Coastal Thermal Gradients and Environmental Justice in Global South Megacities: A Multi-Climate LightGBM–SHAP Analysis of Istanbul, Lima, Lagos, and Kolkata - AI-JISR PDF / Print
AI-Driven Journal of Interdisciplinary Social Research (AI-JISR)
e-ISSN: 3150-0951 • DOI: 10.5281/zenodo.24224228
Volume 1, Issue 2 (2026)
Pages: 126-159
OPEN ACCESS • CC-BY 4.0

Coastal Thermal Gradients and Environmental Justice in Global South Megacities: A Multi-Climate LightGBM–SHAP Analysis of Istanbul, Lima, Lagos, and Kolkata

1 Necmettin Erbakan University, Türkiye
Abstract (English)

Coastal proximity is commonly assumed to moderate urban heat, yet whether this effect is consistent across climate settings and whether it reaches residents equitably remain poorly understood. This study examines summer land surface temperature (LST) across four coastal megacities — Istanbul (Mediterranean–oceanic), Lima (arid Pacific), Lagos (humid tropical) and Kolkata (humid subtropical) — using a 1 km × 1 km grid analysis that combines Landsat 8/9 thermal imagery, VIIRS nighttime light data as a socioeconomic proxy, WorldPop population density, and ESA WorldCover land-use classification. Separate LightGBM models are trained for each city under a spatially blocked cross-validation design, and Shapley Additive Explanations (SHAP) are used to identify city-specific feature importance rankings, coastal cooling gradients and non-linear threshold behaviour. The results indicate that the cooling contribution of coastal proximity is neither universal nor equitable. A measurable coastal cooling effect — confined to a 3 km nearshore band — is detected only in Istanbul (SHAP slope: +0.22 °C km⁻¹), while Lima, Lagos and Kolkata show near-zero or negligible coastal SHAP contributions. The population density threshold beyond which vegetation cooling is suppressed ranges from approximately 1,460 persons km⁻² in Kolkata to 7,620 persons km⁻² in Istanbul, confirming that this threshold is city-specific rather than transferable. The socioeconomic gradient in LST diverges markedly across cities: Lima shows the expected pattern (higher-income neighbourhoods 2.46 °C cooler), consistent with coastal thermal privilege concentrated among wealthier Pacific-adjacent districts; Istanbul and Kolkata exhibit a reversal (+3.03 °C and +2.64 °C, respectively), where higher-income areas coincide with dense commercial zones that negate coastal and vegetation cooling advantages. Lagos shows no meaningful gradient. Post-hoc Spatial Error Model correction fails to reduce residual spatial autocorrelation in Istanbul and Lagos, suggesting that these cities’ thermal patterns contain non-linear spatial structure beyond the reach of linear spatial econometric tools. These findings challenge the assumption that coastal cities are inherently more thermally equitable and demonstrate that the socioeconomic distribution of heat exposure is shaped by the intersection of climate, urban morphology and waterfront development patterns. The results carry direct implications for green infrastructure planning and environmental justice policy in Global South cities: density-corrected cooling thresholds should be estimated locally, coastal location should not be treated as a proxy for thermal relief, and heat adaptation strategies must account for the possibility that waterfront development may concentrate — rather than distribute — coastal cooling benefits.

Keywords: Land Surface Temperature, Urban Heat Island, Environmental Justice, Shap, LightGBM, Coastal Cities, Blue Gentrification, Global South, Spatial Cross-Validation

