Identifying predictive thresholds and marginal benefit boundaries of urban heat environments via explainable AI

Abstract

Urban heat island (UHI) effects have become a critical challenge constraining urban sustainability, as accelerating global urbanization has intensified thermal stress in cities worldwide. However, existing studies predominantly rely on correlation analysis and linear regression to explore relationships between urban spatial features and land surface temperature (LST). While recent machine learning studies have improved predictive accuracy, they lack interpretable quantitative methods to systematically identify the marginal benefit boundaries of key driving factors for urban heat environment regulation, thereby hindering the translation of model outputs into actionable, data-driven planning frameworks.Taking Shijiazhuang, a typical high-density city in northern China, as the study area, this study integrates multi-source heterogeneous data including remote sensing imagery, street-view-derived sky view factor (SVF), traffic flow and points of interest, to construct a random forest model combined with explainable AI (XAI) techniques. An XAI-driven “data—model—threshold—decision” framework is developed to characterise nonlinear predictive relationships between urban elements and summer LST via three core quantitative indicators.Results demonstrate that SVF is the most temporally stable LST predictor; key predictive thresholds of 0.270 for building density and 2,800 vehicles/hour for traffic flow are identified; enhanced vegetation index and building morphology exhibit diurnal complementary patterns, with socioeconomic variables showing pronounced nonlinear co-variation with the thermal environment. These findings provide quantitative references for evidence-based, climate-adaptive urban planning.

Keywords

Urban heat island effect / Explainable AI / Nonlinear predictive thresholds / Marginal benefit boundaries / Land surface temperature

References

[1] Chen X, Chen M, Li X, Gao S, Yang Q, Wen Y, Sun Q. Advancing climate resilience through a geo-design framework: Strengthening urban and community forestry for sustainable environmental design. Journal of Forestry Research, 2024.

[2] Chen H, Zheng Y, Xing J, Zhou C. Urban planning priorities analysis based on urban heat risk supply. Sustainable Cities and Society, 2025, 130. ArticleID: 106502

[3] Cheng Q, Jin H, Ren Y. Compound daytime and nighttime heatwaves for air and surface temperature based on relative and absolute threshold dynamic classified in Southwest China, 1980–2019. Sustainable Cities and Society, 2023, 91. ArticleID: 104433

[4] Esposito A, Pappaccogli G, Donateo A, Salizzoni P, Maffeis G, Semeraro T, Buccolieri R. Urban morphology and surface urban heat island relationship during heat waves: A study of Milan and Lecce (Italy). Remote Sensing, 2024, 16(23. ArticleID: 4496

[5] Friedman, J. H., & Popescu, B. E. (2008). Predictive learning via rule ensembles. The Annals of Applied Statistics, 2(3). https://doi.org/10.1214/07-aoas148

[6] Gaonkar VP, Nadaf FM, Kapale V. Decoding urban heat: A decadal analysis (1991–2021) of land surface temperature and thermal comfort dynamics in coastal Taluka of Bardez, Goa, India. Disaster Advances, 2024, 18(2): 1991.

[7] Ghanghermeh A, Roshan G, Asadi K, Attia S. Spatiotemporal analysis of urban heat islands and vegetation cover using emerging hotspot analysis in a humid subtropical climate. Atmosphere, 2024, 15(2. ArticleID: 161

[8] Han W. Analyzing the scale dependent effect of urban building morphology on land surface temperature using random forest algorithm. Scientific Reports, 2023.

[9] Han L, Zhao J, Gao Y, Gu Z. Prediction and evaluation of spatial distributions of ozone and urban heat island using a machine learning modified land use regression method. Sustainable Cities and Society, 2022, 78. ArticleID: 103643

[10] He X, Yuan Q, Qin Y, Lu J, Li G. Analysis of surface urban heat island in the Guangzhou-Foshan metropolitan area based on local climate zones. Land, 2024, 13(10. ArticleID: 1626

[11] Hong T, Heo Y. Exploring the impact of urban factors on land surface temperature and outdoor air temperature: A case study in Seoul, Korea. Building and Environment, 2023, 243. ArticleID: 110645

[12] Hu L, Brunsell NA. The impact of temporal aggregation of land surface temperature data for surface urban heat island (SUHI) monitoring. Remote Sensing of Environment, 2013.

