Summary

Recharge is a vital part of the ground-water budget. As competition grows for limited water resources, water managers increasingly look to the ground-water system as a source for possible development. Most available methods to estimate ground-water recharge depend on data that are generally unavailable or difficult to obtain. Methods are also limited by the application of scale. Methods either simulate at a point or site scale or simulate a large area as a single value, thereby limiting the ability to accurately scale down to a local area or distribute the recharge estimates spatially.

Process-based models that compute distributed water budgets on a watershed scale have demonstrated an ability to calculate accurately recharge rates at varying scales using readily available databases. These watershed models should be evaluated to determine what parameters have a dominant control of the estimated recharge rates.

The U.S. Geological Survey analyzed the sensitivity of estimated ground-water recharge to parameters in seven existing watershed models in different humid regions of the United States to gain an understanding of the watershed-model parameters that control recharge estimates. A nonlinear regression model, UCODE, was coupled with the Modular Modeling System and Precipitation-Runoff Modeling System to generate a suite of diagnostic statistics. Dimensionless scaled sensitivities, composite scaled sensitivities, and parameter correlation coefficients tested the sensitivity of simulated recharge to parameter values and the ability of two parameters to be uniquely estimated. The objectives of the study were to determine (1) which watershed-model parameters were the dominant controls in determining recharge, (2) if regional differences existed in the sensitivity of recharge to watershed-model parameters, (3) if specific computer models used to simulate recharge affect parameter sensitivities, and (4) if objectives and approach of a study can affect ground-water recharge estimates and parameter sensitivities.

Simulated recharge in the Lost River and Big Creek watersheds, in Washington State, was sensitive to small changes in air temperature. The precipitation falls predominantly as snow, and the spring snowmelt produces much of the recharge and streamflow. A change of 0.3 degree Fahrenheit in the simulated maximum daily temperature would affect the form of precipitation (rain or snow) on 20 days in the Lost River watershed and 25 days in the Big Creek watershed over the 3-year simulation period.

The model for the Hamden watershed, in west-central Minnesota, was developed to investigate the relations that wetlands and other landscape features have with runoff processes. Different modules for wetlands were compiled to achieve this objective. Excess soil moisture was preferentially routed to wetlands instead of to the ground-water system, resulting in very little sensitivity of any parameters to recharge.

Recharge in the North Fork Pheasant Branch watershed, Wisconsin, was most sensitive to parameters related to evapotranspiration (ET). Recharge occurs during winter and spring, when snow is melting, or during autumn, when ET is reduced. Recharge is limited in summer, when ET is greatest.

Three watersheds were simulated as part of the Model Parameter Estimation Experiment (MOPEX). Amite River near Denham Springs, Louisiana; English River at Kalona, Iowa; and South Branch Potomac River near Springfield, West Virginia, serve as benchmark watersheds to address the parameter uncertainty of ungaged watersheds and assess new techniques to decrease the uncertainty. Parameter sensitivities for the three MOPEX watersheds were remarkably similar. Tmax_allsnow and soil2gw_max, a user-defined flow-routing parameter, were the dominant controls for recharge.

Although the primary objective of this study was to identify, by geographic region, parameters important to ground-water recharge simulation, the secondary objectives proved to have valuable applications in developing future models. The value of a rigorous sensitivity analysis can (1) make the calibration process more efficient, (2) guide additional data collection, (3) identify model limitations, and (4) help explain simulated results.

Back to Table of Contents

For more information about USGS activities in Washington, visit the USGS Washington Water Science Center home page .

Accessibility FOIA Privacy Policies and Notices

U.S. Department of the Interior | U.S. Geological Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065041
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 18:55:21 EST