The USGS Groundwater Model for the Owens Valley

ICWD Home > Research and Reports > Annual Reports > 2004-2005 Monitor >

by Bob Harrington, Hydrologist
August 2005

In a 19th century decision regarding groundwater rights, the Ohio Supreme Court decreed that "Because the existence, origin, movement, and course of such waters, and the causes which govern and direct their movements are so secret, occult, and concealed that any attempt to administer any set of legal rules in respect to them would be involved in hopeless uncertainty, and therefore would be practically impossible." Since then, much has been learned about the physical processes affecting groundwater, and there now exist methods that allow for the evaluation of the movement of groundwater. Among the most useful of the modern tools for evaluating groundwater levels and flows are groundwater flow models.

Groundwater levels respond to inputs and withdrawals of water from the groundwater system. For example, when it rains on the valley floor or when springtime snowmelt runoff percolates into the alluvial fans flanking the Sierra Nevada, water levels rise. Groundwater withdrawals, such as pumping of groundwater or water use by plants, cause groundwater levels to decline. The balance of inputs and withdrawals determines whether groundwater levels rise or fall, much like the balance of deposits and withdrawals determines whether a bank account rises or falls. The nature of the response of the groundwater system to inputs and withdrawals is determined by the groundwater flow properties of the materials that the groundwater moves through. For example, water flows more readily through sand and gravel than through clay, so a groundwater pumping well will preferentially withdraw water from sand and gravel layers rather than clay layers.

In order to make sense of all the inputs, withdrawals, and flow properties in a groundwater flow system, hydrologists use a groundwater flow model to simulate the water levels and movement of groundwater. Groundwater flow models are computerized renditions of groundwater flow systems, incorporating the basic physical principles of groundwater flow and information about the inputs, withdrawals, and flow properties of the particular groundwater system under study. Computer simulations of environmental systems can never completely capture the complexity of the real world, and therefore are never perfectly accurate. Though we have some information from monitoring of flows and from tests done on wells, we never have complete or exact knowledge of what the inputs, withdrawals, and flow properties of a groundwater system really are. Design of groundwater models necessitates a certain amount of guesswork about how the groundwater system works, and the results generated by the model have to be interpreted with consideration of the assumptions and uncertainties included in the model design.

Figure 1: Shown above are modeled and observed hydrographs of well 419T. Well 419T is located east of Aberdeen near the LA Aqueduct intake. Its observed hydrograph, the thick line, shows water table fluctuations that are typical of the past thirty years in the Owens Valley: a depressed water table during the late 1970's, a peak in the mid-1980's, a rapid decline in the early-1990's, followed by gradual recovery to a peak in the late-1990's, followed by a gradual decline. The missing parts of the observed hydrograph are due to the well being dry. The modeled hydrograph, the thin line with circles, is consistently offset a few feet higher than the observed hydrograph, but the modeled year-to-year changes correspond closely to the observations. Thus, at well 419T, while the model produces water levels that are erroneous by a few feet, it reproduces the response to pumping stresses and recharge quite faithfully. The errors are present for two reasons: first, the model is an imperfect and approximate representation of reality, and second, the model represents average water levels over a 2000 by 2000 foot square, which need not be exactly the same as at the location of well 419T.

The period from 2003 to 2020 was modeled using mean runoff conditions and no pumping. Notice that the water table is intermittently modeled as being above the land surface - this is not entirely unrealistic, as this area was formerly the site of significant groundwater discharge (the presently dry vent of Hines Spring is located nearby). Rather than ponding to the depths indicated by the modeled hydrograph, the emerging groundwater would flow down-slope in the many sloughs and channels in the area forming shallow ponds and wetlands.

As part of a joint Inyo County/LADWP cooperative study, the U.S. Geological Survey (USGS) constructed a groundwater model for the Owens Valley in the 1980's for the purpose of examining different groundwater management scenarios. This model simulated the groundwater system for the period 1963 through 1988. Inyo County, working with LADWP and the USGS, has recently completed an update of the USGS model to make it into a tool for evaluating current and future groundwater levels. Tasks completed as part of the update of the USGS model include extending the range of the model out to the year 2020, implementing a more detailed accounting of the groundwater inputs and withdrawals, and examination of five scenarios designed to verify that the model was performing correctly. Figure 1 compares the USGS model's prediction of water table elevation to observations in a shallow monitoring well (well 419T) near the LA Aqueduct intake, a few miles east of Aberdeen. Though the model does not exactly reproduce the observed water table elevations, clearly the model provides us with a general prediction about how the water table responds to hydrologic stresses. The results of the update generally strengthen the conclusions from the initial model development effort:

The primary influence on water levels in most areas of the Valley is groundwater pumping.
The time frame for the groundwater system to respond to regional stresses is five to ten years. This means that there is a time-lag of as much as a decade before the groundwater system completely responds to a change in natural conditions, such as the amount of runoff, or a change in management, such as a change in pumping.
Pumping is a relatively short-term and localized effect, whereas recharge from runoff is a relatively diffuse and long-term effect.

The updated model promises to be a valuable tool for both Inyo County and LADWP. The effects of a proposed annual pumping plan, a proposed well or well test, or a change in the volume of runoff can be evaluated using this model. Additionally, retrospective questions, such as what has been the effect of the Drought Recovery Policy, or alternative futures, such as what would be the effect of stopping groundwater pumping (Figure 1), can be analyzed. Groundwater science has progressed since the 19th century, and no longer is the movement of groundwater considered "secret, occult, and concealed;" groundwater flow modeling is a key tool in the analysis of groundwater movement in the Owens Valley.