Estimating and Interpreting Fine-Scale Gridded Population Using Random Forest Regression and Multisource Data

Yun Zhou,Zhenyu Peng,Mingguo Ma,Kaifang Shi

doi:10.3390/ijgi9060369

Yun Zhou, Zhenyu Peng + Show 2 more

Open Access

https://doi.org/10.3390/ijgi9060369

Copy DOI

Abstract

Gridded population results at a fine resolution are important for optimizing the allocation of resources and researching population migration. For example, the data are crucial for epidemic control and natural disaster relief. In this study, the random forest model was applied to multisource data to estimate the population distribution in impervious areas at a 30 m spatial resolution in Chongqing, Southwest China. The community population data from the Chinese government were used to validate the estimation accuracy. Compared with the other regression techniques, the random forest regression method produced more accurate results (R2 = 0.7469, RMSE = 2785.04 and p < 0.01). The points of interest (POIs) data played a more important role in the population estimation than the nighttime light images and natural topographical data, particularly in urban settings. Our results support the wide application of our method in mapping densely populated cities in China and other countries with similar characteristics.

Highlights

Gridded population data can reflect the explicit size and detailed distribution of the population
As some of the most popular social sensing data, points of interest (POIs) data are helpful for getting high accuracy gridded population results, especially after the kernel density estimation was used to generate various kinds of POI surfaces [27]
With the help of the feature importance analysis and the PDP, the results indicate that POI data had a more crucial effect on population mapping than the remote sensing data, like the Luojia1-01 nigh-time light (NTL) images, in cities [26]

Summary

Introduction

Gridded population data can reflect the explicit size and detailed distribution of the population. They are indispensable for scientific research and policy making, such as in city development planning [1], disaster prevention [2], public health including novel coronavirus disease 2019 (COVID-19) [3] and environmental conservation [4]. The high resolution population distribution census data linked with small areas were restricted [9,10]. Thanks to the development of computer science and new models, there have been some studies that have used novel methods to obtain high spatial resolution population data without eroding privacy

Objectives

Results

Discussion

Conclusion