Prediction of nitrogen and phosphorus transport in surface runoff from agricultural watersheds

Abstract

Agricultural surface drainage water can deposit nitrogen and phosphorus into surrounding rivers and streams, therefore accelerating eutrophication which threatens the ecosystem. Surface drainage water, from paddy fields and other agricultural lands, is influenced by numerous factors such as spatial and temporal distribution of rainfall, land topography and soil characteristics. A Generalized Regression Neural Network (GRNN) model was used to define the influence of rainfall and surface drainage water on nutrient load into the neighboring water systems by predicting the surface water quality and quantity. The data was obtained from the 15 ha paddy fields surrounded by drainage and irrigation channels. Simulations showed reasonably good predictions of surface drainage water based on historical data of rainfall (R=0.84). However, the resulting predictions for nutrient concentrations corresponding to surface drainage were somewhat varied (R=0.72 and 0.40 in total nitrogen and total phosphorus, respectively). It is suspected that the model’s prediction on nutrient concentrations consists of both natural and artificial variations of nutrient content in irrigation streams. Therefore, recommendations include providing a more.