References

  1. Akiba, T., Sano, S., Yanase, T., Ohta, T., & Koyama, M. (2019). Optuna: A Next-generation Hyperparameter Optimization Framework. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2623–2631. https://doi.org/10.1145/3292500.3330701
  2. Anselin, L. (1988). Spatial Econometrics: Methods and Models (Vol. 4). Springer Netherlands. https://doi.org/10.1007/978-94-015-7799-1
  3. Ascencio, E. J., Barja, A., Benmarhnia, T., & Carrasco-Escobar, G. (2023). Disproportionate exposure to surface-urban heat islands across vulnerable populations in Lima city, Peru. Environmental Research Letters, 18(7), 074001. https://doi.org/10.1088/1748-9326/acdca9
  4. Avni, N., & Moser, S. (2026). Mapping waterfront (re)developments in Southeast Asia: Speculation, entrepreneurial urbanism, and blue gentrification. Cities, 170, 106714. https://doi.org/10.1016/j.cities.2025.106714
  5. Bera, D., Das Chatterjee, N., Mumtaz, F., Dinda, S., Ghosh, S., Zhao, N., Bera, S., & Tariq, A. (2022). Integrated Influencing Mechanism of Potential Drivers on Seasonal Variability of LST in Kolkata Municipal Corporation, India. Land, 11(9), 1461. https://doi.org/10.3390/land11091461
  6. Blythe, J. L., Gill, D. A., Claudet, J., Bennett, N. J., Gurney, G. G., Baggio, J. A., Ban, N. C., Bernard, M. L., Brun, V., Darling, E. S., Di Franco, A., Epstein, G., Franks, P., Horan, R., Jupiter, S. D., Lau, J., Lazzari, N., Mahajan, S. L., Mangubhai, S., … Zafra-Calvo, N. (2023). Blue justice: A review of emerging scholarship and resistance movements. Cambridge Prisms: Coastal Futures, 1, e15. https://doi.org/10.1017/cft.2023.4
  7. Chakraborty, T., Hsu, A., Manya, D., & Sheriff, G. (2019). Disproportionately higher exposure to urban heat in lower-income neighborhoods: A multi-city perspective. Environmental Research Letters, 14(10), 105003. https://doi.org/10.1088/1748-9326/ab3b99
  8. Chen, Z., Ye, J., Liu, Y., Hu, T., & Dong, J. (2025). Global assessment in the effect of urban built-up environment on cooling efficiency of green spaces. Communications Earth & Environment, 6(1), 968. https://doi.org/10.1038/s43247-025-02925-7
  9. Cheng, X., Liu, Y., Dong, J., Corcoran, J., & Peng, J. (2023). Opposite climate impacts on urban green spaces’ cooling efficiency around their coverage change thresholds in major African cities. Sustainable Cities and Society, 88, 104254. https://doi.org/10.1016/j.scs.2022.104254
  10. Deppner, J., & Cajias, M. (2024). Accounting for Spatial Autocorrelation in Algorithm-Driven Hedonic Models: A Spatial Cross-Validation Approach. The Journal of Real Estate Finance and Economics, 68(2), 235–273. https://doi.org/10.1007/s11146-022-09915-y
  11. Dissanayake, D., Morimoto, T., Murayama, Y., Ranagalage, M., & Handayani, H. H. (2018). Impact of Urban Surface Characteristics and Socio-Economic Variables on the Spatial Variation of Land Surface Temperature in Lagos City, Nigeria. Sustainability, 11(1), 25. https://doi.org/10.3390/su11010025
  12. Georganos, S., Grippa, T., Niang Gadiaga, A., Linard, C., Lennert, M., Vanhuysse, S., Mboga, N., Wolff, E., & Kalogirou, S. (2021). Geographical random forests: A spatial extension of the random forest algorithm to address spatial heterogeneity in remote sensing and population modelling. Geocarto International, 36(2), 121–136. https://doi.org/10.1080/10106049.2019.1595177
  13. Guo, Z., Esperon-Rodriguez, M., Davin, E., Huang, H., Chen, B., Hejazi, M., Wu, J., Wang, J., Ge, Y., Song, G., Zhao, Y., Feng, K., Lin, C., Gong, P., & Zhou, Y. (2026). Global urban vegetation exhibits divergent thermal effects: From cooling to warming as aridity increases. Science Advances, 12(1), eaea9165. https://doi.org/10.1126/sciadv.aea9165
  14. Ibsen, P. C., Borowy, D., Dell, T., Greydanus, H., Gupta, N., Hondula, D. M., Meixner, T., Santelmann, M. V., Shiflett, S. A., Sukop, M. C., Swan, C. M., Talal, M. L., Valencia, M., Wright, M. K., & Jenerette, G. D. (2021). Greater aridity increases the magnitude of urban nighttime vegetation-derived air cooling. Environmental Research Letters, 16(3), 034011. https://doi.org/10.1088/1748-9326/abdf8a
  15. Ke, G., Meng, Q., Finley, T., Wang, T., Chen, W., Ma, W., Ye, Q., & Liu, T.-Y. (2017). LightGBM: a highly efficient gradient boosting decision tree. Proceedings of the 31st International Conference on Neural Information Processing Systems, NIPS’17, 3149–3157.
  16. Kopczewska, K. (2022). Spatial machine learning: New opportunities for regional science. The Annals of Regional Science, 68(3), 713–755. https://doi.org/10.1007/s00168-021-01101-x
  17. Li, T., Li, Z., & Yu, Y. (2026). Regulatory Effects of Urban Vegetation and Urban Forests on the Thermal Environment of Megacities: A Comparative Study Based on Explainable Machine Learning. Forests, 17(3), 296. https://doi.org/10.3390/f17030296
