Monitoring The Spatial Pattern Of High Temperature In The Yangtze River Delta By Remote Sensing

Under the background of global warming,the relationship between climate change and urbanization is growing more and more closer.High temperature disaster is one of the most prominent manifestations of climate change,which has a great impact on resident health,social life and ecosystems.Monitoring the spatial pattern of high temperature from the perspective of urban agglomerations is of great significance for high temperature phenomena in continuous areas.This paper will take the Yangtze Delta urban agglomeration as the study area.Based on multi-source remote sensing data,the average maximum temperature of the study area was estimated,and the spatial pattern characteristics of temperature in the study area were discussed,which provide references for the research of high temperature in urban agglomerations.The main conclusions are as follows:(1)Compared with the multiple linear regression model,using random forest model to estimate the near-surface temperature has obvious advantages,and the result has higher accuracy.The importance of each independent variable factor is analyzed,and it is found that the surface temperature(LST)has a significant effect on the model error.The random forest(RF)model is applied to the study area,and the average maximum temperature of the study area estimated was obtained by remote sensing.The accuracy evaluation of the results shows that the average maximum temperature estimated by RF model has a good correlation with the average maximum temperature observed by the station,indicating that the average temperature of remote sensing estimated under clear sky conditions can be used to reasonably characterize the spatial distribution of air temperature during high temperature.(2)The mean absolute error(MAE)between the observed temperature and the estimated temperature by remote sensing under cloudless conditions during high temperature is calculated and analyzed.The results show that the temperature estimated by remote sensing can better reflect the spatial pattern of the study area.(3)With the Heat Intensity Index as the standard,the high temperature phenomena in the study area were graded and evaluated.The data shows that the average maximum temperature in the study area has significant spatial differences.Discussing this,the results show that the proportion of heat intensity gradeshas a great correlation with the development of the local economic level,topographical factors and the location of land and sea.(4)The Heat Proportion Index of each city in the study area was calculated and analyzing the results.The results show that HPI is significantly affected by high temperature area,vegetation coverage area,land and sea location,and topographical factors.