Urban heat island dynamics in mid-sized cities: a longitudinal field study

Thesis Abstract

Urban heat island (UHI) effects in mid-sized cities pose rising energy demands, degraded outdoor comfort, and exacerbated air quality issues amid rapid urbanization. This study addresses the gap in longitudinal, small- to medium-city analyses by examining spatiotemporal UHI dynamics, drivers, and mitigation potentials over a five-year period in three mid-sized urban cores with populations between 150,000 and 500,000. The aims are to quantify temporal trends in surface and near-surface air temperatures, identify key urban morphology and land-use determinants, assess the effectiveness of green and blue infrastructure interventions, and model potential cooling benefits under different policy scenarios. Specific objectives include (i) estimating annual and seasonal UHI intensity using ground-based meteorological stations (n = 12 sites per city) and high-resolution Landsat and Sentinel-2 thermal and reflectance data (2019–2023), (ii) evaluating the influence of built form (albedo, sky-view factor, insulation of buildings), land cover change, and anthropogenic heat flux on UHI magnitude through multilevel regression analyses, (iii) analyzing the performance of implemented mitigation measures (urban parks, green roofs, permeable pavements) via a quasi-experimental design and difference-in-differences approach, and (iv) integrating socio-ecological perspectives to explore thermal comfort, energy consumption, and equity implications across neighborhoods. Methodologically, the study adopts a mixed-methods longitudinal design. The population comprises urban neighborhoods within the three cities, with a stratified random sample of 36 neighborhoods per city (total n ? 108). Data collection combines quantitative and qualitative instruments (a) continuous meteorological monitoring data from municipal stations and portable sensor packs (hourly air temperature, humidity, wind speed) and (b) remotely sensed thermal infrared imagery (12-bit Landsat data at 30 m resolution and Sentinel-2-derived land surface temperature products) supplemented by aerial drone surveys for microclimate mapping in representative blocks. Built-form and land-use variables are compiled from municipal GIS layers, satellite proxies, and field verification surveys (n ? 1,500 measurements). Intervention data include timing and specifications of park renovations, green roof deployments, and porous pavement installations. Climate and thermal comfort perceptions are captured via standardized surveys (n ? 1,200 respondents per city annually) and a subset of in-situ comfort trials (n ? 360 observations). Data analysis employs (i) descriptive statistics and time-series decomposition to delineate seasonal UHI signals, (ii) multilevel mixed-effects regression models to quantify the relationships between UHI intensity and predictors at parcel, block, and neighborhood levels, (iii) variance decomposition to apportion effects among climate, morphology, and land-use changes, (iv) a difference-in-differences framework to evaluate intervention efficacy, and (v) regression-based mediation analysis to explore pathways linking urban form to energy use and comfort. Model diagnostics include residual analysis, multicollinearity checks, and cross-validation. The theoretical grounding draws on the Urban Metabolism framework and the Theory of Urban Microclimates, incorporating concepts from the Surface Urban Energy Balance model for kernel land-surface temperature estimation and the Green Infrastructure Typology for intervention classification. Expected findings indicate that mid-sized cities exhibit pronounced diurnal UHI with seasonal amplification during summer, modulated by nighttime cooling rates as a function of sky-view factor and vegetative cover. Built form and land-use change are anticipated to explain a substantial proportion of variance in UHI intensity (R2 change > 0.40 in several models), while green infrastructure is expected to yield measurable cooling effects (2–4°C reductions in localized zones) contingent on scale, connectivity, and maintenance. The study will reveal inequities in thermal exposure correlating with neighborhood socioeconomic status and housing quality, and it will quantify associated energy consumption differentials, informing targeted mitigation prioritization. The contribution to knowledge includes (i) a robust longitudinal empirical framework for UHI dynamics in mid-sized cities integrating ground, aerial, and survey data; (ii) empirically derived estimates of intervention effectiveness across different urban morphologies, informing scalable, low-cost adaptation strategies; and (iii) enhanced understanding of the socio-spatial distribution of thermal stress and its implications for energy policy and urban equity. The study concludes with policy recommendations emphasizing prioritized green-blue infrastructure investments, urban design guidelines to optimize albedo and ventilation, and data-driven planning tools for ongoing UHI monitoring.