Adopting Explainable-AI to investigate the impact of urban morphology design on energy and environmental performance in dry-arid climates

Urban design choices directly impact energy consumption: The study found that three specific design parameters—building shape, window-to-wall ratio, and commercial ratio—are the most critical factors affecting how much energy buildings need for cooling, heating, and lighting in dense urban areas of dry climates like Tehran.

Neighboring buildings matter more than you might think: The heights and distances of surrounding buildings strongly influence both cooling demand and solar access for individual buildings, suggesting that urban planning decisions must consider block-level interactions rather than treating buildings in isolation.

AI can make complex urban planning decisions more transparent: By using Explainable AI techniques (specifically SHAP), researchers were able to identify which urban design factors matter most and why, making it possible for city planners and policymakers to understand the reasoning behind AI recommendations rather than treating them as "black boxes."

Machine learning can rapidly evaluate urban design options: The XGBoost model achieved high accuracy (R² of 0.92) while training in just 3.64 seconds, demonstrating that AI-powered tools could help cities quickly test and compare different urban morphology scenarios to find climate-responsive designs.

Findings from one city can help others: The research on Tehran's dense urban landscape produced generalizable insights applicable to other dry-arid regions, suggesting that this explainable AI framework could be adapted as a scalable tool for sustainable urban planning in similar climates worldwide.