Current flood forecasting practices conducted by NWS hydrologists in Columbia consist of receiving forecasts from the Atlanta River Forecast Center and refining this information to make local flood warnings. The process presently involves manual overlays wherein high hazard dams, population, and other variables are considered within specific river basins. The primary task of this project was to develop a system to import WSR-88D precipitation into a geographic information system (GIS) and merge these data with existing databases that are relevant to flood forecasting.
Participants in this project included the NWS Columbia WSFO, the University of South Carolina Geography Department, and the Southeast Regional Climate Center. The first efforts on the project were dedicated to developing an efficient method for importing WSR-88D data into a geographic information system. Next, other appropriate GIS databases were constructed. Using a global positioning system, the South Carolina State Climatology Office determined precise locations of all precipitation gages in the state. In other cases, digital databases existed, but had to be geographically registered. These included the state road network and a hydrography database derived from the US Geological Survey. NOAA's Hydrological Research lab and the South Carolina Water Resources Commission also produced a digital inventory of all lakes and dams in the state which was imported into a GIS database. Including it should significantly improve the forecast office's ability to produce accurate flood watches and warnings. In addition to providing information about the location of dams across the state, the database includes information about the age, condition, and ownership of each dam, as well as the volume of lake water behind each. When these databases are merged in a GIS, they allow detailed analysis of potential flood hazard based on a variety of important controlling variables.
The work conducted enabled timely study of the use of 88D data prior to a heavy precipitation event in June 1994. The impact of three dam failures and the use of 88D rainfall estimates to analyze the intensity, duration, and coverage of the heavy precipitation underscored the usefulness of this product for assessment and planning. In addition the project has led to the identification of other areas of possible research that can use a combination of WSR-88D products and GIS, including hydrologic and agricultural modeling and drought monitoring.