Drivers of spatiotemporal patterns of surface water inputs in a catchment at the rain-snow transition zone of the water-limited western United States

Authors: HALE, KATE - University Of Colorado; KIEWIET, LEONIE - Colorado State University; TRUJILLO, ERNESTO - Boise State University; KROHE, CLARISSA - Ramboll; Hedrick, Andrew; MARKS, DANIEL - Retired ARS Employee; KORMOS, PATRICK - National Oceanic & Atmospheric Administration (NOAA); Havens, Scott; MCNAMARA, JAMES - Boise State University; LINK, TIMOTHY - University Of Idaho; GODSEY, SARA - Idaho State University

Submitted to: Journal of Hydrology
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/20/2022
Publication Date: 11/10/2022
Citation: Hale, K., Kiewiet, L., Trujillo, E., Krohe, C., Hedrick, A., Marks, D., Kormos, P., Havens, S.C., McNamara, J., Link, T., Godsey, S. 2022. Drivers of spatiotemporal patterns of surface water inputs in a catchment at the rain-snow transition zone of the water-limited western United States. Journal of Hydrology. 616. Article 128699. https://doi.org/10.1016/j.jhydrol.2022.128699.
DOI: https://doi.org/10.1016/j.jhydrol.2022.128699

Interpretive Summary: The agricultural water supply in the Western U.S. is inextricably tied to when, where, and how much rainfall and snowmelt reach the ground surface. The seasonal snowpack acts as a natural water reservoir and varies substantially from year to year. In semi-arid mountainous rangelands, which cover a large area of the West, the annual snowpack can alternate between a deep, healthy snowpack in one year to an intermittent snowpack in the next. This study uses an advanced snow hydrology model called iSnobal over two years to determine how watersheds may react under future warmer conditions. Results show that warm wintertime storms in which rain fell on the snowpack produced large pulses of water input to the watershed and that most water input occurred after peak snow accumulation but before the snowpack had melted. In addition, snow drifts at the top of the watershed held and released twice as much water than scoured areas – areas that in turn received less precipitation than the non-drift areas lower in the watershed. The major finding of this study is that the annual water supply is heavily controlled by the presence of snow drifts, while the seasonal or event-based water supply is controlled by the energy inputs to the watershed.

Technical Abstract: When, where and how much rainfall and snowmelt (i.e., surface water inputs or SWI) reaches the ground surface controls fundamental hydrologic variables such as soil moisture, groundwater recharge and streamflow. Quantifying the spatial and temporal distribution of SWI and its drivers at the annual, seasonal and event scale can therefore yield valuable insights in catchment water resources. This is especially relevant at the rain-snow transition zone, an area that comprises a large portion of the western United States, and may also serve as a climate warming indicator for currently snow-dominated catchments that could shift toward a transitional or intermittent snowpack. To evaluate SWI distribution, we modeled rainfall and snowpack dynamics in a headwater catchment spanning the rain-snow transition in southwestern Idaho, USA. Precipitation magnitude, at the pixel-scale, drove spatial patterns of annual SWI and differences in SWI magnitude within and across two modeled water years (“wet” 2011 and “dry” 2014). Snow drifts generated more SWI (901-2080 mm) than high-elevation scour locations (442-640 mm) in both years. Scour zones received slightly less precipitation, and thus generated less SWI, than mid-elevation, non-drift locations (452-784 mm). Temporal energy flux patterns in both years were similar, with an average 15% difference in magnitude, across aspects and elevations until the melt season. Snowmelt was driven by higher net radiation at lower elevations and south-facing slopes and by higher turbulent fluxes at higher elevations and north-facing slopes in the catchment. A total of 15-20% of annual SWI occurred during rain-on-snow (ROS) events, and most excess SWI was produced, on average, after peak snow water equivalent (SWE), when rainfall fell onto relatively warm snowpacks with relatively high liquid water content. Before peak SWE, ROS events generated more SWI at lower elevations and on south-facing slopes, whereas drifts generated up to 30 mm of total SWI during ROS events occurring after peak SWE. Catchment water resources will thus depend on SWI magnitude, location, and timing; which are moderated by drift persistence at the annual scale and seasonal and event energetics at shorter time-scales.