Urban climate simulation for extreme heat events – A comparison between WRF and GEM

The Urban Heat Island (UHI) effect and extreme heat events (EHEs) pose significant challenges to urban areas under a changing climate, emphasizing the need for accurate urban climate modeling to guide mitigation strategies. This study evaluates the Weather Research and Forecasting (WRF) and Surface Prediction System (SPS) of the Global Environmental Multiscale (GEM) model during the 2018 Montreal heatwave. Both publicly available models are compared to assess their performance in modeling extreme heat events. WRF incorporates a Building Effect Parameterization (BEP) and the Building Energy Model (BEM), while GEM's Surface Prediction System (SPS) integrates Town Energy Balance (TEB) for urban surfaces. SPS achieved strong accuracy for wind speed and near-surface air temperature predictions (MAE: 1.1682 m/s, 1.6328 °C), while WRF demonstrated good results in land surface temperature estimates (spatially averaged differences <2 °C from MODIS observations). For relative humidity, SPS showed lower error values (STDE: 7.51–10.74 %) compared to WRF (STDE: 10.64–13.37 %), with both models exhibiting negative bias in humidity predictions. Analysis of diurnal urban heat island intensity showed WRF effectively captured overall patterns across different urban morphologies, while SPS performed well during transition periods. These findings highlight the importance of meteorological models' cross-comparison to address model-specific uncertainties and provide comprehensive insights into extreme heat events in urban settings, emphasizing the value of considering different urban meteorological models for resilient urban planning. • WRF and SPS models evaluated for simulation of an extreme heat event. • SPS predicts air temperature (MAE: 1.6328 °C) and wind speed (MAE: 1.1682 m/s) with high accuracy. • WRF excels in urban land surface temperature prediction, outperforming SPS in dense areas. • A multi-model approach can address uncertainties in urban climate simulations. • Both models demonstrate effective UHI modeling to aid urban planning and heat mitigation strategies.