###
气象:2021,47(1):60-70
本文二维码信息
码上扫一扫!
基于卷积神经网络的逐时降水预报订正方法研究
陈锦鹏,冯业荣,蒙伟光,文秋实,潘宁,戴光丰
(福建省灾害天气重点实验室,福州 350001; 数据科学与统计重点实验室,漳州 363005; 福建省漳州市气象局,漳州 363005; 中国气象局广州热带海洋气象研究所/广东省区域数值天气预报重点实验室,广州 510640; 福建省气象台,福州 350001)
A Correction Method of Hourly Precipitation Forecast Based on Convolutional Neural Network
CHEN Jinpeng,FENG Yerong,MENG Weiguang,WEN Qiushi,PAN Ning,DAI Guangfeng
(Fujian Key Laboratory of Severe Weather, Fuzhou 350001; Fujian Key Laboratory of Data Science and Statistics, Zhangzhou 363005; Zhangzhou Meteorological Office of Fujian Province, Zhangzhou 363005; Guangzhou Institute of Tropical and Marine Meteorology, CMA/Key Laboratory of Regional Numerical Weather Prediction of Guangdong Province, Guangzhou 510640; Fujian Meteorological Observatory, Fuzhou 350001)
摘要
图/表
参考文献
相似文献
本文已被:浏览 82次   下载 641
投稿时间:2020-02-09    修订日期:2020-10-20
中文摘要: 应用2017—2018年5—9月福建省观测资料对华南区域中尺度模式(GTRAMS-3 km-RUC)预报进行站点检验,建立和训练基于卷积神经网络的逐时降水分级订正模型,并与频率匹配法进行2017—2018年测试集的对比试验和2019年数据集的模拟业务检验,探讨了试验过程中遇到的样本不均衡、特征变量选取以及模型过拟合问题。结果表明:模式对于15 mm·h-1以上降水的预报能力弱,各订正方法对原始预报均有不同程度的改进作用。从评估指标来看,基于卷积神经网络的订正方法比频率匹配法表现出优势,其中相关系数判别方案下的网络模型对强降水预报的订正效果显著优于其他方法;在输入特征变量选取方面,应用主成分分析方案的模型训练收敛速度比相关系数判别方案更快,最佳训练期有所提前,但也更早进入严重的过拟合状态,而相关系数判别方案能够使网络模型的训练拥有更长的提升期以达到更具“潜力”的状态;基于卷积神经网络的订正方法对减少分类降水预报的漏报率、晴雨和弱降水预报的空报率具有显著作用,其优化程度明显超过频率匹配法。
Abstract:In this article, precipitation forecasts from the South China regional hourly updated cycle mesoscale model (GTRAMS-3 km-RUC) are verified based on observations of Fujian Province during May to September in 2017-2018. A correction approach on categorical hourly precipitation forecast using convolutional neural network (CNN) is built and trained, taking into account of issues such as sample imbalance, selection of feature and over-fitting of the system. Comparisons of test between CNN and frequency matching (FM) method on the 2017-2018 test set and the 2019 practical data set are also made for their correction effects. The results show that all these correction methods can improve the original forecast at different degrees, but underperform with precipitation over 15 mm·h-1. CNN performs better than FM, and CNN with correlation coefficient discrimination (CCD) is the most effective for heavy rainfall category. In addition, we apply two different solutions to extract input features of the neural network system. The CNN model converges faster when principal component analysis (PCA) solution is used, but it leads to over-fitting earlier and severely, which implies that the system’s generalization ability needs to be improved further. In contrast, CCD solution displays its longer promotion period and more potential state. It seems that CNN correction method improves forecast ability mainly by reducing the missing ratio for categorical precipitaition forecasts and the false alarm ratio for the rainfall and light rain forecasts.
文章编号:     中图分类号:    文献标志码:
基金项目:国家自然科学基金联合基金项目(U1811464)、泛珠三角区域数值预报联合发展专项(201804和201904)、中国气象局预报员专项(CMAYBY2020-061)、广州市科技计划项目(201903010104)和数据科学与统计重点实验室开放课题(42010702)共同资助
引用文本:
陈锦鹏,冯业荣,蒙伟光,文秋实,潘宁,戴光丰,2021.基于卷积神经网络的逐时降水预报订正方法研究[J].气象,47(1):60-70.
CHEN Jinpeng,FENG Yerong,MENG Weiguang,WEN Qiushi,PAN Ning,DAI Guangfeng,2021.A Correction Method of Hourly Precipitation Forecast Based on Convolutional Neural Network[J].Meteor Mon,47(1):60-70.