PREDICTIVE MODELLING OF URBAN TEMPERATURE UNDER GREEN AREA ENCROACHMENT IN AMAC, ABUJA, NIGERIA
Keywords:
Urban heat island, land surface temperature, green area encroachment, Random Forest, SHAP, scenario modelling, AMAC, AbujaAbstract
Rapid urbanisation in Abuja Municipal Area Council (AMAC), Nigeria, has driven extensive green area encroachment and a rise in Land Surface Temperature (LST). Between 2010 and 2022 the built-up area increased by 43.2 per cent while vegetation declined by 29.6 per cent, and mean LST rose from 31.4°C to 34.1°C, yet predictive tools capable of guiding climate-responsive planning in AMAC remain lacking. This study developed a Random Forest and SHAP predictive model of urban temperature under green area encroachment. Landsat 8 and 9 imagery for 2013, 2018 and 2023 was used to derive LST together with NDVI, NDBI, Green Area, Albedo and Elevation. A Random Forest model was trained on 50,000 pixels using spatial block cross-validation in order to avoid the inflated accuracy that arises from spatial autocorrelation, and SHAP was applied to quantify the contribution of each driver. Future LST for 2035 was projected under three scenarios, namely Business-as-Usual, Moderate Conservation and Aggressive Greening, using CA-Markov simulated land cover. Under spatial block cross-validation the model achieved an R-squared of 0.84 and a root mean square error of 1.72°C, compared with 0.91 and 1.34°C under a conventional random split, confirming the inflation attributable to spatial leakage. SHAP ranked NDVI at 35 per cent, NDBI at 30 per cent and Green Area at 25 per cent as the leading predictors, with every 10 per cent increase in green area reducing LST by 0.8°C on average. Under Business-as-Usual, mean LST is projected to rise by 2.3°C to 36.4°C by 2035, while the Moderate Conservation and Aggressive Greening scenarios would limit warming to 1.1°C and achieve 0.2°C of net cooling respectively. Targeted greening in districts with less than 15 per cent vegetation yielded the highest cooling benefit of 2.1°C to 2.5°C. The study concludes that policy-driven greening can reverse projected warming in AMAC, and that the framework provides a transparent tool for quantifying cooling benefits and prioritising interventions.
References
Adebayo, S., & Akande, T. (2024). Heat-related health burden in West African cities. The Lancet Planetary Health, 8(3), e189–e198. https://doi.org/10.1016/S2542-5196(24)00012-3
Akinyemi, F. O., Adepoju, M. O., & Odufuwa, B. O. (2024). Explainable machine learning for modeling Land Surface Temperature in Lagos Metropolis, Nigeria. Environmental Science and Pollution Research, 31(12), 17842–17859. https://doi.org/10.1007/s11356-024-32517-8
Akinyemi, F., Ogunjobi, K., & Adeola, A. (2025). Machine learning for urban climate projection in Africa. Remote Sensing of Environment, 298, 113842. https://doi.org/10.1016/j.rse.2024.113842
Breiman, L. (2001). Random forests. Machine Learning, 45(1), 5–32. https://doi.org/10.1023/A:1010933404324
Chukwuma, E., Zhang, Y., & Adeyemi, B. (2022). Green space and LST dynamics in tropical cities. Urban Climate, 45, 101254. https://doi.org/10.1016/j.uclim.2022.101254
Federal Capital Development Authority. (2021). Abuja master plan implementation report. FCDA.
Ibrahim, M., & Bello, A. (2023). Spatio-temporal analysis of LST in Abuja using Landsat data. Journal of African Earth Sciences, 201, 104892. https://doi.org/10.1016/j.jafrearsci.2023.104892
Li, X., Zhou, W., & Ouyang, Z. (2024). Advances in RF modeling for urban thermal environment. ISPRS Journal of Photogrammetry and Remote Sensing, 208, 45–61. https://doi.org/10.1016/j.isprsjprs.2024.01.005
Li, W., Zhang, H., & Guo, Q. (2025). Predicting Land Surface Temperature with Random Forest and multi-source remote sensing: A case study of urban agglomerations in China. ISPRS Journal of Photogrammetry and Remote Sensing, 218, 245–261. https://doi.org/10.1016/j.isprsjprs.2024.11.018
Lundberg, S. M., & Lee, S. I. (2017). A unified approach to interpreting model predictions. In I. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, & R. Garnett (Eds.), Advances in neural information processing systems 30(pp. 4765–4774). Curran Associates. https://proceedings.neurips.cc/paper/2017/hash/8a20a8621978632d76c43dfd28b67767-Abstract.html
Mensah, C., & Owusu, G. (2021). Land cover change and UHI in Accra, Ghana. _Remote Sensing Applications: Society and Environment, 23, 100589. https://doi.org/10.1016/j.rsase.2021.100589
Musa, I., Tijjani, B., & Yusuf, H. (2021). Land use change and LST in Abuja, Nigeria. GeoJournal, 86(4), 1567–1582. https://doi.org/10.1007/s10708-020-10189-2
Owoeye, J., & Popoola, A. (2023). Urban expansion and vegetation loss in AMAC, Abuja. Land Use Policy, 128, 106592. https://doi.org/10.1016/j.landusepol.2023.106592
Rao, K., Singh, P., & Kumar, A. (2024). Scenario modeling of urban heat in Delhi using RF. Science of the Total Environment, 912, 169104. https://doi.org/10.1016/j.scitotenv.2023.169104
Roberts, D. R., Bahn, V., Ciuti, S., Boyce, M. S., Elith, J., Guillera-Arroita, G., & Dormann, C. F. (2017). Cross-validation strategies for data with temporal, spatial and phylogenetic structure. Ecography, 40(8), 913–929. https://doi.org/10.1111/ecog.02881
Suleiman, R. (2022). Correlation between NDVI and LST in Abuja Metropolis. Nigerian Journal of Geography, 15(2), 33–49.
United Nations, Department of Economic and Social Affairs, Population Division. (2024). World urbanization prospects: The 2024 revision. https://population.un.org/wup/
UN-Habitat. (2024). World cities report 2024: Cities and climate change. United Nations Human Settlements Programme. https://unhabitat.org/wcr/
World Bank. (2023). Nigeria urbanization review: Productive cities for jobs and prosperity. https://doi.org/10.1596/978-1-4648-1972-8
Zhang, L., Wang, Q., & Liu, H. (2023). Scenario projection of LST under greening policies in Shanghai. Sustainable Cities and Society, 88, 104321. https://doi.org/10.1016/j.scs.2022.104321
Downloads
Published
Issue
Section
License
Copyright (c) 2026 This is an Open Access article distributed under the terms of the Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, adaptation, and reproduction in any medium, provided that the original work is properly cited.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Authors are permitted to post their work online in institutional/disciplinary repositories or on their own websites. Pre-print versions posted online should include a citation and link to the final published version in Journal of Librarianship and Scholarly Communication as soon as the issue is available; post-print versions (including the final publisher's PDF) should include a citation and link to the journal's website.
