Research On Early Warning Of Agricultural Drought Based On Remote Sensing Drought Index And Chlorophyll Fluorescence

Drought is one of the frequent natural disasters in recent decades.It not only endangers the environment of human life,but also has an extremely negative effect on the social and economic development.Inner Mongolia and Mongolia are located in arid and semi-arid regions,which are affected by drought all the year round.In addition,Inner Mongolia and Mongolia are dominated by agriculture and animal husbandry,and water resources are scarce,which will affect the development of agriculture and animal husbandry as well as the normal life of human beings.In order to strengthen the study of drought monitoring and warning in the study area,it has a positive effect on drought monitoring and warning,drought resistance and drought prevention in the study area.This paper selects the MOD11A2 data and MOD13A2 data from 2014 to 2018,and establishes the TVDI drought monitoring model through the relationship between LST and NDVI.Through the classification of drought levels,we can understand the distribution of drought,so that drought monitoring can be effectively carried out in the study area.It also combines sunlight-induced Solar Induced Fluorescence(SIF)products.Two typical areas of the eastern Mongolian Plateau and the northeastern part of Inner Mongolia were selected.According to the relationship between NDVI and SIF,a drought early warning index model based on the percentage of vegetation anomalies and the percentage of fluorescence anomalies was established to conduct drought monitoring and early warning research.The results show that in this paper,the surface temperature data and vegetation index data are used to establish the Ts-NDVI characteristic space through band operation.And the temperature vegetation drought index(TVDI)is calculated.Through the analysis of the data in the past five years,the dry and wet edge fitting figure is roughly trapezoidal shape,which conforms to the theoretical model research.Through the classification of drought grades,the overall drought-stricken area accounted for more than 90% of the total study area from 2014 to 2018.Compared with the distribution of drought-stricken areas in five years,the drought-stricken areas are the most severe in May,followed by June.Therefore,the drought-stricken measures should be strengthened from may to avoid greater impacts.Through the comparative analysis of soil moisture through SIF,the correlation coefficient R2>0.5.SIF and soil moisture shows that there is a good correlation between them.Comparing SIF with NDVI,it is found that there is a certain difference between SIF and NDVI.The highest value of NDVI appears around mid-July,and the highest value of SIF appears at about the end of June.SIF The highest value occurs slightly earlier than the time when the NDVI value appears.SIF is more sensitive than NDVI and it can respond to vegetation changes in advance.A comprehensive drought early warning model based on the percentage of vegetation anomaly and the percentage of fluorescence anomaly is established.The model can be used to predict the occurrence of drought in the next month.The model can effectively predict the drought in the study area.The combination of remote sensing drought index and chlorophyll fluorescence compensates for the shortcomings of the single method,making the drought monitoring and early warning results more reasonable and accurate,and it has important significance for agricultural drought warning and loss assessment.