大气电场数据与雷电相关性的深度学习算法

雷电是一种对流天气现象，也是发生频繁的自然灾害之一，破坏力极大。由于雷暴的时间短、范围小，精准的雷暴预警在天气预报中较为困难。目前各种气象监测设备很多，但数据利用率较低，其中利用大气电场仪数据的变化来判断雷电是否发生的研究较少。因而，通过探索二者间的相关性，对于建立更有效的雷电预警模型是有意义的。而本文利用卷积神经网络分析并验证了大气电场变化和雷电的关系，模型的分类结果更直观地反映了二者存在相关性。 

Deep Learning Algorithm for the Correlation between Atmospheric Electric Field Data and Thunderstorm

Thunderstorm is a kind of convective weather phenomenon, which is one of the frequent natural disasters and has great destructive power. Due to its short duration and small range, precise thunderstorm warning in the weather forecast is difficult. At present, there are many kinds of weather monitoring equipment, but the data utilization is low, and the research which use the atmospheric electric field data to judge the occurrence of lightning is relatively less. Thus, it is significant to establish a more effective thunderstorm forecasting model by exploring the correlation between the atmospheric electric field data and lightning. In this paper, we use the convolution neural network to analyze and verify the relationship between them. The classification results of the model reflect the correlation between those two kinds of data more directly.