闵行智能网格预报降水订正数据产品

基于机器学习的闵行区智能网格降水预报订正数据是对智能网格降水预报采用新型的机器学习模型进行订正后的每日08时起报未来24小时降水预报数据，订正后的网格降水基本接近实况降水，每天1期，每期约3～70KB，有较高的参考价值，针对闵行区局地降水预报效果有较大改进，提升了闵行区降水预报服务精确度。

The machine learning-based corrected dataset for intelligent grid precipitation forecasting in Minhang District refers to 24-hour ahead precipitation forecast data initialized at 08:00 daily, which is generated by correcting the original intelligent grid precipitation forecasts using a novel machine learning model. The corrected gridded precipitation data is generally consistent with observed precipitation. This dataset is released once per day, with each file sized approximately 3 to 70 KB and holds considerable reference value. It significantly improves the forecasting performance of local precipitation in Minhang District and enhances the accuracy of precipitation forecast services in the region.