Reducing redundancy among flow metrics in support of flow-ecology modeling in the Cape Fear and Pee Dee/Yadkin river basins
Flow-ecology relations are being investigated in the Cape Fear and Pee Dee/Yadkin river basins in North and South Carolina as part of the USGS’s Water Smart program. State fish and invertebrate bioassessment data are being combined with modeled (SWAT) daily flow data as a means of projecting responses (flow and biological) to climate change and evaluating possible flow criteria. Flow metrics are calculated from modeled flows using the USGS’s EflowStats R-package which calculates 198 metrics encompassing flow magnitude, duration, timing, frequency, and rate of change. Many of these metrics are highly correlated, therefore, it is necessary to reduce these flow metrics to a more manageable and less redundant set of metrics prior to modeling flow-ecology relations. A combination of literature review, correlation, principle component analysis, and local stake holder input was used to reduce the number of metrics from 189 to 30 metrics minimally redundant metrics that represent the five major components of flow. These assessments will assist in building models that help local water resource managers plan for future scenarios of water use and availability.
Jason May (Primary Presenter/Author), U.S. Geological Survey, California Water Science Center, firstname.lastname@example.org;
Jonathan Kennen ( Co-Presenter/Co-Author), U.S. Geological Survey, New Jersey Water Science Center, 3450 Princeton Pike, Suite 110, Lawrenceville, NJ 08648, email@example.com;
Thomas Cuffney ( Co-Presenter/Co-Author), U.S. Geological Survey, South Atlantic Water Science Center, 3916 Sunset Ridge Rd., Raleigh, NC 27607, firstname.lastname@example.org;