[13] Huang Y, Lin J, He X, Lin Z, Wu Z, Zhang X. Assessing the scale effect of urban vertical patterns on urban waterlogging: An empirical study in Shenzhen. Environmental Impact Assessment Review, 2024, 106. ArticleID: 107486

[14] Huang, X., & Wang, Y. (2019). Investigating the effects of 3D urban morphology on the surface urban heat island effect in urban functional zones by using high-resolution remote sensing data: A case study of Wuhan, Central China. ISPRS Journal of Photogrammetry and Remote Sensing, 152. https://doi.org/10.1016/j.isprsjprs.2019.04.010

[15] Jiang X, Ma T, Liang J, Wang Y. The greener the cooler? Nonlinear thresholds in the cooling effects of urban blue-green spaces and implications for sustainable planning in the Yangtze River Delta. Urban Climate, 2025, 64. ArticleID: 102679

[16] Kakooei M, Baleghi Y. Spatial-temporal analysis of urban environmental variables using building height features. Urban Climate, 2023, 52. ArticleID: 101736

[17] Kulworatit, C., & Petchhan, J. (2025). Analysis of Urban Heat Islands Using Nighttime Land Surface Temperature Data. In 2025 International Technical Conference on Circuits/Systems, Computers, and Communications (ITC-CSCC). IEEE. https://doi.org/10.1109/itc-cscc66376.2025.11137603

[18] Li J, Liu Z, Tang X, Liu Y, Qin C, Wu Y, Huang J. How building height affects the global urban surface heat island from local climate zone perspective. Sustainable Cities and Society, 2025, 130. ArticleID: 106660

[19] Liao S, Liu Z. Explaining and Reducing Urban Heat Islands Through Machine Learning: Evidence from New York City. Buildings, 2026, 16(1): 186.

[20] Luo T, He Q, Wang W, Fan X. Response of summer land surface temperature of small and medium-sized cities to their neighboring urban spatial morphology. Building and Environment, 2024, 250. ArticleID: 111198

[21] Meyer H, Pebesma E. Machine learning for spatial data: The importance of spatial cross-validation. Geographical Analysis, 2022, 54(1): 3-20.

[22] Mokarram M, Taripanah F, Pham TM. Spatial-temporal analysis of atmospheric environment in urban areas using remote sensing and neural networks. Sustainable Computing: Informatics and Systems, 2024, 42. ArticleID: 100987

[23] Molnar, C., Freiesleben, T., König, G., Herbinger, J., Reisinger, T., Casalicchio, G., … Bischl, B. (2023). Relating the Partial Dependence Plot and Permutation Feature Importance to the Data Generating Process. In Communications in Computer and Information Science. Springer Nature Switzerland. https://doi.org/10.1007/978-3-031-44064-9_24

[24] Nakaegawa T, Kobashi T. Indonesia climate change: Observations and future projections in IPCC AR6 WG I and beyond. Forum Geografi, 2025.

[25] Noviani R, Saputra AE, Wijayanti P, Koesoma S. Analysis of land use land cover and land surface temperature in Karst Area: A case study Wonogiri Regency. Indonesian Journal of Applied Physics, 2024, 14(1): 89.

[26] O’Malley C, Kikumoto H. An investigation into heat storage by adopting local climate zones and nocturnal-diurnal urban heat island differences in the Tokyo Prefecture. Sustainable Cities and Society, 2022, 83. ArticleID: 103959

[27] Ofia, I. C., Husain, O. A., Jetta, S. E., Bello, R. A., Adegbo, G. A., & Chinonyerem, C. A. (2024). Urban heat island analysis using GIS and remote sensing in Koshoffe Lagos State. International Journal of Nature and Science Advance Research. https://doi.org/10.70382/mejnsar.v6i9.002

[28] Rana VK, Suryanarayana TMV. Quantifying the effects of urban land forms on land surface temperature and modelling the spatial variation using machine learning. Geocarto International, 2022.

[29] Reshef DN, Reshef YA, et al.. Detecting novel associations in large data sets. Science, 2011, 334(6062): 1518-1524.

