SOLWEIG-GPU: GPU-Accelerated Thermal Comfort Modeling Framework for Urban Digital Twins

Several tools exist for modeling outdoor thermal comfort and radiative environments in cities.For example, RayMan performs computations at the point scale using simplified representations of urban geometry and is computationally efficient; however, it does not provide spatially continuous fields.ENVI-met is a three-dimensional prognostic model that resolves spatially explicit urban microclimate processes, but it is computationally expensive and has a closed-source implementation.In this context, we present Solar and LongWave Environmental Irradiance Geometry-Graphics Processing Unit (SOLWEIG-GPU), a Python package that enables execution of SOLWEIG version 2022a on GPUs via a single-line code interface, command-line interface (CLI), or graphical user interface (GUI).The package enables GPU-parallelized human thermal comfort modeling at meter-scale resolution across city-scale domains by computing key variables, including sky-view factor (SVF), mean radiant temperature (TMRT), ground shading, and the universal thermal climate index (UTCI).SOLWEIG-GPU is intended for use by academicians and researchers, as well as city weather, sustainability, and heat-risk practitioners.Sample data are available on Zenodo, and a quick-start guide is provided in the documentation.