Inversion Of Agricultural Drought With MODIS Data In Liaoning Region

MODIS can offer surface temperature data of high attitude resolution(MOD11A2)and vegetation indices of 16-day resolution(MOD13A2).Based on the above remote sense data,the study picked NDVI and EVI in Liaoning between 2013 and 2014,and used RVI with LST to construct the feature space of 16-day interval from March to September in 2013 and 2014.From the three feature spaces,TVDI,TEDI and TRDI were calculated so as to measure the soil moisture in Liaoning.Combining the 27 practical observations of soil moisture and 61 observed data of precipitation,the results of inversion was analyzed optimal-visually,finding nice consistency exists between inversion results and practical soil moisture.The proper feature spaces would be classified on the drought grade,with soil moisture,precipitation and yearbooks of natural disasters,availability of the mathematics model could be seen.At last,the study found things as follows:(1)A11 of the three vegetation indices had close relationships with surface temperature.The drought model constructed on NDVI-LST feature spaces could also be applied to other two vegetation indices selected in the paper;as to the intercepts of dry and wet sides,EVI-LST constructed an effective feature space,generally conforming to theoretical effect,DNVI-LST less,RVI-LST least;three feature spaces produced consistent intercept changes,showing there was regularity,caused by vegetation covering,in fluctuation of maximum temperature and minimum temperature.(2)Comparison was conducted on the correlations between soil moisture and three indices of TVDI,TEDI and TRDI,the result showed that drought index could present soil moisture,and there was negative correlation between them,moreover,TEDI had highest correlation with soil moisture,also with highest proportion of passing the test,TVDI came second and TRDI last.(3)Three temperature vegetation drought index and precipitation on the space all have good consistency,but in this three kinds of drought index and precipitation had no obvious correlation.(4)According to feature space comparison and correlation analysis between drought and the two indices of soil moisture and precipitation,the optimal TEDI for Liaoning was chosen to classify the drought in the province,then effect of the drought grade was tested on the basis of disasters yearbooks and meteorological data from 2013 to 2014,showing TEDI could accurately reflect the space-and-time distribution in Liaoning,and that the TVDI based on EVI-LST could be used to monitor the agricultural drought in areas of Liaoning.(5)Vulnerability of TEDI and LST,NDVI were evaluated,showing TEDI correlated with LST strongly and positively,but correlated with NDVI weakly and negatively,indicating vegetation indices could better present drought change than surface temperature.