  18. Liu, X., Kounadi, O., & Zurita-Milla, R. (2022). Incorporating Spatial Autocorrelation in Machine Learning Models Using Spatial Lag and Eigenvector Spatial Filtering Features. ISPRS International Journal of Geo-Information, 11(4), 242. https://doi.org/10.3390/ijgi11040242
  19. Liu, Y., Huang, X., Yang, Q., & Cao, Y. (2021). The turning point between urban vegetation and artificial surfaces for their competitive effect on land surface temperature. Journal of Cleaner Production, 292, 126034. https://doi.org/10.1016/j.jclepro.2021.126034
  20. MacBride, C., Davies, V., & Lee, D. (2025). A spatial autoregressive random forest algorithm for small-area spatial prediction. The Annals of Applied Statistics, 19(1). https://doi.org/10.1214/24-AOAS1969
  21. Manoli, G., Fatichi, S., Schläpfer, M., Yu, K., Crowther, T. W., Meili, N., Burlando, P., Katul, G. G., & Bou-Zeid, E. (2019). Magnitude of urban heat islands largely explained by climate and population. Nature, 573(7772), 55–60. https://doi.org/10.1038/s41586-019-1512-9
  22. Meyer, H., & Pebesma, E. (2021). Predicting into unknown space? Estimating the area of applicability of spatial prediction models. Methods in Ecology and Evolution, 12(9), 1620–1633. https://doi.org/10.1111/2041-210X.13650
  23. Meyer, H., & Pebesma, E. (2022). Machine learning-based global maps of ecological variables and the challenge of assessing them. Nature Communications, 13(1), 2208. https://doi.org/10.1038/s41467-022-29838-9
  24. Naserikia, M., Hart, M. A., Nazarian, N., & Bechtel, B. (2022). Background climate modulates the impact of land cover on urban surface temperature. Scientific Reports, 12(1), 15433. https://doi.org/10.1038/s41598-022-19431-x
  25. Oke, T. R. (1987). *Boundary Layer Climates* (2nd ed.). London: Methuen.
  26. Ploton, P., Mortier, F., Réjou-Méchain, M., Barbier, N., Picard, N., Rossi, V., Dormann, C., Cornu, G., Viennois, G., Bayol, N., Lyapustin, A., Gourlet-Fleury, S., & Pélissier, R. (2020). Spatial validation reveals poor predictive performance of large-scale ecological mapping models. Nature Communications, 11(1), 4540. https://doi.org/10.1038/s41467-020-18321-y
  27. Roberts, D. R., Bahn, V., Ciuti, S., Boyce, M. S., Elith, J., Guillera‐Arroita, G., Hauenstein, S., Lahoz‐Monfort, J. J., Schröder, B., Thuiller, W., Warton, D. I., Wintle, B. A., Hartig, F., & Dormann, C. F. (2017). Cross‐validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography, 40(8), 913–929. https://doi.org/10.1111/ecog.02881
  28. Roth, M. (2007). Review of urban climate research in (sub)tropical regions. International Journal of Climatology, 27(14), 1859–1873. https://doi.org/10.1002/joc.1591
  29. Ünsal, Ö., Lotfata, A., & Avcı, S. (2023). Exploring the Relationships between Land Surface Temperature and Its Influencing Determinants Using Local Spatial Modeling. Sustainability, 15(15), 11594. https://doi.org/10.3390/su151511594
  30. Valavi, R., Elith, J., Lahoz‐Monfort, J. J., & Guillera‐Arroita, G. (2019). BLOCK CV: An R package for generating spatially or environmentally separated folds for k ‐fold cross‐validation of species distribution models. Methods in Ecology and Evolution, 10(2), 225–232. https://doi.org/10.1111/2041-210X.13107
  31. Yang, Q., Huang, X., Tong, X., Xiao, C., Yang, J., Liu, Y., & Cao, Y. (2022). Global assessment of urban trees’ cooling efficiency based on satellite observations. Environmental Research Letters, 17(3), 034029. https://doi.org/10.1088/1748-9326/ac4c1c
  32. Yang, Q., Huang, X., Yang, J., & Liu, Y. (2021). The relationship between land surface temperature and artificial impervious surface fraction in 682 global cities: Spatiotemporal variations and drivers. Environmental Research Letters, 16(2), 024032. https://doi.org/10.1088/1748-9326/abdaed
  33. Yazar, M., Daloglu Cetinkaya, I., Iban, M. C., & Bilgilioglu, S. S. (2023). The green divide and heat exposure: Urban transformation projects in istanbul. Frontiers in Environmental Science, 11, 1265332. https://doi.org/10.3389/fenvs.2023.1265332
  34. Zhang, M., Zhou, D., Xu, D., Liu, K., Gu, Z., Yang, Y., Zhang, Q., & Chen, L. (2026). How do built environment factors influence urban heat islands across local climate zones? Evidence from an interpretable XGBoost-SHAP model. Sustainable Cities and Society, 136, 107078. https://doi.org/10.1016/j.scs.2025.107078
  35. Ziter, C. D., Pedersen, E. J., Kucharik, C. J., & Turner, M. G. (2019). Scale-dependent interactions between tree canopy cover and impervious surfaces reduce daytime urban heat during summer. Proceedings of the National Academy of Sciences, 116(15), 7575–7580. https://doi.org/10.1073/pnas.1817561116