Estimating GDP by Fusing Nighttime Light and Land Cover Data
Accurate information on gross domestic product (GDP) is essential for better understanding the dynamics of regional economies and urbanization processes. Satellite based nighttime light datasets can well track GDP in urban areas, however, they are difficult to be used in suburban, rural and sparsely...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2024-12-01
|
| Series: | Canadian Journal of Remote Sensing |
| Online Access: | http://dx.doi.org/10.1080/07038992.2024.2377641 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846095188639875072 |
|---|---|
| author | Nan Xu Shiyi Zhang Shuai Jiang |
| author_facet | Nan Xu Shiyi Zhang Shuai Jiang |
| author_sort | Nan Xu |
| collection | DOAJ |
| description | Accurate information on gross domestic product (GDP) is essential for better understanding the dynamics of regional economies and urbanization processes. Satellite based nighttime light datasets can well track GDP in urban areas, however, they are difficult to be used in suburban, rural and sparsely populated areas. Thus, this study explored the potential of GlobeLand30 and relief degree for improving the ability of VIIRS Nighttime Light data of GDP estimation. Firstly, we calculated the Moran’s Index (Moran’s I) to measure spatial auto-correlation of GDP. At provincial level, Moran’s I Index of GDP is 0.14, Z value is 2.29. While at the prefecture city level, it is 0.11 and 14.54, respectively. Then, we compared the results derived from geographically weighted regression (GWR) and OLS models (i.e., R2, root mean square error, corrected Akaike information criteria and residuals). Both models suggest that land cover information can significantly improve GDP estimation performance, and total nighttime light (TNL) is the most important economic indicator for estimating GDP. The coefficients of the GWR model for TNL at the provincial and prefecture levels are 1.75 and 1.19, respectively, which are significantly larger than the coefficients for other factors such as land cover and terrain undulation. In addition, the GWR model performed better than OLS model in GDP estimation at both provincial and prefecture levels, and prefecture-level models can better depict the spatial variation in detail. In provincial-level models, GWR could account for 93% of economic development, while OLS could only reflect 82%. Likewise, in prefecture-level models, R2 of GWR model improved almost 50% compared with that of OLS model. |
| format | Article |
| id | doaj-art-a20b0d9ad5104185a5d274d9aefc551b |
| institution | Kabale University |
| issn | 1712-7971 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Canadian Journal of Remote Sensing |
| spelling | doaj-art-a20b0d9ad5104185a5d274d9aefc551b2025-01-02T11:34:20ZengTaylor & Francis GroupCanadian Journal of Remote Sensing1712-79712024-12-0150110.1080/07038992.2024.23776412377641Estimating GDP by Fusing Nighttime Light and Land Cover DataNan Xu0Shiyi Zhang1Shuai Jiang2School of Earth Sciences and Engineering, Hohai UniversityCollege of Geography and Remote Sensing, Hohai UniversityState Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan UniversityAccurate information on gross domestic product (GDP) is essential for better understanding the dynamics of regional economies and urbanization processes. Satellite based nighttime light datasets can well track GDP in urban areas, however, they are difficult to be used in suburban, rural and sparsely populated areas. Thus, this study explored the potential of GlobeLand30 and relief degree for improving the ability of VIIRS Nighttime Light data of GDP estimation. Firstly, we calculated the Moran’s Index (Moran’s I) to measure spatial auto-correlation of GDP. At provincial level, Moran’s I Index of GDP is 0.14, Z value is 2.29. While at the prefecture city level, it is 0.11 and 14.54, respectively. Then, we compared the results derived from geographically weighted regression (GWR) and OLS models (i.e., R2, root mean square error, corrected Akaike information criteria and residuals). Both models suggest that land cover information can significantly improve GDP estimation performance, and total nighttime light (TNL) is the most important economic indicator for estimating GDP. The coefficients of the GWR model for TNL at the provincial and prefecture levels are 1.75 and 1.19, respectively, which are significantly larger than the coefficients for other factors such as land cover and terrain undulation. In addition, the GWR model performed better than OLS model in GDP estimation at both provincial and prefecture levels, and prefecture-level models can better depict the spatial variation in detail. In provincial-level models, GWR could account for 93% of economic development, while OLS could only reflect 82%. Likewise, in prefecture-level models, R2 of GWR model improved almost 50% compared with that of OLS model.http://dx.doi.org/10.1080/07038992.2024.2377641 |
| spellingShingle | Nan Xu Shiyi Zhang Shuai Jiang Estimating GDP by Fusing Nighttime Light and Land Cover Data Canadian Journal of Remote Sensing |
| title | Estimating GDP by Fusing Nighttime Light and Land Cover Data |
| title_full | Estimating GDP by Fusing Nighttime Light and Land Cover Data |
| title_fullStr | Estimating GDP by Fusing Nighttime Light and Land Cover Data |
| title_full_unstemmed | Estimating GDP by Fusing Nighttime Light and Land Cover Data |
| title_short | Estimating GDP by Fusing Nighttime Light and Land Cover Data |
| title_sort | estimating gdp by fusing nighttime light and land cover data |
| url | http://dx.doi.org/10.1080/07038992.2024.2377641 |
| work_keys_str_mv | AT nanxu estimatinggdpbyfusingnighttimelightandlandcoverdata AT shiyizhang estimatinggdpbyfusingnighttimelightandlandcoverdata AT shuaijiang estimatinggdpbyfusingnighttimelightandlandcoverdata |