Optimization And Evaluation Of Temperature Vegetation Drought Index (TVDI)

Drought is the most widely affected natural disaster in the world,and it is also the most affected agricultural natural disaster in our country.Due to its long duration,wide impact,and large disaster losses,drought has gradually become one of the most severe natural disasters in the world.Severe drought can lead to crop failure,frequent forest fires,and reduction of water bodies.Accurately monitoring drought conditions and grasping the laws of its occurrence and development can provide decision-making departments with a scientific and reasonable basis,which is of great significance for disaster reduction and prevention.Remote sensing technology has the advantages of wide coverage,fast acquisition speed,low cost,and abundant information.It can effectively obtain drought information in a large area and dynamically in real time,and calculate various parameters or indicators that can directly or indirectly reflect drought conditions.It has become a common method for drought monitoring.A variety of drought monitoring methods have been formed.Located in the northwest of China,Shanxi Province is a typical arid and semi-arid area,known as the Nine Droughts in Ten Years.This paper takes Shanxi Province as the study area,based on MODIS surface temperature(MOD11A2),surface reflectance(MOD09A1)and leaf area index(MOD15A2H)data,using the temperature vegetation drought index(TVDI)model construction principles,the surface temperature and vegetation index in the model were improved and optimized respectively.The main contents of the study are as follows:(1)In order to reduce the influence of elevation and latitude changes on LST,the accuracy was higher after LST was corrected based on DEM and latitude;(2)In order to overcome the shortcomings of NDVI being easily saturated and greatly affected by water vapor,the EVI-Ts and LAI-Ts feature spaces were constructed,and the NDVI-Ts(TVDI),EVI-Ts(TEDI)and LAI-Ts(TLDI))was compared and verified.Then,the optimal model for drought monitoring in Shanxi Province was obtained;(3)Considering the influence of night temperature on the surface temperature,VI-Ts and VI-?Ts feature spaces based on day-night temperature difference(?Ts)and day-night average temperature(Ts)are constructed,and they are verified and evaluated;(4)Using the optimized model to monitor the agricultural drought in Shanxi Province from 2001 to 2015,and analyze the response relationship between meteorological factors and agricultural drought.Through research,the following main conclusions are obtained:(1)After improving the elevation and latitude of the surface temperature,the accuracy of TVDI for monitoring drought in Shanxi Province has been effectively improved,and the correlation with the measured soil relative humidity data has been greatly improved.It has also effectively avoided the occurrence of southern drought and north wetness and high drought and low humidity errors.(2)EVI and LAI replace NDVI to construct a TVDI model,which can effectively avoid the shortcomings of NDVI that are easily saturated in high vegetation coverage areas,and in the feature spaces of NDVI-Ts,EVI-Ts and LAI-Ts,parabola is used to fit dry edges and wet edges is better.Among the six different TVDI models constructed using linear and parabolic NDVI-Ts,EVI-Ts and LAI-Ts feature spaces,the parabolic feature space TLDIp constructed by LAI and land surface temperature has the best monitoring effect on drought.(3)ATLDIp constructed by VI-Ts feature space which based on the day and night average temperature cannot represent the drought information in Shanxi Province.And DTLDIp constructed by VI-?Ts feature space which based on the day and night temperature difference has a lower accuracy for drought monitoring in Shanxi Province than TLDIp,indicating that the day surface temperature is most suitable for constructing temperature vegetation drought monitoring model.(4)From 2001 to 2020,the drought,in Shanxi Province,was in a state of northern drought and southern wetness,and with the increase of months,the drought situation first eased and then aggravated,with obvious seasonal changes.The most severe droughts is in spring and winter.The summer drought is the most relaxed,and in the past 20 years,the drought has shown a trend of alleviation.(5)The drought in Shanxi Province is jointly affected by precipitation and high temperature.This phenomenon is related to the topographical distribution of Shanxi Province:drought in northern Shanxi is affected by precipitation and high temperature,drought in central and southwestern Shanxi is mainly affected by high temperature,however,due to the large rainfall and numerous rivers in the southeast of Shanxi,the drought is basically not affected by climatic factors.