Study Of Comprehensive Drought Monitoring Model Of Yunnan Province Based On Multi-source Remote Sensing Data

Drought is a common and extremely harmful global natural disaster,which has a significant impact on human society and ecosystems.In the context of global warming,drought events may become more severe and frequent.Therefore,it is of great significance to carry out research on drought monitoring techniques and methods,to understand the drought situation in a timely manner,to take effective measures to resist drought,and to scientifically guide agricultural production.The drought index of meteorological or remote sensing data is mainly focused on a single drought response factor,which cannot better express the comprehensive information of drought.TRMM precipitation data from downscaling to 1 km×1 km can not only make up for the error caused by meteorological precipitation interpolation,but also meet the regional scale drought monitoring research.Yunnan is a drought-prone area,and agricultural production,people's life,and socio-economic conditions are severely affected.Therefore,Yunnan Province was used as the research area,and remote sensing data such as MODIS,TRMM,and DEM were used as data sources.In addition,the factors of precipitation,vegetation growth,surface temperature and elevation were considered,and the objective information of the samples was retained and the existing knowledge and experience were integrated with the combination of principal component analysis(PCA)and analytic hierarchy process(AHP).The comprehensive drought monitoring model of Yunnan Province was constructed,and the spatial and temporal variation characteristics of drought in Yunnan Province from 2009 to 2018 were analyzed.The main contents and conclusions are as follows:(1)Study on drought monitoring indicators based on MODIS data.Vegetation condition index(VCI)and temperature condition index(TCI)have poor ability to monitor drought in Yunnan Province.VCI has a poor monitoring effect in areas with low coverage,such as in the non-vegetative growing season,and is only suitable for vegetation areas with moderate vegetation coverage and stable growth,however,TCI's monitoring of drought conditions in areas with higher temperatures differs from the actual situation.Neither index is suitable for drought monitoring in Yunnan Province alone.(2)Study on drought monitoring indicators based on TRMM data.The accuracy of TRMM 3B43 data evaluation is high,and it has good applicability in Yunnan Province,and the R2 of monthly,season,annual and single site precipitation data is greater than 0.58.Combined with NDVI data,a geographic weighted regression(GWR)model was established to reduce the accuracy of the data after time scale and single station.Precipitation of Yunnan Province show the spatial distribution of north and east lack while south and west abundant,with an average annual precipitation of 1082 mm,mainly from May to October.Both the precipitation anomaly percentage(Pa)and the tropical rainfall condition index(TRCI)better reflect the drought conditions in Yunnan Province,and the trends of the two trends are basically the same,except that the drought levels monitored by the two indicators are different(3)Comparison of accuracy of drought monitoring indicators.From the standardized precipitation index(SPI),standardized precipitation evapotranspiration index(SPEI)and VCI,TCI,Pa,TRCI,and DEM(elevation)relationships,it can be found that the correlation between TRCI and SPI?SPEI is the strongest,followed by the Pa index,and both precipitation and drought indexes pass the 0.01 significance test.The correlation between TCI,VCI,DEM(elevation)and SPI,SPEI is weak,but both have passed the 0.05 significance test,so the TRCI,VCI,TCI,elevation can be used as the construction index of the comprehensive drought monitoring model.(4)Study on comprehensive drought monitoring model.Considering the effect of elevation factors on drought,the elevation factors of SRTM_DEM data were added to the model construction,and the PCA and AHP were used to determine the weights of the indicators of the drought model(TRCI,VCI,TCI,DEM(elevation)),and obtained comprehensive drought monitoring model(CDI),which is more in line with the actual drought situation in Yunnan Province.CDI has showed good positive correlation with SPI,SPEI,all of which passed the significance test of 0.01.There is a negative correlation between CDI and temperature vegetation dryness index(TVDI),and CDI can well express the ability of remote sensing drought monitoring.The R of CDI and soil moisture is between 0.39-0.73,which can well reflect the soil moisture status.(5)Drought analysis in Yunnan Province.The characteristics of the spatiotemporal changes of drought conditions from 2009 to 2018 are basically the same.Droughts occurred in most parts of the province,and the rates of droughts were relatively high in January to March and November to December.The average drought frequency in Yunnan Province in the past ten years was 29%,and the drought frequency in most regions was mainly distributed between 15-45%,and there were four obvious droughtprone centers.Of the average monthly drought coverage in the past 10 years,the area covered by January was the largest,and the range of drought from 2009 to 2014 was wider than that from 2015 to 2018.Among the 25 stations in Yunnan Province from 2009 to 2018,except for October,the drought in most of the stations in other months was weakening.