Linking environmental indicators to heat mitigation strategies: A statistical analysis of urban heat island and urban design indicators in Ho Chi Minh City

Abstract

Urban Heat Island (UHI) poses challenges in rapidly urbanizing cities like Ho Chi Minh City (HCMC). The complexity of urban morphology and the inevitable urban expansion process significantly impact land surface temperature (LST). This study investigates the spatial relationship between LST and seven urban environmental indicators, including both two-dimensional (Normalized Difference Vegetation Index (NDVI), Normalized Difference Built-up Index (NDBI), Normalized Difference Water Index (NDWI), impervious surface, albedo) and three-dimensional metrics (building volume, canopy height). Firstly, using high-resolution remote sensing data, the study models land surface temperature (LST) using Ordinary Least Squares (OLS) and Spatial Error Models (SEM). Secondly, diagnostic tests and k-fold cross-validation are applied to identify the better fit model. Results reveal that the SEM with quadratic NDVI outperforms OLS in addressing spatial autocorrelation and heteroskedasticity. The SEM with robust standard errors is the better fit model for inference and urban design interpretation. NDVI captures the non-linear relationship in HCMC, indicating that a threshold of approximately 0.147 provides a significant reduction in surface temperature. Built-up intensity, imperviousness, and albedo are the dominant warming factors, while an appropriate value of canopy height, water surface, and suitable building volume, combined with green vegetation coverage, can reduce the LST, depending on context. The importance of integrating these indicators as dense, structured green-blue infrastructure is emphasized to enable effective urban cooling in design solutions. The outcome highlights the need for localized urban design interventions to mitigate urban heat.