (T-P) in irrigation water, rain water, ponded water, surface drainage water, and percolated water were analyzed. The coefficient of determination (r(2)) and root mean square error (RMSE) between observed and model predicted surface drainage volume relative to the 1:1 line were 0.66 and 66.6 mm, respectively. The r(2) and RMSE between observed and predicted T-N concentrations in the ponded water were 0.74 and 1.43 mg l(-1), respectively. The same statistics for T-P were 0.64 and 0.03 mg l(-1), respectively. The observed and predicted total T-N loadings from surface drainage were 72.4 and 72.6 kg ha(-1), respectively. The observed and predicted total T-P loadings were 1.70 and 1.69 kg ha(-1), respectively. The r2 and RMSE between observed and predicted T-N concentrations in the surface drainage water were 0.79 and 2.71 mg l(-1), respectively. The same statistics for T-P were 0.73 and 0.04 mg l(-1) respectively. Comparisons of observed and model predicted water balance components, nutrient concentrations, and loading rates exhibited reasonably good agreement. Hence, the GREAMS-PADDY model can be used to predict nutrient loading from paddy fields and to develop BMPs in paddy rice culture to reduce surface water pollution from paddy field drainage water. (C) 2003 Elsevier Science B.V. All rights reserved.