| Lightning is a convective weather phenomenon that frequently occurs in nature.Fires such as forests and grasslands caused by lightning(lightning fires)occur from time to time,which have a serious impact on human life and natural resources.Therefore,understand the characteristics of lightning activity and identify lightning fire is vital to the prevention of lightning disasters.At present,the research on lightning activity in China is mainly based on several Lightning high-prone areas such as South China and the Qinghai-Tibet Plateau.Using geostatistical methods to carry out national-scale lightning activity characteristics research and the exploration of lightning fire identification methods are relatively rare.Considering the availability of data and other factors,the study selected 2015,when strong convective weather was more frequent,as an example(the year was the warmest year since 1961),and used the World Wide Lightning Location Network(WWLLN)data?European Center for Medium-Range Weather Forecast(ECMWF)and National Center for Environmental Prediction(NCEP)to reanalyzed the data of 2015,and the analysis of the spatial-temporal distribution of lightning activities and spatial autocorrelation characteristics of the year was carried out to study the correlation characteristics of lightning and meteorological elements.Based on this,combined with the fire data monitored by satellite remote sensing,natural and socio-economic data such as population distribution and Land-use in 2015,a set of lightning fire identification algorithms considering multiple factors was designed,and the relationship between lightning fire and meteorological elements was analyzed.The relationship between lightning and fire provides a reference for the risk prediction and prevention of lightning fire.The main research content and results of the article are as follows:(1)The analysis of the lightning spatial-temporal distribution in 2015 was carried out on the grid and county scales.The results showed that the spatial-temporal distribution characteristics of lightning frequency in China during that year were similar to those in previous years,with a spatial distribution pattern of"high in the south and low in the north"in space.South China,East China and Southwest China are the three main lightning-prone areas and there were obvious seasonal differences in time.The lightning frequency in spring and summer is higher than that in autumn and winter.The highest peak of lightning activity in South China and Central China occurs in spring,while the lightning activity in Southwest,East,Northwest,Northeast and North China occurs in summer.Further,through spatial autocorrelation analysis,it is found that there are 367 counties in China with significant"high-high"clusters of lightning frequency,which distributed mainly in northern and eastern Tibet,Western Sichuan,Southern Yunnan,Jiangxi,Fujian and most of southern China.The region should strengthen the joint prevention and control of lightning activities.(2)The relationship between lightning frequency and meteorological elements in different seasons in different lightning-prone areas on a national scale is studied.The results show that due to the complex relationship between lightning and meteorological elements,the correlation between lightning frequency and meteorological elements in different regions and in different seasons is obviously different.The lightning-meteorological element association model constructed by the commonly used Ordinary Linear Regression method(OLS)does not consider the regional heterogeneity,and it is inferior to the Geographically Weighted Regression Model(GWR)based on Geostatistics in terms of accuracy indicators such as the R-Square(R~2)and Root Mean Square Error(RMSE).Take China as an example,compared with the R~2 of OLS model,the GWR model in spring,summer,autumn, and winter has increased by 0.37?0.21?0.36 and 0.77 and the RMSE has decreased by0.73?0.31?0.06?0.33,indicating that the GWR model can better reflect the spatial heterogeneity between the large-scale lightning frequency and meteorological elements.(3)The lightning fire identification algorithm based on GIS and Remote sensing technology is designed.The algorithm combines natural and socio-economic data such as WWLLN data,fire data monitored by satellite remote sensing,population and Land-use.Users only need to set certain time and space parameters to screen out suspected lightning fire points from a large amount of fire point data.With reference to related research,the article uses 15KM as the spatial parameters and the longest 7 days as the time parameters(the time difference between lightning and ignition),combined with1KM population,Land-use data and historical fire data,selects from 12,046 fire points nationwide in 2015,115 fire points were suspected of being struck by lightning.On this basis,the correlation between lightning fire and meteorological elements was further analyzed based on the ground weather station data.It was concluded that Temperature and Wind speed are positively correlated with lightning fire.Precipitation and Relative humidity are negatively correlated with lightning fire.This establishes the probability prediction model of lightning fire,which can provide a reference for the meteorological risk prediction of lightning fire. |
PDF Full Text Request