The Study Of Drought Characteristics Of Henan Province Based On Multi-source And Heterogeneous Data And Model Prediction

Drought is an agrometeorological disaster caused by long-term precipitation with little or no precipitation,strong evaporation,and insufficient water in the soil,leading to the destruction of crop water balance,resulting in crop yield reduction or even failure.With the intensification of global warming,global meteorological disasters occur frequently,increasing climate instability,and global droughts become more and more frequent.Therefore,the study of drought has attracted more and more attention from domestic and foreign scholars.Henan Province is located in the monsoon climate zone and at the junction of the humid zone and the semi-humid zone.In this area,there are Qinling and Huaihe Rivers passing through,resulting in uneven spatial and temporal distribution of precipitation and sensitive climate.So that drought disasters occur more frequently.Based on its climate characteristics,drought research in Henan province is of great significance to climate change,agricultural production and people's life in this region.This study qualitatively analyzes the temporal and spatial variation characteristics of drought in Henan Province by quantitatively studying the different time scales of each drought index and the changes of different regions in Henan Province.On this basis,this paper systematically analyzes the response of each drought index based on remote sensing data and meteorological station data to drought conditions in Henan Province,and uses soil moisture data as an evaluation index to analyze the applicability of each drought index to Henan Province.The advantage of multi-source and heterogeneous data sources in drought research is that they can not only integrate the advantages of remote sensing spatial data and meteorological data,but also reflect the information of the characteristics of various drought factors.The comprehensive drought index model established by integrating the advantages of multi-source and heterogeneous data sources can not only improve the precision of drought monitoring,but also further analyze and evaluate the drought situation.This study has obtained the following results and conclusions:(1)Based on the drought index of remote sensing,it was found that the most drought prone growth period of winter wheat in Henan province was from the spring returning period to the flowering period,while the most drought prone growth period of summer corn was from tasseling period to milk-ripening period.The most drought prone seasons in Henan province are spring and summer.Combined with the advantages of remote sensing monitoring,such as high spatial resolution and large area coverage,the drought analysis of Henan province shows that both vegetation drought index(TVDI)and vegetation water supply index(VSWI)calculated based on MODIS data can reflect the drought situation of Henan province from different perspectives,but TVDI is relatively more suitable for the study of drought in Henan province.TVDI for drought monitoring mainly reflect the crops in different growth periods of drought conditions,Based on the index,it is found that the most prone to dry growth stage of winter wheat is the greening stage-flowering stage,and the most prone to dry stage of summer maize is the tasseling stage-milk stage.VSWI tends to reflect the distribution characteristics of drought in Henan province in different seasons.Based on VSWI,it is found that the seasons most prone to drought in Henan Province are spring and summer.According to the precipitation anomaly percentage(Pa)analysis calculated with the scale reduction data of TRMM,the driest year in Henan province from 1998 to 2018 is 2014,and the climate in this year is extremely unstable,and the time when the whole province of Henan is severe drought is in July and December.In addition,it is feasible for Henan Province to use the geographically weighted regression model to introduce the normalized vegetation index to scale down the TRMM data.(2)Drought monitoring based on meteorological site data found that SPI is more suitable for drought research in Henan Province,and the drought characteristics and trend of different regions in Henan province are quite different.Based on the advantages of high accuracy and strong continuity of meteorological station data,the drought analysis of Henan province shows that the regional applicability of standardized precipitation index(SPI)and standardized precipitation evapotranspiration index(SPEI)in drought research is different.SPI has general applicability,while SPEI has better applicability for relatively dry areas.Comparing the drought index at the 1 month,3 month,12 month and 24 month time scales,the 3 month scale drought index is more suitable for drought research in the region.With 10 years as the time scale,the climate in the first five years of Henan Province is not as stable as the last five years.In terms of space:it shows that the overall drought situation in the west and north of Henan province is decreasing,while the drought situation in the east,central and south of Henan is increasing.(3)Based on the comprehensive drought index obtained from multi-source heterogeneous data(CDI)model integrated the advantages of remote sensing data and meteorological data,In this paper,the correlation analysis of soil moisture is used to prove that the model is reliable for drought research in Henan Province.)There are many models built based on remote sensing and site data,but there are few comprehensive drought models based on multiple parameters and multi-data source information such as vegetation index,temperature and precipitation.Therefore,SPI,TVDI and Pa are selected to build comprehensive drought model(CDI)as one of the research features and innovation points in this paper.Overall,the correlation coefficient of site data is higher than that of remote sensing data,and the comprehensive drought index(CDI)can well reflect the distribution of drought characteristics in Henan province.Microwave remote sensing SMAP L3 soil moisture data has poor applicability for small range drought monitoring.(4)Based on the application of the comprehensive drought index,it is found that the affected area of crops is inversely proportional to the index;Sunshine,precipitation and air humidity are the three factors that have the largest influence on drought weight in Henan province.When analyzing the correlation between the area affected by crops in Henan Province and the comprehensive drought index(CDI).it was found that:the comprehensive drought index and crop is inversely proportional to the affected area.According to the drought classification of CDI,the correlation coefficient R~2 between the affected area of crops and the comprehensive drought index is greater than 0.5,which indicates that drought is one of the main causes of crop disaster in Henan province.Based on principal component analysis,this paper studies the drought-causing factors in Henan Province and finds that sunshine,precipitation and air humidity are the three factors that affect the maximum drought weight in the region.Based on the correlation analysis of CDI with the above larger sensitivity factors,it was found that the correlation coefficient between CDI and sensitivity factors was higher than 0.5(the absolute value of sunshine factor),indicating that the comprehensive drought index model could reflect the information of drought-causing sensitive factors in Henan province and was suitable for the study of drought in Henan province.