Topographic Gradient Effect Of Population Distribution Changes In Karst Poverty-stricken Counties In Southern Chin

China is the country with the largest regional differences in physical geography,population resources,economic and social development in the world.The complete victory in fighting against poverty has eliminated absolute poverty,but the topographical conditions and population problems that hinder the development of outof-poverty mountainous areas with complex terrain and fragile ecology still exist,and absolute poverty has turned to relative poverty.The research on regional poverty is often carried out directly for typical regions or parts from them,lacking joint research and comparative analysis with similar regions;the research on population spatialization generally selects plain urban agglomeration as the research object,the applicability of existing representative datasets in special mountainous areas needs to be systematically considered,and the basic research on poverty geography needs to be more developed.In this paper,the core of the study area is out-of-poverty counties in the three special difficult mountainous areas on the concentrated and contiguous Karst landforms in Southern China,establishes a refined spatial distribution dataset of population suitable for Karst mountain areas,and classifies the commonly used topographical factor of the study area using the topographical factor classification based on the feature spectrum of landform morphology,discuss the terrain gradient effect of population distribution changes from special mountainous areas at different scale units,focus on a class of out-of-poverty counties including some the key counties of national rural revitalization,and provide policy suggestions for promoting sustainable development of regional rural revitalization for reference.The main conclusions are as follows:(1)Qualitative and quantitative evaluation were conducted on three representative population spatial distribution datasets for World Pop,Land Scan and GWP v4 within the study area,and it was found that Land Scan had the highest overall accuracy.(2)The slope spectrum classification method based on merging by Quantile is suitable for grid scale classification of elevation,relief amplitude and slope;Compared with the common threshold division method,this classification results have a more balanced number of levels,better spatial continuity,and more comprehensive terrain gradients that can be effectively expressed.By using this classified method to grid scale classification of three types of topographical factors,a number of out-of-poverty counties with relatively poor topographical conditions are obtained.(3)On the elevation gradient,the elevation space suitable for population distribution of the study area is not unique;on the gradient of relief amplitude and slope,the population density of the study area as a whole and different county attribute areas decreased in a power function,but the population density of the key counties in almost all relief amplitude gradient and slope gradient was higher than that of overall out-ofpoverty counties and the whole study area in the same period,and the gap were magnified in higher levels.Based on this gap,two types of out-of-poverty counties with contradictory population density of relief amplitude and slope spectrum classification are obtained.