选用长江三峡水库区间流域的历史雨洪资料与短期定量降水预报资料，编制三峡水库区间流域洪水预报模型，并将区间定量降水预报与区间洪水预报模型相耦合，研究了预见期降水对洪水预报的影响。提出了一个随机降水模型，随机生成 500组序列作为降水预报值输入到区间洪水预报模型，并以均值作为预报结果发布。方案比较结果表明，考虑预见期内的降水预报可提高三峡水库的洪水预报精度。

By using the historical data of the Three Gorges Reservoir (TGR), a flood forecasting model for TGR intervening basin was proposed and established, with which the quantitative precipitation predicted data was coupled to study the influence of flood forecasting accuracy. A stochastic precipitation generator was developed to generate 500 precipitation series as the inputsof proposed flood forecasting model. The mean values of the forecasted discharges were used as the final forecasted results. It showed that the proposed flood forecasting model coupled with the stochastic precipitation generator could increase the TGR flood forecasting precision.