[30] Roberts DR, Bahn V, Ciuti S, Boyce MS, Elith J, Guillera-Arroita G, Hooten MB, Warton DI, Thuiller W. Cross-validation strategies for data with temporal, spatial, hierarchical, or phylogenetic structure. Ecography, 2017, 40(8): 913-929.

[31] Snaiki R, Merabtine A. Recent advances on machine learning techniques for urban heat island applications: A review and new horizons. Sustainable Cities and Society, 2025, 134. ArticleID: 106943

[32] Salam, M., Islam, Md. K., Jahan, I., & Chowdhury, Md. A. (2024). Assessing the impacts of vegetation loss and land surface temperature on Surface Urban Heat Island (SUHI) in Gazipur District, Bangladesh. Computational Urban Science, 4(1). https://doi.org/10.1007/s43762-024-00136-y

[33] Son TH, Weedon Z, Yigitcanlar T, Sanchez T, Corchado JM, Mehmood R. Algorithmic urban planning for smart and sustainable development: Systematic review of the literature. Sustainable Cities and Society, 2023, 94. ArticleID: 104562

[34] Strobl C, Boulesteix A-L, Kneib T, Augustin T, Zeileis A. Conditional variable importance for random forests. BMC Bioinformatics, 2008.

[35] Wang X, Yang J, Zhang R, Xin J, Yu W, Ren J, Xiao X. Assessment of thermal vulnerability and analysis of influencing factors in urban functional areas of Foshan City, China. Chinese Geographical Science, 2025.

[36] Wang Y, Xie X, Zhao X, Liang S, Zhu B, Tursun A, Zhang X. Four-decade response of land surface temperature to urban expansion in Beijing. Agricultural and Forest Meteorology, 2023, 341: 109653.

[37] Xi Y, Wang S, Zou Y, Zhou X, Zhang Y. Seasonal surface urban heat island analysis based on local climate zones. Ecological Indicators, 2024, 159. ArticleID: 111669

[38] You M, Huang J, Guan C. Are new towns prone to urban heat island effect? Implications for planning form and function. Sustainable Cities and Society, 2023, 99. ArticleID: 104939

[39] Yu B, Pan J. How do 2D and 3D urban morphology impact spatial patterns of thermal environment? A nested multi-scale local climate zone perspective. Building and Environment, 2026, 288. ArticleID: 114014

[40] Yu S, Chen Z, Yu B, Wang L, Wu B, Wu J, Zhao F. Exploring the relationship between 2D/3D landscape pattern and land surface temperature based on explainable eXtreme Gradient Boosting tree: A case study of Shanghai. China. Science of the Total Environment, 2020, 725. ArticleID: 138229

[41] Zhang N, Ding Q, Wang H. Investigating the influence of urban green spaces on urban heat island mitigation – Taking four districts in Shijiazhuang as an example. Journal of Environmental Engineering and Landscape Management, 2025.

[42] Zhang M, Zhou D, Xu D, Liu K, Gu Z, Yang Y, Zhang Q, Chen L. How do built environment factors influence urban heat islands across local climate zones? Evidence from an interpretable XGBoost-SHAP model. Sustainable Cities and Society, 2026, 136. ArticleID: 107078

[43] Zhang, R. (2025). Research on Urban Heat Island Effect in Chengdu Based on Landsat Images and Analysis of Influencing Factors. Science and Technology of Engineering, Chemistry and Environmental Protection, 1(4). https://doi.org/10.61173/4gs40439

[44] Zhou S, Xu X, Xu H, Zhao Z, Yuan H, Wang Y, Wu T, Wang M, Wu Z. From heat resilience to sustainable co-benefits: Adaptive urban morphology generation based on multimodal data fusion and a novel generative framework. Sustainable Cities and Society, 2025, 127. ArticleID: 106452

[45] Zhu Z, Shen Y, Fu W, Zheng D, Huang P, Li J, Yao X. How does 2D and 3D of urban morphology affect the seasonal land surface temperature in Island City? A block-scale perspective. Ecological Indicators, 2023, 150. ArticleID: 110221

[46] Zong Y, Yu Y, Peng K, Zhang R, Zhou W. Quantifying non-linearities and interactions in urban forest cooling using interpretable machine learning. Forests, 2025, 16(10. ArticleID: 1514

RIGHTS & PERMISSIONS

The Author(s)