Alioua et al. (2025) Quantitative Analysis of Urban Heat Island and ISA Density Effects on Land Surface Temperature: A Remote Sensing Study Using Local Morphological Density Analysis and Hybrid Optimization Model

Identification

• Journal: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
• Year: 2025
• Date: 2025-11-19
• Authors: Nor El Houda Alioua, Akila Kemmouche, Alessandra Capolupo, Eufemia Tarantino
• DOI: 10.1109/jstars.2025.3634721

Research Groups

• Laboratory of Image Processing and Radiation, Department of Telecommunication, Faculty of Electrical Engineering, University of Sciences and Technology Houari Boumediene, Algiers, Algeria.
• Department of Civil, Environmental, Land, Construction and Chemistry, Politecnico di Bari, Bari, Italy.

Short Summary

This study developed a novel hybrid optimization model (Genetic Algorithm-Simulated Annealing with Generalized Additive Models) to quantify the nonlinear relationship between impervious surface area (ISA) density and Land Surface Temperature (LST) and identify Urban Heat Island (UHI) thresholds in Mediterranean cities, revealing a strong positive correlation between ISA density and LST and a mitigating effect of green spaces.

Objective

• To develop and implement a novel method for modeling the relationship between Impervious Surface Area (ISA)/Green Space (GS) densities and Land Surface Temperature (LST) and identify Urban Heat Island (UHI) thresholds using a hybrid optimization approach (GA-SA/GAM).

Study Configuration

• Spatial Scale: Two Mediterranean cities: Taranto, Italy (study area 150 km²) and Algiers, Algeria (study area 558 km²). Satellite imagery resolutions: Landsat 7 ETM+ (30 m visible/NIR, 60 m thermal), ASTER (15 m visible/NIR, 60 m/90 m thermal), and GLC_FCS30D (30 m LULC).
• Temporal Scale: Taranto (2000–2014) and Algiers (2012–2021), focusing on summertime imagery. LULC data from 1985 to 2022.

Methodology and Data

• Models used:
  • Hybrid Optimization Model: Genetic Algorithm (GA) - Simulated Annealing (SA) integrated with Generalized Additive Models (GAM).
  • Local Morphological Density (LMD) analysis for ISA and Green Space density mapping.
  • Planck function for LST retrieval.
  • Parametric Two-tailed Student’s t-test and nonparametric Mann–Whitney–Wilcoxon Rank sum test (MWWRs test) for LST temporal change assessment.
  • Akaike Information Criterion (AIC) for model fitness evaluation within the hybrid optimization.
• Data sources:
  • Satellite: ASTER and Landsat 7 ETM+ imagery (sourced from USGS platform).
  • Validation: Earth Skin Temperature (EST) data from NASA’s Prediction of Worldwide Energy Resource (POWER) daily time series.
  • Land Use/Land Cover (LULC): GLC_FCS30D dataset.

Main Results

• A distinct positive nonlinear relationship was observed between ISA density and LST, with critical thresholds (e.g., 20–30% ISA density) beyond which LST escalates sharply.
• Green space density consistently exhibited a negative relationship with LST, confirming its cooling effect.
• UHI zones showed significantly higher mean LST values:
  • Taranto: Increased from 38.78 °C in 2000 to 42.11 °C in 2014.
  • Algiers: Increased from 33.74 °C in 2012 to 38.38 °C in 2021.
• The mean LST difference between UHI and non-UHI areas intensified over time:
  • Taranto: From 9.93 °C (2000) to 10.29 °C (2014).
  • Algiers: From 10.32 °C (2012) to 11.39 °C (2021).
• ASTER-based LST retrieval generally demonstrated higher accuracy compared to Landsat-based retrieval (e.g., Taranto 2014 ASTER RMSE: 1.84 °C, MAE: 1.76 °C).
• ISA area increased in both cities: Taranto from 36.11% (2000) to 43.33% (2014); Algiers from 53.37% (2012) to 71.25% (2021).
• Green Space (GS) coverage declined: Taranto from 61.47% (2000) to 56.60% (2014); Algiers from 46.57% (2012) to 28.74% (2021).

Contributions

• Introduces a novel hybrid optimization model (GA-SA/GAM) for dynamic and data-driven UHI threshold identification, offering a significant advancement over traditional static methods.
• Integrates robust global search capabilities of Genetic Algorithms and local refinement efficiency of Simulated Annealing with the flexible nonlinear modeling of Generalized Additive Models to precisely detect ISA/LST relationships.
• Develops a transferable computational framework that enhances UHI detection in urban remote sensing applications.
• Provides improved capacity to assess climate-sensitive urban transformations in Mediterranean cities, supporting sustainable urban planning and climate adaptation strategies.

Funding

• Space It Up project, funded by the Italian Space Agency (ASI) and the Ministry of University and Research (MUR) under Contract 2024-5-E.0 - CUP I53D24000060005.

Citation

@article{Alioua2025Quantitative,
  author = {Alioua, Nor El Houda and Kemmouche, Akila and Capolupo, Alessandra and Tarantino, Eufemia},
  title = {Quantitative Analysis of Urban Heat Island and ISA Density Effects on Land Surface Temperature: A Remote Sensing Study Using Local Morphological Density Analysis and Hybrid Optimization Model},
  journal = {IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  year = {2025},
  doi = {10.1109/jstars.2025.3634721},
  url = {https://doi.org/10.1109/jstars.2025.3634721}
}