A Review of Recent Advances in Lightning Fire Prediction
Lightning fires are the most common natural fires,which is the main cause for burned areas on global vegetation.The lightning fire occurrence is influenced by various factors,so it is difficult to predict accurately.Its prediction has been gradually applied in wildfire management with the improvement of lightning monitoring capabilities.Statistical regression model and machine learning methods are popularly used in lightning fire prediction,including logical regression,random forest,generalized linear,etc.This paper summarized recent research advances from three aspects,such as the main drivers,lightning characteristics,and prediction methods,and elaborates on the driving factors and forecasting models.The future researches were proposed incorporating with the current problems in lightning fire prediction.It is necessary to develop prediction models to satisfy various management needs according to fuel characteristics and lightning fire occurrences in different ecological zones,so as to improve the prediction ability of lightning fire occurrence.
NETL
NSTL
万方数据
