Estimating the deep seepage component of the hillslope and catchment water balance within a measurement uncertainty framework
Article first published online: 19 JUL 2010
Copyright © 2010 John Wiley & Sons, Ltd.
Volume 24, Issue 25, pages 3631–3647, 15 December 2010
How to Cite
Graham, C. B., van Verseveld, W., Barnard, H. R. and McDonnell, J. J. (2010), Estimating the deep seepage component of the hillslope and catchment water balance within a measurement uncertainty framework. Hydrol. Process., 24: 3631–3647. doi: 10.1002/hyp.7788
- Issue published online: 26 NOV 2010
- Article first published online: 19 JUL 2010
- Manuscript Accepted: 10 MAY 2010
- Manuscript Received: 15 SEP 2009
- hillslope hydrology;
- water balance;
- experimental uncertainty;
- deep seepage;
- sprinkling experiments
Deep seepage is a term in the hillslope and catchment water balance that is rarely measured and usually relegated to a residual in the water balance equation. While recent studies have begun to quantify this important component, we still lack understanding of how deep seepage varies from hillslope to catchment scales and how much uncertainty surrounds its quantification within the overall water balance. Here, we report on a hillslope water balance study from the H. J. Andrews Experimental Forest in Oregon aimed at quantifying the deep seepage component where we irrigated a 172-m2 section of hillslope for 24·4 days at 3·6 ± 3 mm/h. The objective of this experiment was to close the water balance, identifying the relative partitioning of, and uncertainties around deep seepage and the other measured water balance components of evaporation, transpiration, lateral subsurface flow, bedrock return flow and fluxes into and out of soil profile storage. We then used this information to determine how the quantification of individual water balance components improves our understanding of key hillslope processes and how uncertainties in individual measurements propagate through the functional uses of the measurements into water balance components (i.e. meteorological measurements propagated through potential evapotranspiration estimates). Our results show that hillslope scale deep seepage composed of 27 ± 17% of applied water. During and immediately after the irrigation experiment, a significant amount of the irrigation water could not be accounted for. This amount decreased as the measurement time increased, declining from 28 ± 16% at the end of the irrigation to 20 ± 21% after 10 days drainage. This water is attributed to deep seepage at the catchment scale. Copyright © 2010 John Wiley & Sons, Ltd.