Addressing Mixed Populations In Flood Frequency Analyses: A Case Study In Eastern Pennsylvania

Three common issues associated with using streamflow records in Bulletin 17C flow frequency analyses are regulation from water control structures, short record lengths, and multiple flood mechanisms contributing to mixed populations. These issues were identified in a flow frequency analysis of the Lehigh River for a U.S. Army Corps of Engineers (USACE) feasibility study. The Lehigh River includes approximately 1,300 square miles of land area in nine counties in eastern Pennsylvania. The river has been subjected to regulation from two USACE dams since 1961. While some of the gages on the Lehigh River gages pre-date construction of the dams, others do not. Flood events that affect the Lehigh River basin include summer thunderstorms, tropical storms, and combinations of snow and rainfall. Events can be localized in scale and primarily affect the Lehigh River basin or regional in scale and affect the larger Delaware River watershed.

Samples of unregulated annual peak flows were developed for rain-on-snow, tropical storm, and thunderstorm events. Regulated streamflow records were converted to unregulated records in the Hydrologic Engineering Center’s Hydrologic Modeling System (HEC-HMS) using (1) a period of record simulation and (2) a hydrologic routing model. Snow water equivalent, temperature, and precipitation datasets were evaluated to split the annual peak flow series by meteorologic process. Streamflow record lengths were extended using new correlation analysis and record extension features in HEC’s Statistical Software Program (HEC-SSP). In addition, relationships between daily average and hourly streamflow records were developed to leverage extended record lengths from daily data. These analyses address topics identified as future study areas in Bulletin 17C.