The Random Forest-Based Method of Fine-Resolution Population Spatialization by Using the International Space Station Nighttime Photography and Social Sensing Data

Kangning Li,Yunhao Chen,Ying Li

doi:10.3390/rs10101650

Abstract

Despite the importance of high-resolution population distribution in urban planning, disaster prevention and response, region economic development, and improvement of urban habitant environment, traditional urban investigations mainly focused on large-scale population spatialization by using coarse-resolution nighttime light (NTL) while few efforts were made to fine-resolution population mapping. To address problems of generating small-scale population distribution, this paper proposed a method based on the Random Forest Regression model to spatialize a 25 m population from the International Space Station (ISS) photography and urban function zones generated from social sensing data—point-of-interest (POI). There were three main steps, namely HSL (hue saturation lightness) transformation and saturation calibration of ISS, generating functional-zone maps based on point-of-interest, and spatializing population based on the Random Forest model. After accuracy assessments by comparing with WorldPop, the proposed method was validated as a qualified method to generate fine-resolution population spatial maps. In the discussion, this paper suggested that without help of auxiliary data, NTL cannot be directly employed as a population indicator at small scale. The Variable Importance Measure of the RF model confirmed the correlation between features and population and further demonstrated that urban functions performed better than LULC (Land Use and Land Cover) in small-scale population mapping. Urban height was also shown to improve the performance of population disaggregation due to its compensation of building volume. To sum up, this proposed method showed great potential to disaggregate fine-resolution population and other urban socio-economic attributes.

Highlights

Population is closely associated with a wide range of developmental and environmental issues such as unbalanced region growth, urban planning, hazard response, shortage of water resources, severe traffic congestion, and carbon-induced air pollution [1,2], especially in some international metropolises like Beijing and Shanghai
The architectural composition data indicated detailed spatial distribution of urban landscape on the building level. This proposed method was divided into three main steps: HSL transformation and saturation calibration of International Space Station (ISS) photography, generation of urban functional zones based on point-of-interest, and population spatialization based on the Random Forest Regression
It was proved that HSL transformation was qualified for changing photography into nighttime lights

Summary

Introduction

Population is closely associated with a wide range of developmental and environmental issues such as unbalanced region growth, urban planning, hazard response, shortage of water resources, severe traffic congestion, and carbon-induced air pollution [1,2], especially in some international metropolises like Beijing and Shanghai. In order to meet requirements of practical application in urban planning and disaster emergencies, population data need to be linked to geographical distribution and high resolution [3]. These criteria are not satisfied by traditional population censuses. Satellite imageries are widely employed at relatively low cost to compensate for the lack of spatial information of census data [4,5]. Multi-temporal DMSP/OLS and VIIRS DNB with good resolution and quality allow for globally consistent research of socioeconomic dynamics like urbanization, population, and GDP (gross domestic product) [6]

Methods

Results

Conclusion