Dynamic Monitoring Of Drought In Shandong Province Based On Remote Sensing And Spatial Statistical Analysis

As a large agricultural province,Shandong Province is an important grain production base in China.It has the characteristics of large inter annual variation of precipitation,uneven distribution within the year and frequent drought.Drought disasters caused by water shortage not only affect agricultural production,but also restrict the development of social economy.The accurate dynamic monitoring of regional drought can realize the reasonable allocation of water resources and reduce the agricultural losses caused by drought disasters.There are some defects in the way of monitoring drought based on surface meteorological stations,such as small coverage and few stations,and the sparse and uneven measurement network caused by the high uncertainty and spatial variability of precipitation will often affect the precipitation.Therefore,the traditional drought monitoring is not enough for high-quality real-time dynamic drought monitoring in large area.The birth of satellite remote sensing drought monitoring makes up for the limitations of traditional drought monitoring methods,and a large number of applications have proved the reliability of remote sensing data.Therefore,in view of the poor spatial accuracy of drought monitoring,it is difficult to achieve real-time dynamic drought monitoring.In this paper,taking Shandong Province as the research area,using remote sensing data and meteorological data as data support,taking regional division,Vegetation Drought Stress and abnormal precipitation reduction response lag research as theoretical support,the remote sensing comprehensive drought index based on ordered weighted average method is established,The research idea of drought monitoring and evaluation is to combine remote sensing data with meteorological data,divide the heterogeneous space of drought region based on drought and terrain factors,apply cross spectrum analysis method to explore the lag response time of Vegetation Drought Stress and precipitation anomaly,and construct remote sensing ordered weighted average drought index based on ordered weighted average method,To realize the dynamic monitoring and evaluation of agricultural drought in Shandong Province.The main achievements are as follows:(1)In order to deal with the spatial heterogeneity of climate and topography in Shandong Province,Shandong Province is divided into five arid areas by combining analytic hierarchy process and GIS technology.The comprehensive drought factor layer of Shandong Province was obtained by calculating and processing four drought factors,such as precipitation,soil moisture content,vegetation index(NDVI)and temperature.The comprehensive drought factor layer of Shandong Province is superimposed with the elevation distribution layer of Shandong Province,and five arid areas in Shandong Province are divided into five dry areas,namely,the heavy agricultural drought area in the north of Northwest plain of Shandong Province,the southern part of Northwest Shandong plain and agricultural light dry area of Jiaolai plain,the wetter area of Jiaodong Peninsula,the humid area of hilly hills in South Central Shandong and the plain wet area of southern coastal area of Shandong Province.(2)In view of the response lag of Vegetation Drought Stress and abnormal precipitation reduction,based on the cross spectrum analysis method,the lag response time of Vegetation Drought Stress and abnormal precipitation reduction was calculated from the perspectives of wet period,dry period,wet year and dry year respectively,showing the difference of the results.Firstly,the NDVI time series is reconstructed by savitzky-Golay filter.Secondly,according to the spatial and temporal characteristics of precipitation in Shandong Province,one year is divided into two stages: wet period and drought period.Finally,the NDVI precipitation response time of dry and wet periods is calculated by using the cross spectrum analysis method of NDVI and precipitation,and the spatial distribution of vegetation lag response time is drawn.The results show that the correlation coefficient of NDVI and precipitation in wet period of Shandong Province is higher than that in dry period(average value is 0.40);The average time lag of drought is 11-15.5 days,and that of wet period is 15-21.5d,and the lag time of wet period is higher than that of drought period,and the lag time of mountain and hill is lower than that of plain area;In dry and dry years,the lag time of Northwest plain of Lu is shorter,while in wet year,the lag time of mountainous areas in southern Shandong and Middle Shandong is shorter,and it is found that the lag time of vegetation response is 0.16 a behind the annual time series compared with precipitation.(3)In view of the problem of real-time dynamic drought monitoring in Shandong Province,the ordered weighted average drought index of remote sensing is established to study the dynamic drought monitoring in Shandong Province.Compared with historical events in time and space,it has strong reliability.The integrated drought index OWA-IDI(Integrated Drought Index)was established by coupling the improved analytic hierarchy process(AGA-AHP),mobile window method and ordered weighted average method(OWA)based on accelerated genetic algorithm,and analyzed the drought process in Shandong Province from 1982 to 2014.The results show that OWA-IDI has a significant correlation with SPI;In the time series,the drought period in history was found,which was consistent with the drought events recorded in the bulletin of China on water and drought disasters;In space,the distribution of regional drought assessment in Shandong Province in 2004 is consistent with the distribution of precipitation data in Shandong statistical yearbook;The 5-d model is compared with the monthly model in time,and the more detailed drought monitoring information of short time series is obtained.In conclusion,in Shandong Province,the similar disaster environment areas such as geography,topography and meteorological conditions are divided into five arid areas,and the heterogeneity of space is described;the relationship between vegetation and precipitation is explored,and the delay time of drought stress caused by the decrease of rainfall is calculated to clarify the interaction relationship between drought indicators;the dynamic drought monitoring index is established.The results show that the combination of modern science and technology with agricultural production can realize real-time dynamic drought monitoring in Shandong Province,in order to provide scientific basis and theoretical support for decision makers to formulate drought prevention and drought prevention measures,and reduce agricultural losses caused by drought disasters.