RELATIONSHIPS BETWEEN N-DAY FLOOD VOLUMES FOR INFREQUENT LARGE FLOODS

Design floods of specified frequency are used to consider relevant hydrologic and economic factors int he evaluation of operation, or changes in operation, of flood-damage mitigation facilities. The design flood hydrograph is constructed using flood volumes, often extrapolated beyond observed records, from flood-volume-duration-frequency curves. We examined the foundation of this approach using flood flow data from seven Pacific Northwest river catchments. Nested occurrence of large-return-period flood volumes (i.e., 1-day and n-day flood volumes occur concurrently), for flood durations important to decision making, held true for all seven basins. Coincidence of floods occurs when the 1-day and n-day flood volumes have the same return period. Flood volume-frequency and duration for the largest two or three flood volumes is strongly related (coincident) in three catchments and moderately related in three others. No apparent relationship was observed in the seventh. No link is apparent for any of the seven catchments between catchment physical and climatological features and flood-volume concurrence or coincidence of frequency. These results indicate the flood-hydrograph-frequency record for each catchment should be evaluated prior to using assumed nested volume relationships to derive a design flood hydrograph having a given exceedance frequency. A general method for constructing design flood hydrographs for low exceedance probabilities (high return periods) is given.