[Article] Machine Learning-Driven Mapping of Heatwave Health Risks Across Local Climate Zones in a Mediterranean Context

What if the greatest heat risk in our cities isn't simply where temperatures are highest... but where heat, urban form, and human vulnerability intersect?

For years, urban heat studies have focused primarily on mapping temperatures. Yet extreme heat becomes a public health crisis only when high temperatures overlap with vulnerable populations and specific urban environments.

Understanding this interaction is becoming increasingly important as Mediterranean cities experience more frequent and intense heatwaves.

Our latest study, published in Earth Systems and Environment, presents a new machine learning framework that combines Heat Health Risk Index (HHRI), Surface Urban Heat Island (SUHI), and Local Climate Zones (LCZs) to identify the neighborhoods facing the greatest heat-related health risks in Algiers, Algeria.

Using satellite observations, meteorological records, demographic information, remote sensing, Principal Component Analysis (PCA), and unsupervised machine learning, we developed high-resolution maps revealing where thermal stress and social vulnerability reinforce one another.

Among our findings:

• Local Climate Zones 4, 5, and 8 exhibit the highest overall heat-health risks, where urban morphology and vulnerable populations combine to create critical hotspots.

• Surface Urban Heat Island intensity is strongest in LCZ 8, followed by LCZ 6 and LCZ 4, demonstrating how different urban forms retain heat very differently.

• Machine learning clustering revealed four distinct urban heat risk profiles, allowing cities to distinguish between areas dominated by thermal exposure, social vulnerability, or both simultaneously.

• The framework provides a transferable methodology for climate adaptation planning in Mediterranean, North African, and other data-scarce cities facing increasing heat extremes.

The implications extend well beyond Algiers.

As climate change accelerates, cities need to move beyond simply measuring urban temperatures. The next generation of climate adaptation requires understanding who is exposed, where risks accumulate, and why certain urban forms amplify vulnerability.

Machine learning, remote sensing, and climate science are becoming essential tools for designing healthier, more resilient cities.

Congratulations to Dyna Chourouk Zitouni, Djihed Berkouk, Mohamed Elhadi Matallah, Mohamed Akram Eddine Ben Ratmia, Ayyoob Sharifi, and all collaborators on this excellent work.

A question for urban planners, climate scientists, public health researchers, and policymakers:

Should future heat adaptation plans prioritize neighborhoods with the highest temperatures, or those where heat exposure and human vulnerability overlap the most?

📘 Full article: https://hdl.handle.net/2268/336924

📚 Learn more about our research: https://www.sbd.uliege.be/

💻 Subscribe to my newsletter: https://lnkd.in/diTVT5eq

🅱️ Bilibili: b23.tv/bzjL3bn | 🎬 YouTube: https://lnkd.in/erHrfkNf

🌐 Explore previous posts and resources: https://www.shadyattia.org/

🔗 Follow all my professional links, including WeChat 🟩💬 微信: https://lnkd.in/eN3xZhhZ

#HeatHealth #UrbanHeatIsland #Heatwave #ClimateChange #MachineLearning #RemoteSensing #LocalClimateZones #UrbanClimate #ClimateResilience #HealthyCities #UrbanPlanning #BuildingScience #ArtificialIntelligence #Mediterranean #Algeria #UniversityOfLiege #SBDLab