There is an increasing demand for the FAA to allow small UAS (sUAS) operations in the US national airspace system. The FAA must understand and quantify the risk of collisions with manned aircraft during desired sUAS operations in order to produce regulations and standards. The FAA has made extensive use of Monte Carlo collision risk analysis simulations in developing standards for manned aircraft collision avoidance systems, as well as separation standards and Detect and Avoid systems for large UAS. Similar Monte Carlo simulations are desired for use in collision risk simulations for sUAS operations and system development. A key component of Monte Carlo simulations are models of aircraft flight. Models for manned aircraft are developed from recorded radar data. These manned aircraft models are modified to model large UAS which fly in similar ways. However, new methods are needed to model sUAS flights since there is no corresponding dataset of flights, and their flight capabilities are significantly different from manned aircraft. We demonstrate a methodology for developing sUAS flight models that leverages OpenStreetMap, an open source geospatial information and map dataset, to generate representative unmanned operations at low altitudes for use in Monte Carlo collision risk simulations. For example we model rail line inspection flights by following rail lines, and hobbyist flights by locating parks and other recreational sites. We automatically identify geographic locations of interest for various use cases, and generate appropriate associated transit, search, and hover trajectories for different sUAS performance dynamics and cruise altitudes. We leverage the map data to identify potential variations in operational density.