Monitoring Methods For Poor Areas Supported By DMSP/OLS Night-time Light Imgery

Elimination of poverty is an important matter of the world. Identification of povertyand analysis of spatial changes is the subject of science and still challenging governmentsand scholars. After opening and reform of China, nation and government starts large-scalepoverty relief and development strategy in rural areas. Without locating the real poorareas and groups in poverty, various poverty relief measures will lose their due effects.Also, it will be very challenging to make efficient development strategies and measures ofpoverty relief without understanding the characteristic of spatial distribution. Astraditional statistical data of social economy is lack of the information of space,time-consuming to be collected, and objectivity hard to be guaranteed, they cannot meetthe demand of large-scale, long-term and dynamic research in areas in poverty. Nightlight is closely associated with the development level of regional economy, populationand technology. DMSP/OLS night-light remote sensing data has been widely applied as akind of new remote sensing data source to the studies of urban expansion, regionaleconomy, population, energy consumption and so on.The study mainly introduces DMSP/OLS data to evaluate the poverty degree ofareas. Different from the direct adoption of total regional strength, pixel gray value andother types of remote sensing data in traditional application, the study defines the conceptof “unit light intensity” and associates DMSP/OLS data with the information of regionalpopulation and construction land.13indexes, classified into two types, are designed todescribe the characteristic of regional light, like “unit light intensity of construction land”and “unit light intensity of population”. Through related analysis on3different types ofunit light intensity indexes in municipal district, country and rural-urban areas in country,as well as gross regional production and average resident income in country, authoridentified the scale effect of different unit light intensity indexes on country economy. Onthis basis and with net income of rural residents as indication of regional poverty degree,author screened out different scales of “unit light intensity of construction land” and “unitpopulation light intensity” as variables with the method of stepwise regression to construct2regression models to measure the poverty degree of areas. Goodness of fits of2models is0.912and0.915respectively.The study sorted out and collected the data of population and construction land incountry, and then evaluated the income level of farmers with2regional poverty models.And then, author compared the spatial scale of areas in relative poverty and centralizedextremely poor areas. Meanwhile, author analyzed the objective poverty status ofidentified national-level poor countries. By comparison, author found that the identifiedconnected areas in extreme poverty basically covered most relatively poor areas.However, actual farmer income levels in national-level poor countries in differentprovinces vary greatly. Further, author analyzed the spatial distribution mode of poorcountries with self-correlation statistics of overall and regional space and the result showsthat the distribution of poor areas still represents obvious spatial assembly in the scale ofcountry. There are comparatively rich assembled country areas in Circum-Bohai SeaEconomic Zone of Beijing-Tianjin-Hebei, Yangtze River Delta Economic Zone and PearlRiver Delta Economic Zone, while the provinces in mid-west of China, such as Sichuan,Yunnan, Guizhou, Hunan, Hubei and Tibet, are obviously involved in relatively poorareas and regional consistency is very high, indicating that they are the areas inwidespread poverty.