AI Models Untangle Cities’ Twin Crises of Heat and Floods

Our rapidly urbanising world is grappling with two critical climate challenges: urban flooding and the Urban Heat Island (UHI) effect, where cities are significantly warmer than surrounding rural areas. These are not separate problems; their interactions create amplified thermal and hydrological stresses on infrastructure and public health.

A systematic review of over seventy-four scientific studies found that researchers often tackle these issues separately. The analysis showed a clear trend in the artificial intelligence tools employed: 60-65% of flood danger studies used 'ensemble' models like Random Forest, while 30-32% of UHI studies favoured 'hybrid' deep learning models.

However, this siloed approach overlooks their combined behaviour. The review identifies a critical gap: the lack of scalable models that can integrate real-time data, for instance from Internet of Things (IoT) sensors, and account for a city's unique physical form. By developing unified, data-driven frameworks, planners can better simulate these compound hazards and foster more sustainable, climate-resilient cities.