AppBenchmark: A User-Friendly Application for Urban Microclimate Models Qualification

As global warming accelerates and urban areas expand, understanding urban overheating and its local variations is critical for effective heat adaptation strategies. Numerical modeling is essential for diagnosing urban overheating and evaluating mitigation strategies. However, validating urban microclimate models remains a challenge due to data limitations and the complexity of urban systems, highlighting the need for a standardized framework for model qualification.

This paper introduces AppBenchmark, a user-friendly tool designed to evaluate urban microclimate models using an academic benchmark [1]. The benchmark methodology is based on intercomparison of microclimate simulations and, where available, comparison with reference data. Four benchmark cases are defined to analyze individual and coupled heat transfer processes in an idealized street canyon: shortwave radiation, longwave radiation, aeraulics, and their coupling with heat conduction and storage in walls and ground.

AppBenchmark processes surface-related variables (e.g., surface temperatures, heat fluxes) from formatted simulation data for analysis across canyon surfaces. To account for varying spatial and temporal resolutions, the tool provides three levels of analysis: surface averages, 1D profiles, and 2D fields. It also calculates RMSE to quantify deviations, offering a robust framework for evaluating and comparing simulation results.

An initial dataset containing simulation results from four microclimate models applied to the benchmark [1] is provided as a basis for comparison. The tool, dataset, and user documentation will be made publicly available on the DIAMS ANR project website. Expanding the benchmark application to additional models will help establish a standardized framework for urban microclimate model qualification. This initiative represents a first step toward a comprehensive validation methodology, with future efforts focused on incorporating additional physical phenomena and realistic urban configurations.

[1] Gresse, T. et al. Qualification of Microclimate Models and Simulation Tools: An Academic Benchmark. Available at SSRN: http://dx.doi.org/10.2139/ssrn.5036556