中文摘要：

通过单向耦合GRAPES-MESO中尺度数值模式与淮河流域精细化水文模型，建立了淮河流域精细化渍涝风险预报模型。首先将GRAPES-MESO的降水和日最高、最低温度预报结果输入精细化可变下渗容量（variableinfiltrationcapacity，VIC）水文模型，然后输出15km×15km网格点上的径流深度和三层土壤湿度，按照全国渍涝风险气象预报预警发布标准计算风险等级。最后选择2008年7月淮河流域两次强降水过程进行试验，同时与原始国家级大尺度50km×50km渍涝风险预报模型对比预报效果。结果表明：该模型有效解决了原始国家级渍涝风险预报模型的漏报问题。在连续雨日可能造成的渍涝风险上，比原始大尺度渍涝模型有更好的预报效果，提高了渍涝风险气象预报模型精度。

英文摘要：

A forecast model of waterlogging and water logged risk for Huaihe basin is established by coupling the GRAPESMESOnumerical model and the hydrological model VIC (Variable Infiltration Capacity) with fine resolution. Firstly, the precipitation and temperature forecasts by the GRAPES-MESO are used to force the VIC in order to calculate the runoff depth and three layers of soil moisture with 15 km×15 km grids. Then the risk level of water logging and water logged are obtained based on the results of VIC according to the national forecast and early warning publishing standards of water logging and water logged. Finally, the forecasts resulted from the highly-resolved model were compared with those from the raw large scale model in 50 km×50 km grids by taking 2 heavy rainfall events in July, 2008 in Huaihe River Basin as test cases. The results show that the highly-resolved model is effective to solve the problem of omission of forecast and performs better than the original large scalemodel with lower resolution when the forecast periods are on continuous rainy days.