Population Spatialization in Beijing City Based on Machine Learning and Multisource Remote Sensing Data

Miao He,Ning Li,Yongming Xu

doi:10.3390/rs12121910

Abstract

Remote sensing data have been widely used in research on population spatialization. Previous studies have generally divided study areas into several sub-areas with similar features by artificial or clustering algorithms and then developed models for these sub-areas separately using statistical methods. These approaches have drawbacks due to their subjectivity and uncertainty. In this paper, we present a study of population spatialization in Beijing City, China based on multisource remote sensing data and town-level population census data. Six predictive algorithms were compared for estimating population using the spatial variables derived from The National Polar-Orbiting Partnership/ Visible Infrared Imaging Radiometer Suite (NPP/VIIRS) night-time light and other remote sensing data. Random forest achieved the highest accuracy and therefore was employed for population spatialization. Feature selection was performed to determine the optimal variable combinations for population modeling by random forest. Cross-validation results indicated that the developed model achieved a mean absolute error (MAE) of 2129.52 people/km2 and a R2 of 0.63. The gridded population density in Beijing at a spatial resolution of 500 m produced by the random forest model was also adjusted to be consistent with the census population at the town scale. By comparison with Google Earth high-resolution images, the remotely-sensed population was qualitatively validated at the intra-town scale. Validation results indicated that remotely sensed results can effectively depict the spatial distribution of population within town-level districts. This study provides a valuable reference for urban planning, public health and disaster prevention in Beijing, and a reference for population mapping in other cities.

Highlights

In recent decades, the world has experienced rapid and widespread urbanization
13 spatial variables were derived from National Polar-Orbiting Partnership (NPP)/Visible Infrared Imaging Radiometer Suite (VIIRS) Night-time light (NTL), ASTER GDEM, Landsat/TM, OpenSreetMap road network and administrative district data, including night-time light (NTL), Built-up coverage (BC), water coverage (WC), forest coverage (FC), bare land coverage (BLC), industrial and mining coverage (IMC), grassland coverage (GC), cropland coverage (CC), land surface temperature (LST), altitude (ALT), flat area coverage (FAC), road network density (RND) and urban function type (UFT)
Traditional linear regression and five machine learning algorithms were introduced for modeling at the town level: multiple linear regression (MLR), least absolute shrinkage and selection operator (LASSO), Bayesian regularized neural networks (BRNN), support vector machines with radial basis function kernel (SVM), random forest (RF) and extreme gradient boosting (XGBoost)

Summary

Introduction

The world has experienced rapid and widespread urbanization. In 2018, 55% of the worlds’ population lived in urban areas [1] and this increase in urbanization has led to economic development and social progress [2,3]. Cities with large and concentrated populations are characterized by heterogeneous and convoluted spatial patterns, which provide challenges for planning and management [6]. Detailed information on socioeconomic and natural characteristics are critical for effective urban management and decision making. Most of the information required for urban management and planning is spatial in nature [7]. It is difficult to acquire detailed spatial information by traditional surveys. Remote sensing, which provides spatially continuous observation across large areas is a good source of complementary data to traditional surveys and improves our understanding of cities [8]. Remote sensing has been widely used in various urban applications, from monitoring urban growth to predicting sustainable development [9,10]

Methods

Results

Conclusion