Abstract: Numerical product has been an important means to improve the anticipation period of hydrological forecast. In the purpose of understanding areal rainfall forecast accuracy and error distribution of ECMWF (European Centre for Medium Range Weather Forecasts) and WRF (Weather Research and Forecasting Model) and providing scientific support for improving the robustness of flood forecast accuracy, we verified the forecast results of areal rainfall in flood season of 2016 in seven subareas of Jialing River Basin by examining indexes involving TS score, false alarm rate, point over rate, and success hit rate. Furthermore, we analyzed the relationship between the abovementioned indexes and the period of forecast validity in the subareas. Results suggest that ECMWF product and WRF product could be both applied to the sunny/rainy forecast in the study area. The forecast accuracy improves with the increase of rainfall grade, but reduces with the decrease of forecast validity period. In general, ECMWF has better forecast efficiency in the Jialing river basin.

Key words: areal rainfall, numerical forecast product, Jialing River Basin, precipitation classification, forecast precision evaluation