Watershed memory at the Coweeta Hydrologic Laboratory: The effect of past precipitation and storage on hydrologic response. Issue 3 (6th March 2016)

• Record Type: Journal Article
• Authors: Nippgen, Fabian; McGlynn, Brian L.; Emanuel, Ryan E.; Vose, James M.
• Is Part Of: Water resources research. Volume 52:Issue 3(2016:Mar.)
• Page Start: 1673
• Page End: 1695
• Publication Date: 2016-03-06
• Subjects: watershed -- water balance -- storage -- watershed memory -- Coweeta -- precipitation; Hydrology -- Periodicals; 333.91
• DOI: 10.1002/2015WR018196
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: The rainfall‐runoff response of watersheds is affected by the legacy of past hydroclimatic conditions. We examined how variability in precipitation affected streamflow using 21 years of daily streamflow and precipitation data from five watersheds at the Coweeta Hydrologic Laboratory in southwestern North Carolina, USA. The gauged watersheds contained both coniferous and deciduous vegetation, dominant north and south aspects, and differing precipitation magnitudes. Lag‐correlations between precipitation and runoff ratios across a range of temporal resolutions indicated strong influence of past precipitation (i.e., watershed memory). At all time‐scales, runoff ratios strongly depended on the precipitation of previous time steps. At monthly time scales, the influence of past precipitation was detectable for up to 7 months. At seasonal time scales, the previous season had a greater effect on a season's runoff ratio than the same season's precipitation. At annual time scales, the previous year was equally important for a year's runoff ratio than the same year's precipitation. Estimated watershed storage through time and specifically the previous year's storage state was strongly correlated with the residuals of a regression between annual precipitation and annual runoff, partially explaining observed variability in annual runoff in watersheds with deep soils. This effect was less pronounced in the steepest watershed that also contained shallow soils. We suggest that the location of a watershed on a nonlinear watershed‐scale storage‐release curve can explain differences in runoff during growing and dormant season between watersheds with different annual evapotranspiration. Key Points: Watershed memory effects quantifiable at varying time scales Variability in annual runoff largely explained by previous year's storage state Storage‐release curve explains differences in Q for watersheds with different ET