Identification Of Poverty Based On Nighttime Light Remote Sensing Date:a Case Study On Contiguous Special Poverty Stricken Areas In Liupan Mountains

Poverty is a worldwide problem faced by mankind.In particular,for large agricultural countries and a large number of poor people in China,the task of comprehensively eradicating poverty in 2020 is extremely difficult.Since 2013,when President Xi Jinping first put forward "precision poverty alleviation",precision identification of poverty has become the most important in the process of poverty alleviation,and poverty identification has become a hot topic in academia.Although statistical data and questionnaire interviews are used to identify poverty,it can help small scale poverty identification,but there are different statistical methods and difficulties in data acquisition.Because the light image reflects the different socioeconomic level of each county and the difference leads to the difference of poverty degree between counties in 2012 years later,DMSP-OLS images and NPP-VIIRS images were respectively applied to poverty identification and a large number of empirical studies have shown the accuracy and timeliness of recognition.In the process of precision poverty alleviation,the problem that traditional data statistic aperture is not unified and nighttime light data to identify poverty is studied in short time usually exist.The article takes the Liupan Mountain contiguous destitute area in China's old revolutionary base areas,ethnic minority settlements,and poverty-stricken areas as examples,Multiple Poverty Index statistical(MPIstatistical)and Average Light Index(ALI)were constructed.The multidimensional poverty index is constructed by using social economic statistics data and basic geographic data to construct a multidimensional poverty index system and applying an objective entropy method to assign index weights and generate a grey correlation model.The MPIstatistical that can truly reflect the level of poverty in each county was constructed by using social economic statistics data and basic geographic data to construct a multidimensional poverty index system and applying an objective entropy method to assign index weights and generate a grey correlation model.The ALI that can objectively reflect the socioeconomic level of each county was generated by using the corrected nighttime light image to calculate the average light value of each county.Based on the principle of identification,a multidimensional poverty estimated index(MPIestimated)was generated through a poverty estimation model.The fitting function of MPIstatistical and ALI creates the model.The results show that:(1)The accuracy of poverty results based on nighttime light data was higher,which can reflect the real poverty degree of the region,and the relative error range between 3.14% and 3.52%.(2)The MPIestimated average value of the contiguous special poverty area respectively are 0.346,0.353,0.353,0.357 and 0.358 in many years.The level of poverty has been reduced year by year.Between 2000 and 2012,there were 39-46 counties with extremely poor conditions and 20-21 counties with highly poor ones.(3)The Moran's I index from 2000 to 2015 respectively were 0.49,0.45,0.47,0.49 and 0.43 indicating that the poverty level in 78 counties has obvious agglomeration.(4)The pattern of poverty is presented with the spatial evolution trend of “relatively less poverty in the eastern and western regions and relatively heavier poverty in the north and south”from 2000 to 2015.