Improving population mapping using Luojia 1-01 nighttime light image and location-based social media data

Area of Science:

• Geographic Information Systems (GIS)
• Remote Sensing
• Socioeconomic Studies

Background:

• Fine-resolution population mapping is crucial for urban planning, public health, and disaster management.
• Existing methods struggle with the heterogeneity of population distribution relative to urban facilities.
• A quantitative relationship between population density and facility popularity/distribution needs further exploration.

Purpose of the Study:

• To develop a novel method for fine-scale population mapping.
• To quantify the relationship between population concentration and the distribution/popularity of urban facilities.
• To improve population mapping accuracy in diverse geographic areas.

Main Methods:

• Utilized Luojia 1-01 nighttime light imagery, points of interest (POI), and social media check-in data.
• Developed a grid-based attraction degree (AD) model to assess population concentration potential.
• Extracted 16 attraction indexes using kernel density estimation and random forest modeling.

Main Results:

• Achieved high accuracy in fine-scale population mapping in Zhejiang, China (R² = 0.75 and 0.58 compared to demographic and WorldPop data).
• Identified optimal parameters for check-in data: 650m search distance and 19:00-08:00 acquisition time.
• Demonstrated superior performance over previous methods in rural and suburban areas with low data density.

Conclusions:

• The proposed method significantly reduces mapping errors caused by heterogeneity.
• This approach shows great potential for accurate fine-scale population mapping.
• The integration of diverse data sources enhances population distribution analysis.