Research On Integrated Drought Monitoring Model Using MODIS Data

Drought is the most serious natural disaster in China.With the complication of global climate,the problem of drought has become frequent and severe,which has seriously affected the development of our national economy and even threatened the people's normal life.Therefore,research on drought monitoring method based on satellite remote sensing technology,to provide scientific basis for effective measures for drought control and disaster mitigation is of great significance.In this paper,the problems of drought monitoring in remote sensing were studied.Based on the description of drought conditions and the analysis of drought factors such as land surface temperature change and vegetation growth status and remote sensing monitoring mechanism,remote sensing drought index is proposed based on vector cosine similarity measure,and is improved by the use of remote sensing drought factor time series data.The method of drought grade classification based on mutual information for drought index is put forward.Combined with the adaptability of different remote sensing drought indices,a integrated remote sensing drought monitoring model is designed and constructed.Overall,the main research results and contributions of this study are as follows:Based on the analysis of drought characteristics and existing drought index,a drought monitoring method combining by drought factors was proposed.The existing combinatorial methods mainly have ratio,sum value or the way of establishing the feature space,but do not fully consider the applicability of drought factor and the influence of spatio-temporal factors on drought factors.Based on the idea of language machine translation,this study chooses the drought factor suitable for describing drought and establishes the drought state feature vector composed of drought factors.From the point of view of similarity of information,a new method of remote sensing drought index is proposed by measuring the similarity of the extreme state characteristic vector,and the validity and applicability of the mothod was proved.Based on the temporal and spatial characteristics of drought factors,the drought index of remote sensing was improved by analyzing the time series data of drought factors.In this paper,MODIS vegetation index and land surface temperature quality control was put on.Based on the analysis of time series data and the support of high-quality time-series data,the extreme state feature are modified in the spatial and temporal features by establishing the spatial-temporal zoning.Accuracy,sensitivity and spatio-temporal adaptation of the remote sensing drought index is improved.Drought grade classification method for remote sensing drought index is usually to establish by the statistical model of drought index and ground observation data.However,the relationship between drought index and ground data is complex.In order to solve this problem,this paper introduces the idea of information theory,selects mutual information maximum partition scheme by calculating the mutual information of various remote sensing drought indexes and the measured drought-grade data,determines the range of drought index corresponding to each drought level.A new idea of classification of remote sensing drought grade is put forward.Considering the temporal and spatial characteristics of remote sensing drought index,this study improves the method by synthesizing different remote sensing drought indices.Based on the analysis of the temporal and spatial adaptability,different remote sensing drought indices were optimized and classified to the same standard,a Integrated remote sensing drought monitoring model is designed and built,and the accuracy and temporal-spatial adaptability of the remote sensing drought monitoring is further improved.