Understanding spatiotemporal evolution of the surface urban heat island in the Bangkok metropolitan region from 2000 to 2020 using enhanced land surface temperature

Abstract

The urbanization process has significantly intensified surface urban heat island (SUHI) effects in the Bangkok Metropolitan Region (BMR). However, understanding the evolution of the urban thermal environment is challenging due to the difficulty in obtaining consistent remote sensing data of the cloud-prone landscape in the BMR. In this study, the data fusion algorithm was utilized to fill cloud-induced data gap and create high spatiotemporal-resolution data by blending Landsat and MODIS remote sensing images. The fused data was used to retrieve land surface temperature (LST) for winter months from 2000 to 2020. The spatiotemporal variations in SUHI were then captured using spatial cluster analysis. Finally, gradient analysis and geographically weighted regression (GWR) were employed to analyse the effects of land cover composition on LST. The SUHI intensity in winter increased from 4.40 °C in 2000 to 5.76 °C in 2020. The areal percentage of SUHI hot spots increased from 24.86% to 29.13%. The gradient analysis results indicated that vegetation with a density higher than 0.3 had a significant effect on LST compared to low-density areas. The woody lands were more effective in lowering LST than cultivated lands. These results provide useful information for developing heat mitigation strategies in the metropolitan regions.

Keywords:

• Spatiotemporal data fusion
• land surface temperature
• surface urban heat island
• sustainable development goals
• megacity

Disclosure statement

No potential conflict of interest was reported by the authors.

Data availability statement

The datasets used in this study is available from the corresponding author (Linlin Lu) upon reasonable request.

Additional information

Funding

The authors are grateful to U.S. Geological Survey for providing Landsat dataset and NASA's Earth Science Data Systems (ESDS) program for providing MODIS products. The authors also would like to express their appreciation to the anonymous reviewers and editors for their helpful comments and suggestions. This research was funded by the Strategic Priority Research Program of the Chinese Academy of Sciences (grant number XDA19030101); the Director Fund of the International Research Center of Big Data for Sustainable Development Goals (grant number CBAS2022DF016); and the National Natural Science Foundation of China (grant number 42071321).