Automated Microclimate Model Generation from Remote Sensing Data

The ongoing climate crisis has highlighted the need for sustainability and resilience in the development and maintenance of urban areas regarding climate comfort. Weather simulation tools can aid researchers in understanding the effects that weather has on the microclimate in urban areas. While simulations are handled autonomously by computers once set up, the creation of the requisite input urban models is still a highly manual process. In this study, a novel method for the automated generation of urban models using land and cadastral remote sensing data is presented. By analyzing grass, trees, buildings, and roads algorithmically, data can be extracted and configured into spatial models compatible with microclimate simulation software such as ENVI-Met. Comparison to a baseline model shows that our method enables the creation of models fit for use for exploring microclimate scenarios in the urban environment, saving time by eliminating the need for manual processing.