The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Share
Publications Copernicus
Download
Citation
Share
Articles | Volume XLVIII-G-2025
https://doi.org/10.5194/isprs-archives-XLVIII-G-2025-1685-2025
https://doi.org/10.5194/isprs-archives-XLVIII-G-2025-1685-2025
02 Aug 2025
 | 02 Aug 2025

Geospatial Solutions for Urban Climate Adaptation: The LCZ-UHI-GEO Project between Italy and Vietnam

Matej Žgela, Alberto Vavassori, Maria Antonia Brovelli, Giovanna Venuti, Deodato Tapete, Patrizia Sacco, Pham Thi Mai Thy, and Lam Dao Nguyen

Keywords: Local Climate Zones, Urban Heat Island, Air Temperature Mapping, PRISMA Hyperspectral satellite, Urban Environments

Abstract. The LCZ-UHI-GEO project is an international collaboration between Italy and Vietnam, aiming to analyse urban climate dynamics, particularly the Urban Heat Island (UHI) effect, through geospatial techniques. The project focuses on four major cities—Rome and Milan in Italy, and Hanoi and Ho Chi Minh City in Vietnam. The project's objectives include mapping Local Climate Zones (LCZ) using Earth Observation (EO) data and mapping high-resolution air temperature based on in-situ and EO data. A key highlight is the integration of hyperspectral (HS) imagery from the PRISMA satellite to improve LCZ classification accuracy compared to traditional multispectral (MS) data. This paper presents the first experiments carried out for LCZ mapping in Rome and air temperature mapping in Milan. In Rome, LCZ mapping achieved 88.1% overall accuracy using PRISMA imagery, outperforming Sentinel-2 data which obtained an accuracy of 73.7%. Air temperature mapping in Milan employed a machine learning (ML) based interpolation combining Sentinel-2 data, GIS data, in-situ official weather station data, and crowdsourced Netatmo sensor measurements. Temperatures were modelled for the summer of 2022 for five specific diurnal periods, namely 4-6 AM, 9-11 AM, 2-4 PM, 6-8 PM, and 9 PM-12 AM. The resulting high-resolution temperature maps revealed distinctive UHI patterns for each period, with peak intensity observed during night-time. The model performed well across different periods, achieving an average root mean square error (RMSE) of 1.4°C. Future work will expand PRISMA data collection and provide multi-temporal LCZ and temperature mapping, particularly in Vietnamese cities. By utilising advanced geospatial technologies, the LCZ-UHI-GEO project contributes to sustainable urban development and addresses the challenges of urbanisation and climate change.

Share