Browse by author
Lookup NU author(s): Professor Raj Ranjan
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND).
© 2024High-resolution spatial distribution maps of GDP are essential for accurately analyzing economic development, industrial layout, and urbanization processes. However, the currently accessible GDP gridded datasets are limited in number and resolution. Furthermore, high-resolution GDP mapping remains a challenge due to the complex sectoral structure of GDP, which encompasses agriculture, industry, and services. Meanwhile, multi-source data with high spatial resolution can effectively reflect the level of regional economic development. Therefore, we propose a multi-scale fusion residual network (Res-FuseNet) designed to estimate the GDP grid density by integrating remote sensing and POI data. Specifically, Res-FuseNet extracts multi-scale features of remote sensing and POI data relevant to different sectors. It constructs a joint representation of multi-source data through a fusion mechanism and accurately estimates GDP grid density for three sectors using residual connections. Subsequently, the high-resolution GDP grid data are obtained by correcting and overlaying grid density for each sector using county-level statistical GDP data. The 100-meter gridded GDP map of the urban agglomeration in the middle reaches of the Yangtze River in 2020 was successfully generated using this method. The experimental results confirm that Res-FuseNet outperforms machine learning models and baseline model significantly in training across different sectors and at the town-level. The R2 values for the three sectors are 0.69, 0.91, and 0.99, respectively, while the town-level evaluation results also exhibit high accuracy (R2=0.75). Res-FuseNet provides an innovative high-resolution mapping method, and the generated high-resolution GDP grid data reveal the distribution characteristics of different sector structures and fine-scale economic disparities within cities, offering robust support for sustainable development.
Author(s): Wu N, Yan J, Liang D, Sun Z, Ranjan R, Li J
Publication type: Article
Publication status: Published
Journal: International Journal of Applied Earth Observation and Geoinformation
Year: 2024
Volume: 129
Online publication date: 09/04/2024
Acceptance date: 01/04/2024
Date deposited: 17/04/2024
ISSN (print): 1569-8432
ISSN (electronic): 1872-826X
Publisher: Elsevier B.V.
URL: https://doi.org/10.1016/j.jag.2024.103812
DOI: 10.1016/j.jag.2024.103812
Data Access Statement: Data will be made available on request.
Altmetrics provided by Altmetric
