This National Science Foundation sponsored hydrological research project conducted at the University of Wyoming examined the role of snow-dominated Rocky Mountain hillslopes in streamflow generation. Snowmelt from the Rocky Mountains is an important water source for agriculture, population centers, industry, and wetland ecosystems. The role of the subsurface in transforming spring snowmelt and summer rainfall into streamflow is poorly understood. The shallow and heterogeneous mountain soils and underlying (weathered) rock are difficult to sample, and the installation of hydrological monitoring equipment is labor intensive.

This project applied hydro-geophysical methods to image the subsurface and estimate structure, properties, and water flow dynamics using inversion techniques. Inversion can be understood as an iterative process wherein different subsurface configurations are tested for their ability to explain the geophysical and hydrological observations. This general approach was applied to seven mountain hillslopes, covering three different geologies (old granitic surfaces, young volcanic surfaces, and glacially altered surfaces), using shallow seismic refraction (SR) and time-lapse electrical resistivity tomography (ERT) data. In addition, for one of the study sites (in the Owyhee Mountains, ID), surface and airborne electromagnetic induction (EMI) data were used to determine subsurface properties in a three-dimensional hydrologic model to predict streamflow.

Relatively intact bedrock was detected within 5 m from the surface at hillslopes in the Laramie Range, WY (granitic), the Owyhee Mountains, ID (volcanic), and the Snowy Range, WY (glacially altered). Two hillslopes had only soil-like material within 5 m from the surface, namely in the Jemez Mountains, NM (volcanic) and the Bighorn Mountains, WY (glacially altered). Not surprisingly, no simple relationship between geological origin and subsurface make-up was found, given differences in past geological events, recent morphological processes, and local climate conditions.

Several inversion techniques were tested on selected sites using different geophysical and hydrological data sets. One study used time-lapse ERT, soil water content, and groundwater level data to estimate the subsurface hydraulic properties for two glacially altered hillslopes by assuming simple two-layer profiles. Using only groundwater level data to inform the inversion yielded better hillslope water storage and outflow predictions than when using only soil water content data. Using only time-lapse ERT resulted in reasonable outflow predictions but relatively poor hillslope water storage predictions with the two-layer structure being too simple to replicate the ERT observations.

A new hydro-geophysical inversion code was developed for the research project to allow for the estimation of fully distributed subsurface hydraulic properties using shallow SR or time-lapse ERT data, or a combination of these. Results for the glacially altered hillslope in the Snowy Range, WY indicated that replacing the two-layer approach with a fully distributed approach allows for a better replication of the ERT observations, and presumably, a more accurate prediction of the subsurface structure and hydraulic properties.

Application of a three-dimensional model at a volcanic headwater in the Owyhee Mountains, ID showed the potential of surface and airborne EMI data to inform hydrologic models. Here, past SR surveys and well core data were used to describe the headwater with five stratigraphic zones, each consisting of different materials (weathered basalt soil, loess soil, fractured basalt, semi-dense basalt, and altered basalt). The hydraulic properties of the different materials were obtained through inversion using different combinations of EMI, soil water content, groundwater level, and streamflow data. Interestingly, hydrologic model calibration with EMI and streamflow data gave reasonable predictions of streamflow and intra-headwater water storage dynamics, suggesting that surface and/or airborne EMI surveys may be an alternative to labor intensive and costly soil water content sensor and groundwater well installations. Future studies that do not rely on well core data to determine stratigraphy are needed to confirm the potential of EMI data in supplementing streamflow data for hydrologic model calibration.

Overall, this research project showed that subsurface configuration and its role in routing water to streams in mountainous regions is difficult to predict from information on geology alone. The project showed how geophysical measurement techniques such as SR, ERT, and EMI can be used to inform hydrologic models at the hillslope and headwater scales. Geophysical measurements provide additional information to models and have a different footprint than hydrological data. Having the ability to quantify subsurface structure, water storage dynamics, and outflow in a relatively cost-effective manner is important to improve our understanding of snow-dominated mountain ecosystems and their role in streamflow generation.

Last Modified: 09/05/2024
Modified by: Thijs J Keleners