Projecting high resolution population distribution using Local Climate Zones and multi-source big data

Abstract

High precision and easily updatable data on population estimates are critical for urban planning, disaster risk assessment, and public health campaigns. However, traditional administrative census data often faces limitations in spatial resolution, making it difficult for further applications. In this study, we construct a fast and efficient population estimation model using Local Climate Zones (LCZ) products at a fine spatial scale. The population totals are estimated by LCZ units, based on a robust linkage between built LCZ types and population density. Validated by city- and county-level census data in 21 cities in China, the new model exhibits a very good fit with R2 values of 0.77 and 0.73 respectively, which confirms the effectiveness of LCZ-based population estimates. To our knowledge, this is the first study that directly estimates the population via LCZ maps. In addition, it was proven that population density can be used as a new property for LCZ type definition. As LCZ products are convenient to obtain, this work provides a simple, economical, and reliable population estimation method that can complement the traditional census.