GFS可利用性研究及其在旬径流预报中的应用

Usability Analysis of Global Forecasting System(GFS) and Its Application in Ten-day Runoff Forecasts

分析了中长期降雨径流预报研究应用现状和存在的主要问题,研究了利用GFS降雨数值预报信息进行旬径流预报的方法,建立了旬降雨—旬初土壤含水量—旬径流预报模型(PSRM)。通过对实例流域历史旬径流资料的模拟优选出模型参数,分析了GFS降雨数值预报信息的可利用性,并给出了利用GFS降雨预报信息和PSRM模型推求径流的方法。结果表明,基于GFS预报降雨信息和PSRM模型推求径流期望值和概率分布用于水库的实时调度可行。

Application status quo of mid-long term precipation and runoff forecasts and main existing problems are analyzed. Ten-days runoff forecast based on precipition forecast information using GFS is studied. Firstly, based on analyzing cause-effect we developed the （PSRM）model within precipitation, runoff quantity in ten days and soil moisture when current ten days started is developed. We optimize model parameters by simulating historical data and the modified results show a high accuracy. Then, based on accuracy analysis of digital precipitation forecast data GFS, we can predict ten-day runoff quantity using PSRM and GFS rainfall forecast results. The applied watershed runoff prediction results showed that runoff possibility distribution corresponding forecasted graded precipitation is clear. And the forecasted runoff can be used for reservoir operation dispatching making.