Event-based Recession Analysis across Scales

Hydrograph recessions have long been a window to investigate hydrological processes and their interactions. The authors conducted an exploratory analysis of about 1000 individual hydrograph recessions in a period of around 15 years (1995-2010) from time series of hourly discharge (USGS IDA stream flow data set) at 27 USGS gauges located in Iowa and Cedar River basins with drainage area ranging from 6.7 to around 17000 km2. They calculated recession exponents with the same recession length but different time lags from the hydrograph peak ranging from ~0 to 96 hours, and then plotted them against time lags to construct the evolution of recession exponent. The result shows that, as recession continues, the recession exponent in first increases quickly, then decreases quickly, and finally stays constant. Occasionally and for different reasons, the decreasing portion is missing due to negligible contribution from soil water storage. The increasing part of the evolution of can be related to fast response to rainfall including overland flow and quick subsurface flow through macropores (or tiles), and the decreasing portion can be connected to the delayed soil water response. Lastly, the constant segment can be attributed to the groundwater storage with the slowest response. The points where recession exponent reaches its maximum and begins to plateau are the times that fast response and soil water response end, respectively. The authors conducted further theoretical analysis by combining mathematical derivation and literature results to explain the observed evolution path of the recession exponent . Their results have a direct application in hydrograph separation and important implications for dynamic basin storage-discharge relation analysis and hydrological process understanding across scales.