Microscale Reynold Averaged Navier Stockes (RANS) models, while providing high-resolution urban climate simulations, are computationally demanding. This restricts their use for comparing and assessing the impact of urban planning scenarios from a climatological perspective. To overcome this limitation and support the development of climatologically resilient cities, we trained a neural network on a comprehensive dataset of RANS model outputs from numerous cities in Germany and Switzerland, enabling it to learn the complex relationships between urban morphology and microclimate. The GIS-based "Klimascanner" delivers rapid and accurate predictions of key urban climate parameters, like temperature and wind speed, and ensures immediate assessment of urban planning impacts. Validated and reliable, its user-friendly interface makes it accessible to planners and decision-makers, facilitating informed urban climate mitigation strategies.
