1 INTRODUCTION 1.1 Project overview Currently, the largest known undeveloped uranium ore deposit in North America, consisting of an esti- mated resource of 120 million pounds of uranium, is located in Pittsylvania County in south central Vir- ginia (Santoy Resources 2009). This uranium reserve is large enough to supply the fuel to all nuclear reac- tors in the United States for two years (U.S. Energy Information Administration 2010). As a result, a number of studies are being conducted to determine the potential implications of developing such a min- ing operation. As part of these investigations, a Probable Maxi- mum Flood (PMF) inundation study was carried out on an approximately 22 km 2 area surrounding the uranium reserve. The primary goal of this PMF analysis was to determine the spatial extent of flood- ing that would occur within the defined study area as a result of the region experiencing the theoretical worst-case scenario precipitation event. A meteorological model representing the Proba- ble Maximum Precipitation (PMP) was developed for the study watershed. The PMP storm data was simulated in an event-based hydrologic model of the study watershed to obtain PMF response hydro- graphs. These hydrographs were then routed through a hydraulic model to determine the extent and the el- evation of the water surface in the area. Finally, these stage values were utilized, in conjunction with topography and a GIS program, to develop a flood inundation map of the study area. This map will be used for site planning purposes if the mining site is developed to ensure that key structures and opera- tions are not impacted by flood waters during an ex- treme event. 1.2 Project site description The ore body around which this study was based is located between two rural streams: Whitethorn Creek to the north and one of its tributaries, Mill Creek, to the south. To capture the inundation effects from these streams, in addition to a third smaller stream, Dry Branch, that enters Whitethorn Creek just south of the Mill Creek confluence, the study watershed outlet point was selected as the conflu- ence of Dry Branch and Whitethorn Creek. This point is approximately 4 km upstream of the mouth of Whitethorn Creek and corresponds to an area where the creek floodplains begin to become con- fined in a narrow valley. Figure 1 depicts the 107 km 2 Whitethorn Creek study watershed delineated upstream of the selected outlet point, as well as the 22 km 2 PMF study area in which the hydraulic mod- eling and inundation mapping took place. The figure also shows the 10 subbasins that the watershed was divided into for hydrologic modeling purposes. The land cover of the watershed characterized primarily as deciduous forest and agricultural land (hay/pasture) (Fry et al. 2011) and over 85% of the watershed is considered to be underlain with soil having a moderately low runoff potential (hydrologic Probable maximum flood inundation modeling: a case study in southern Virginia W.J. Kingston, C.F. Castro-Bolinaga, E.R. Zavaleta, & P. Diplas Baker Environmental Hydraulics Laboratory, Department of Civil and Environmental Engineering, Virginia Polytechnic Institute and State University, Blacksburg, VA 24061, USA ABSTRACT: A Probable Maximum Flood inundation study was conducted on a 22 km 2 area located within the Whitethorn Creek watershed in south central Virginia, USA. For this analysis, a HEC-HMS hydrologic model was developed for a 107 km 2 study watershed to simulate the rainfall-runoff response resulting from the Probable Maximum Precipitation storm event. This event is representative of the worst-case rainfall event that is theoretically possible in the drainage. Outflow hydrographs from the hydrologic model were routed through a HEC-RAS hydraulic model of three area creeks to determine the extent of the inundation. An inun- dation map of the study area was generated using ArcGIS and topography from the National Elevation Dataset (1/3 arc-second resolution). Model development was facilitated through the use of HEC-GeoHMS and HEC- GeoRAS to assimilate several geospatial data sources and generate a number of modeling inputs.