Abstract: Capturing a fine-scale population spatial distribution, gridded data is of great significance for supporting research on geography, resources, environment, ecology, disasters, and sustainable development. On the basis of urban-rural partition and township-level demographic data, this study uses the spatialization method that combines enhanced vegetation index adjusted nighttime light index and land use data to produce the gridded population data of Shandong Province, namely, SDpop2000 and SDpop2010 (100 m resolution). The model verification results showed that 78.0% of the administrative units in 2000 and 78.4% in 2010 had an absolute relative error of less than 20%. When validated against global WorldPop population data products, our dataset demonstrated a correlation coefficient and a %RMSE of 0.932 and 38.42% in 2000 respectively, and 0.948 and 40.46% in 2000.
Keywords: population; spatialization; nighttime light data; land use data; township-level; Shandong Province