Effect of uncertainties in agricultural working schedules and Monte-Carlo evaluation of the model input in basin-scale runoff model analysis of herbicides

In the prediction of time-series concentrations of herbicides in river water with diffuse-pollution hydrological models, farming schedules (the dates of herbicide application and drainage of irrigation water from rice paddies) greatly affect the runoff behavior of the herbicides. For large catchments, obtaining precise data on farming schedules is impractical, and so the model input inevitably includes substantial uncertainty. This paper evaluates the effectiveness of using the Monte-Carlo method to generate sets of estimated farming schedules to use as input to a GIS-based basin-scale runoff model to predict the concentrations of paddy-farming herbicides in river water. The effects of using the Monte-Carlo method to compensate for uncertainty in the evaluated parameters for herbicide decomposition and sorption were also evaluated.