Evaluation of Streamflow Depletion Related to Groundwater Withdrawal, Humboldt River Basin


Streamflow Depletion Maps

Streamflow depletion maps (generically known as capture maps) are a mapping tool designed to help characterize the effects of groundwater withdrawal on the timing and rates of streamflow depletion. Depletion maps are created through repeated simulations of a calibrated groundwater flow model, where each simulation computes the streamflow depletion as a result of pumping at varying locations over time.

The middle and lower HRB models will be used in sequence to evaluate streamflow depletion based on changes in streamflow at the Imlay gage caused by upstream groundwater withdrawals. The lower HRB model will develop streamflow depletion maps based on changes in Humboldt River water budgets along entire reach below Rye Patch reservoir. Streamflow depletion maps will be developed by:

  • Starting with reference scenario (no pumping).
  • Systematically pumping one model cell at a time.
  • Evaluating change in streamflow (stream depletion) as result of pumping from each cell for durations of 10, 25, 50, and 100 years.
  • Developing a contour map of stream depletion for each pumping duration evaluated (depletion map).

The complexity of the Humboldt River Basin flow system may present unique challenges that can limit the practicality and accuracy of direct interpretation of streamflow depletion from standard streamflow depletion map analysis. This is due to nonlinearities introduced by other head-dependent flow processes, in particular evapotranspiration of shallow groundwater from phreatophytic areas away from the river (this does not include riparian ET), and simulation of unconfined aquifer conditions. Depletion maps are developed by simulating withdrawals in only one model cell at a time. Because of this, depletion maps are likely to represent more efficient capture of groundwater evapotranspiration and less efficient capture of streamflow than is realized by the combined influence of all pumping from areas in which evapotranspiration is a component of the overall capture. This limitation pertains more to areas distant from the Humboldt River and in particular, to areas with shallow groundwater and groundwater evapotranspiration discharge between pumping locations and the river.

One approach for addressing this limitation is to evaluate and present the bias associated with estimation of streamflow depletion from the developed depletion maps. This is done by estimating streamflow depletion for existing distribution of pumping using the streamflow depletion maps and comparing with the streamflow depletion determined from the calibrated model. The difference between these two methods of determining streamflow depletion is the bias associated with use of a non-linear model to develop the depletion map. This bias is a function of the aquifer properties, depth to water table, and distribution of actual pumping. It is anticipated that the overall bias associated with use of non-linear models to develop streamflow depletion maps will be relatively small (<10%). However, this is an area of active research and therefore will be further researched and documented as part of this project.

The bias associated with use of non-linear flow models in determination of streamflow depletion will be researched in collaboration with the Desert Research Institute using a variety of test models that have similar representation to many of the valleys within the Humboldt River system. The test models will simulate groundwater flow in systems containing rivers (and/or streams), groundwater evapotranspiration from shallow groundwater areas adjacent to rivers (riparian ET) and away from rivers (phreatophyte ET) and a variety of pumping distributions and rates. Bias associated with use of streamflow depletion maps to estimate streamflow depletion will be summarized and then methods applied in a real-life example using an existing model of the Fernley area in Nevada. The Fernley area groundwater flow model was developed and documented by the Desert Research Institute in Stevick and others.