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 intelligent grid precipitation forecast dataset for Minhang District comprises 24-hour precipitation forecasts initialized at 08:00 each day, which are obtained by correcting the initial intelligent grid precipitation forecasts using a novel machine learning model. The corrected grid precipitation outputs are generally consistent with ground-truth precipitation observations. It is released once daily, with each file size ranging from approximately 3 to 70 KB and possessing high reference value. It yields significant improvements in local precipitation forecast performance for Minhang District, thereby enhancing the accuracy of precipitation forecast services in the district.

搜集汇总

背景与挑战

背景概述

该数据集是基于机器学习模型对闵行区智能网格降水预报进行订正的数据产品，每日08时起报未来24小时降水预报，订正后网格降水更接近实况，有效提升了闵行区局地降水预报的精确度和服务效果。