Forecasting and Classification of Severe Convective Weather Based on Numerical Forecast and Random Forest Model
(Meteorological Observatory of Zhejiang Province)
本文已被:浏览 124次   下载 26
投稿时间:2017-09-15    修订日期:2018-03-14
中文摘要: 提要:随机森林算法是当前得到较为广泛应用的机器学习方法之一,有着很高的预测精度,训练结果稳定,泛化能力强,解决多分类问题有明显优势。本文将随机森林算法应用于强对流的潜势预测和分类,分短时强降水、雷暴大风、冰雹和强对流四种类别,基于2005-2016年NCEP1×1°再分析资料计算的对流指数和物理量,开展强对流天气的分类训练、0-12 h预报和检验,经2015-2016年独立测试样本检验表明,针对强对流发生站点的点对点检验,整体误判率为21.9%,85次强对流过程基本漏报,模型尤其适用于较大范围强对流天气。随机森林算法筛选的因子物理意义较为明确,和主观预报经验基本相符,模型准确率高,可用于日常业务。
Abstract:Abstract: The random forest algorithm is currently one of the more widely used machine learning methods, with a high prediction accuracy, stable training results and generalization ability. Solving the problem of multi-classification has obvious advantages. Random forest algorithm is applied to the prediction and classification of severe convective weather, which is divided into four categories: short-duration heavy rainfall, thunderstorm wind, hail and no severe convection. Based on the data of convection index and physics calculated from the NCEP data of 2005-2016 , then the training, 0-12 h forecasting and testing of classified severe convection are carried out. The results show that: The whole misjudgment rate is 21.9% that calculated from the independent data of 2015-2016. It has almost no omission in 85 examples of severe convective weather and the model is especially suitable for a larger range of severe convective weather. The physical meaning of the factors used in random forest algorithm is relatively clear, and basically consistent with the subjective forecasting experience, can be used for daily business.
文章编号:201709150424     中图分类号:    文献标志码:
李文娟,0.[en_title][J].Meteor Mon,():-.
LI WEN JUAN,0.Forecasting and Classification of Severe Convective Weather Based on Numerical Forecast and Random Forest Model[J].Meteor Mon,():-.