Population spatialization is widely used for spatially downscaling census population data to finer-scale. The core idea of modern population spatialization is to establish the association between ancillary data and population at the administrative-unit-level (AU- level) and transfer it to generate the gridded population. However, the statistical char- acteristic of attributes at the pixel-level differs from that at the AU-level, thus leading to prediction bias via the cross-scale modeling (i.e. scale mismatch problem). In addition, integrating multi-source data simply as covariates may underutilize spatial semantics, and lead to incorrect population disaggregation; while neglecting the spatial autocorrelation of population generates excessively heterogeneous population distribution that contra- dicts to real-world situation. To address the scale mismatch in downscaling, this paper proposes a Cross-Scale Feature Construction (CSFC) method. More specifically, by grading pixel-level attributes, we construct the feature vector of pixel grade proportions to narrow the scale differences in feature representation between AU-level and pixel-level. Meanwhile, fine-grained building patch and mobile positioning data are utilized to adjust the population weighting layer generated from POI-density-based regression modeling. Spatial filtering is furtherly adopted to model the spatial autocorrelation effect of popula- tion and reduce the heterogeneity in population caused by pixel-level attribute discretiza- tion. Through the comparison with traditional feature construction method and the ablation experiments, the results demonstrate significant accuracy improvements in popu- lation spatialization and verify the effectiveness of weight correction steps. Furthermore, accuracy comparisons with WorldPop and GPW datasets quantitatively illustrate the advantages of the proposed method in fine-scale population spatialization.

Yuao Mei;Zhipeng Gui;Jinghang Wu;Dehua Peng;Rui Li;Huayi Wu;Zhengyang Wei

School of Remote Sensing and Information Engineering,Wuhan University,Wuhan,China;State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan,China;Collaborative Innovation Center of Geospatial Technology,Wuhan University,Wuhan,China;Institute for Computational and Mathematical Engineering,Stanford University,Stanford,CA,USA

地球空间信息科学学报（英文版）

[1]Yuao Mei;Zhipeng Gui;Jinghang Wu;Dehua Peng;Rui Li;Huayi Wu;Zhengyang Wei-.Population spatialization with pixel-level attribute grading by considering scale mismatch issue in regression modeling)[J].地球空间信息科学学报（英文版）,2022(03):365-382

underutilize,contra,dicts,CSFC,discretiza

Population,spatialization,pixel,level,grading,by,considering,scale,mismatch,issue,regression,modeling,widely,spatially,downscaling,census,population,finer,core,idea,modern,establish,association,between,ancillary,administrative,unit,AU,transfer,gridded,However,statistical,char,acteristic,attributes,differs,from,that,thus,leading,prediction,bias,via,cross,problem,In,addition,integrating,multi,source,simply,covariates,may,semantics,incorrect,disaggregation,neglecting,autocorrelation,generates,excessively,heterogeneous,distribution,real,world,situation,To,address,this,paper,proposes,Cross,Scale,Feature,Construction,method,More,specifically,feature,vector,grade,proportions,narrow,differences,representation,Meanwhile,grained,building,patch,mobile,positioning,are,utilized,adjust,weighting,layer,generated,POI,density,Spatial,filtering,furtherly,adopted,reduce,heterogeneity,caused,Through,traditional,construction,ablation,experiments,results,demonstrate,significant,accuracy,improvements,verify,effectiveness,correction,steps,Furthermore,comparisons,WorldPop,GPW,datasets,quantitatively,illustrate,advantages,proposed

0.529346