COMPARING MODELED AND OBSERVED GROUNDWATER RESIDENCE TIMES IN SUPPORT OF SUSTAINABLE GROUNDWATER MANAGEMENT IN BUTTE BASIN, CALIFORNIA

California’s 2014 Sustainable Groundwater Management Act (SGMA) provides opportunity for innovative approaches of assessing groundwater models and water budgets. Advanced models calibrated with groundwater elevation alone may not accurately reflect residence times or aquifer storage turnover times. Uncertainty remains over recharge and seepage rates to deep aquifers. This study compares isotopic signatures in groundwater to examine residence times with output from the Butte Basin Groundwater Model (BBGM).

Our study focuses on production and monitoring wells screened over the heavily pumped Tuscan Formation in Butte Basin. Groundwater with a subsurface residence time of less than 50 years is dated using the ³H-³He method, and estimates of longer residence times are estimated using radiogenic ⁴He and ¹⁴C.

Independent estimates of residence times are based on output from BBGM, Butte County’s version of the California Department of Water Resource’s Integrated Water Flow Model (IWFM) that is calibrated using groundwater elevation. We applied the Z-budget subroutine of IWFM that allows examination of the water budget over specific layers that comprise the Tuscan Formation. The output budget from the model gives the volume of water in storage in each aquifer layer and the inflow and outflow rates; therefore, the mean residence time for each model layer, over a subregion, is readily calculated.

The mean ³H-³He groundwater ages and modeled residence times are in approximate agreement, with residence times of one to five decades in the heavily pumped portion of the basin. However, about half of the Tuscan wells produce water that recharged entirely over 70 years ago, based on tritium levels below the detection limit. Many of these wells produce fossil water, with ⁴He concentrations and ¹⁴C activities that indicate groundwater ages of 2000 to 5000 years.

Using the model residence times, and assuming an exponential age distribution, this translates to 0 to 28% of all water being older than 70 years for the different model subregions, a fraction significantly less than determined from isotopic ages (50%). More importantly, the exponential age distribution predicts no water older than 500 years, whereas the isotopic data indicates fossil water in 32% of the wells. This difference in model and isotope-derived ages and turnover times provides an opportunity to improve the model calibration.