Building structures in cities affect the energy balance of urban microclimate. To understand and quantify effects of urban form on local microclimate, high resolution urban parameter inputs to numerical simulation models are needed. The Neighborhood Adaptive Tissues for Urban Resilience Futures (NATURF) model addresses this need by creating WRF-readable binary files that contain parameters for urban morphologies at sub-kilometer resolution. The parameters are generated from the physical dimensions of actual buildings so that their structures may be represented in three dimensions and spatially geolocated within modeling environments. NATURF was developed under the Integrated Multisector Multiscale Modeling (IM3) Scientific Focus Area led by Pacific Northwest National Laboratory. It has produced county-coverage data sets at 100m horizontal and 5m vertical resolution using various three-dimensional building input data sets. Most recently, NATURF has used ORNL’s Model America building shapefiles as input to support the ongoing city-scale experiments of IM3. This model also participates in an IM3 model chain that will produce ensembles of new neighborhoods that may be built in the future for each city. The IM3 model chain consists of population predictions, high-resolution city-scale Land Use Land Cover Change projections, the NATURF Generative Adversarial Network (GAN) component to generate potential building footprints, heights and placement for new neighborhoods based on historical training examples (paired images of LULC and building footprints with grayscale encoded heights); and the Population-Urban Morphology Attribution (PMA) model, which will provide building volume constraints based on urban scaling for the potential new neighborhood morphologies. This talk will discuss NATURF, its capabilities, and its connection to the upstream and downstream models in the IM3 chain.