Water storage at the Panola Mountain Research Watershed, Georgia, USA

Abstract

AbstractStorage is a major component of a catchment water balance particularly when the water balance components are evaluated on short time scales, that is, less than annual. We propose a method of determining the storage–discharge relation using an exponential function and daily precipitation, potential evapotranspiration (PET) and baseflow during the dormant season when evapotranspiration (ET) is low. The method was applied to the 22‐year data series of the 0.41‐ha forested Panola Mountain Research Watershed, Georgia. The relation of cumulative daily precipitation minus daily runoff and PET versus baseflow was highly significant (r2 = 0.92, p < 0.0001), but the initial storage for each year varied markedly. For the 22‐year study period, annual precipitation and runoff averaged 1240 and 380 mm, respectively, whereas the absolute catchment storage range was ~400 mm, averaging 219 mm annually, which is attributed to contributions of soil water and groundwater. The soil moisture of a catchment average 1‐m soil depth was evaluated and suggests that there was an active (changes in soil storage during stormflow) and passive (a longer‐term seasonal cycle) soil water storage with ranges of 40–70 and 100–120 mm, respectively. The active soil water storage was short term on the order of days during and immediately after rainstorms, and the passive or seasonal soil storage was highest during winter when ET was lowest and lowest during summer when ET was highest. An estimate of ET from daily changes in soil moisture (ETSM) during recessions was comparable with PET during the dormant season (1.5 mm day−1) but was much lower during the growing season (June through August); monthly average SMET and PET ranged from 2.8 to 4.0 mm day−1 and from 4.5 to 5.5 mm day−1, respectively. The growing season difference is attributed to the overestimation of PET. ETSM estimates were comparable with those derived from hillslope water balances during sprinkling experiments. Master recession curves derived from the storage–discharge relation adjusted seasonally for ET (1.5 and 4.0 mm day−1 during the dormant and growing seasons, respectively) fit actual recessions extremely well. Copyright © 2011 John Wiley & Sons, Ltd.