The invention discloses a 
lightning intensity value predicting method based on atmospheric 
electric field data, which is applied to the field of 
lightning early warning. The disclosed 
lightning intensity value predicting method is used for converting original one-dimensional atmospheric 
electric field data into a two-dimensional 
image representation through extracting depth features of the atmospheric 
electric field data, mining depth feature information of the image, establishing a relationship model of the atmospheric electric 
field data and a lightning intensity value by combining with a non-
linear regression analysis method, and predicting the lightning intensity value. According to the lightning intensity value predicting method, various kinds of detection data is used for performinglightning early warning, i.e., the atmospheric electric 
field data and lightning data; the one-dimensional atmospheric electric 
field data is converted into the two-dimensional 
image representation byadopting a dimension 
conversion method, so as to enhance the structure representation of the atmospheric electric field data; the depth feature information of the atmospheric electric field data is extracted by using a depth model; and the lightning intensity value is predicted by using the atmospheric electric field data, thereby filling a gap of low utilization of the atmospheric electric fielddata in other predicting methods and a gap of lightning intensity value prediction.