Abstract.  We present a visualization-based method for a decision tool that enables exploring and understanding the link between urban land use planning policies and urban weather, in particular thunderstorms, heavy rainfall, and temperature changes. Our work combines recent work from urban planning, weather and climate studies, and visualization and computer graphics. Results from meteorological studies show that regional climate is related to regional land-use (i.e., the shape of urban and green spaces). We use work in procedural and urban modeling to demonstrate how to generate 3D city models from only coarse specifications of 2D fields. The result of our work is the compelling ability to quickly and interactively generate models of current cities and of potential future city configurations with greening patterns which are then used in simulating regional meteorology. We enable, and demonstrate, a first feedback loop of urban land use planning for mitigating extreme weather and climate change impacts in urban areas, yielding quick and intuitive controls and visualizations of how cities affect weather andvice versa.