Prediction of chemicals runoff m watershed based on the combined water and mass transfer model

点击翻译按钮获取中文摘要

A model analysis of chemicals runoff in watershed was constructed, based on the combination of chemicals transfer model and water transfer model. Watershed area was divided into grids of 1km x 1km and each grid was further divided vertically and horizontally to 13 compartments. The Chikugo River basin was analyzed in which agrochemical compounds were monitored frequently (five times per week). Runoff of agrochemicals used in paddy field was predicted.Data of land utilization were collected from digital map and information published by the National Geographical Survey Institute. Weather information was obtained from the AMEDAS data. River flow rate was estimated from the observed values at the observation points and the reserved volume and discharge flow rate at the reservoir dam.Data on farm work such as date and place of agrochemical sprinkling and spraying, date of rice transplanting, water management schedule are regarded uncertain. Characteristics of the agrochemical compound such as adsorption coefficient and biological decomposition rate constant are also regarded uncertain. The value of these parameters was set using Monte Carlo method based on the probability distribution of matters taking place. The conventional simulation by a deterministic algorithm was also applied for comparison. As a result, the following results were obtained.(1)Estimated concentration level of agrochemicals by the Monte Carlo simulation are more dosed to the actual observed value in rivers than that estimated by the deterministic algorithm. Thus, the Monte Carlo simulation was proved to give much accurate prediction. (2)It was indicated that runoff of an agrochemical compound can probably be estimated if the total amount of consumption of the compound in the whole watershed area is given, even though the precise schedule and accurate amount of agrochemical sprinkling in each paddy field are unknown.