Submitted to: Hydrological Processes
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/2/2008
Publication Date: 1/1/2009
Citation: Seyfried, M.S., Grant, L., Marks, D.G., Winstral, A.H., McNamara, J. 2009. Simulated Soil Water Storage Effects on Streamflow Generation in a Mountainous Snowmelt Environment. Hydrological Processes, 23, 858-873, doi:10.1002/hyp.7211.
Interpretive Summary: Snowmelt is the primary source of water in the intermountain west. In the past, estimates of water supply have been based strictly on estimates of the amount of snow. While this is the primary variable, soil water storage conditions (i.e., wet or dry) also affect the timing and amount of streamflow generated from snowmelt. In this paper we link a model simulating snow melt with one that simulates soil water processes. We verify the accuracy of both models with field measurements of snow depth and soil water content made over a two year period. We found that the combined modeling approach accurately described soil water movement at the measurement sites. More importantly, we found that, by expanding the modeling approach over the entire watershed, we could use soil properties to describe the initiation and cessation of streamflow. We think that this approach, when extended to larger scales, will provide information required for improved management of water that is in increasing demand.
Technical Abstract: Soil processes affect the timing and amount of streamflow generated from snowmelt and are often overlooked in estimations of snowmelt generated streamflow in the western USA. We investigated the use of a soil water balance modeling approach for incorporating the effects of soil processes, in particular soil water storage, on the timing and amount of snowmelt generated streamflow. The study was conducted in the Reynolds Mountain East (RME) watershed, a 38 ha, snowmelt dominated watershed in southwest Idaho. Snowmelt or rainfall inputs to the soil were determined using a well established snow accumulation and melt model (Isnobal). The soil water balance model was first evaluated at a point scale, using periodic soil water content measurements made over two years at 14 sites. In general, the simulated soil water profiles were in agreement with measurements as indicated by high R2 values, y-intercept values near 0, slopes near 1 and low average differences between measured and modeled values. In addition, observed soil water dynamics were generally consistent with critical model assumptions. Spatially distributed simulations over the watershed for the same two years indicate that streamflow initiation and cessation are closely linked to the overall watershed soil water storage capacity, which acts as a threshold. During periods of rapid streamflow response contributing cells are highly interconnected. Incorporation of soil water storage effects may potentially improve estimation of the timing and amount of streamflow generated from mountainous watersheds dominated by snowmelt.