by Alvar Escriva-Bou, Jay Lund, Josue Medellin-Azuara, and Thomas Harter

Earlier this year, the first local Groundwater Sustainability Plans (GSPs) were submitted to California’s Department of Water Resources for basins with the most severe groundwater overdraft.  To comply with the Sustainable Groundwater Management Act, these plans must address any “significant and unreasonable” impacts of groundwater overdraft that occurred after January 1, 2015, including lowering groundwater levels and other “undesirable results.” The math for ending overdraft is simple: groundwater basins must balance their budgets, by increasing groundwater recharge and reducing pumping.

In principle, evaluating the adequacy of these plans to achieve sustainability should also be simple: Does the anticipated reduction in pumping plus increase in recharge equal or exceed the basin’s long-term rate of overdraft?

In practice, however, it’s not so simple. Water supply in California is dynamic and variable, and groundwater overdraft can vary widely from year to year. In dry years more groundwater is pumped to replace reductions in surface water deliveries, while in wet years more water is recharged. Cropping patterns also vary annually with market conditions, agronomic conditions (fallowing, salinity, etc.), and labor availability, all of which can affect regional water use. Even estimates of long-term average overdraft have considerable uncertainty, due to California’s hydrologic variability and uncertainties in data and models used to estimate water balances. For instance, model estimates of average statewide groundwater overdraft for different periods range from 1.4 to 9.1 million acre-feet (maf) per year (Brush et al., 2013; Escriva-Bou, 2019; Famiglietti et al., 2011; Faunt et al., 2009; Xiao et al., 2017). Considering all these uncertainties, plans based on rigid and static assumptions are bound to miss their targets.

So how do we know if a proposed Groundwater Sustainability Plan will succeed? Or, put another way, how can we plan for sustainability given all these uncertainties?

Although we can’t ever be sure, we can make reasonable estimates that account for the range of uncertainties. In a recent paper, we outline a method for estimating the likelihood of achieving groundwater sustainability. This method highlights the relationship between restrictions on groundwater pumping and the likelihood of maintaining groundwater levels, trade-offs between pumping restrictions and agricultural revenue losses, and implications of defining thresholds—such as minimum groundwater levels—as part of sustainability plan rules.

Our analysis highlights the dynamic character of hydrology and water use in California’s Central Valley. Although we illustrate the method using regional modeling estimates of overdraft and hydrology that were available prior to the release of the new GSPs, the analysis is relevant for groundwater sustainability agencies as they implement and refine their plans going forward. The method can be applied to help estimate uncertainties in these plans, reduce these uncertainties over time, and plan adaptively for sustainability, considering the economic consequences of varying pumping restrictions and thresholds.

Here we highlight some of the paper’s main lessons.

Greater pumping restrictions increase the likelihood of ending overdraft

Uncertainty in overdraft arises from the state’s highly variable climate, uncertainties in future water use decisions, and imperfections in models and data representing water balances.

We estimated uncertainty in annual overdraft for 21 Central Valley sub-regions using the results of recent hydrologic history (1975-2003) of two Central Valley groundwater models (USGS’ CVHM and DWR’s C2VSIM).  Each model’s results show variability in annual water balances and overdraft. The differences between these models’ results can represent the uncertainty of data and models.

Our analysis directly combines both sources of uncertainty. A simple approach is to estimate average annual overdraft as the average of model results, and estimate the variability of annual overdraft from the standard deviation of these results.  A simple statistical equation can be used to estimate the probability of accumulated future overdraft for a given net reduction in pumping over a specified number of years. By defining sustainability as the inter-annual fluctuation of the groundwater table that does not exceed a minimum threshold (Figure 1), it is possible to estimate the probability of achieving sustainability as a function of the net reduction in groundwater use.

Figure 1: Hypothetical representation of changes of groundwater storage over time and definition of minimum thresholds to define sustainability

Figure 2 shows results for all 21 Central Valley sub-regions. A net reduction in water use equaling the average level of overdraft has a 50% chance of achieving sustainability in any time period. Greater reductions in net water use increase the probability of ending overdraft, although the rates (the slope of the curves) vary across sub-regions depending on local conditions. For example:

• Sub-regions that are sustainable (but that could become unsustainable with unfavorable conditions). Sub-regions 9 (Delta) and 12 (Turlock Basin) are sustainable on average, with positive average annual change in groundwater storage. Without net water use reductions their chances of balancing groundwater levels in 20 years exceed 70%, but with unfavorable hydrologic conditions they would have to reduce net water use to ensure sustainability.
• Sub-regions with low levels of unsustainable pumping. Sub-regions 10 (Delta-Mendota Basin) and 11 (Modesto and southeast San Joaquin Basin) have some average annual overdraft (< 10 taf/year). Without net water use reduction, they would have almost a 50% chance achieving sustainability over next 20 years. To ensure sustainability for any future hydro-climatic conditions, they would have to reduce water use by much more (~ 80 taf/year).
• Sub-regions with large values of unsustainable pumping. Sub-regions 8 (the valley floor east of the Delta) and 13 (Merced, Chowchilla, and Madera Basins) have larger average annual overdraft (55 and 116 taf/year respectively). Their larger variances also make the curves flatter. These sub-regions still have a 20% chance of becoming sustainable without reducing water use. Net use reductions at least three times the current level of overdraft are needed to reliably achieve sustainability.

For longer regulatory horizons, the range of pumping reductions for different probabilities narrows around the estimated average annual overdraft because the statistical average becomes more likely to dominate over rare favorable or unfavorable hydrology over longer periods.

Two caveats are worth mentioning. First, this analysis assumes all overdraft is addressed by net reductions in groundwater use. Smaller reductions will be needed if groundwater recharge can be increased in wet seasons and years. Some estimates find that 15-30% of long-term overdraft in the Central Valley could be supplied by additional groundwater replenishment (DWR 2018; Kocis and Dahlke 2017, Escriva-Bou and Hanak 2018). This could significantly reduce the economic impact of pumping restrictions. Second, our results are based on DWR’s C2VSim and USGS’s CVHM models, developed for the Central Valley. These models may have some difficulties representing local groundwater basins—the relevant geographic focus for GSPs. Even if GSPs are able to incorporate more accurate local estimates of overdraft and historic hydrology, it will still be important for these plans to estimate the uncertainties of these and other parameters in evaluating the likelihood of attaining groundwater sustainability.

Figure 2. Probability of ending groundwater overdraft in year 2040 for a range of net groundwater pumping reductions for all Central Valley sub-regions

Increasing the likelihood of achieving sustainability rapidly raises agricultural economic losses

We also estimated economic costs and probabilities of achieving sustainability over different periods. Here, we assumed the sub-regions achieve sustainability only by reducing agricultural water use, with internal water markets within each sub-region.

Trade-off curves of the reliability of achieving sustainability versus economic costs of reducing farm water use follow an S curve (Figure 3). Small water use reductions have low likelihood of achieving sustainability, and lower economic costs for agriculture. Achieving sustainable groundwater levels with higher reliability requires more pumping reductions and much higher economic costs, because this requires curtailing water for more valuable crops. For example, the average cost to achieve a 50% chance of groundwater sustainability in sub-region 15 over 20 years is $10 million/year. However, to achieve nearly 100% reliability of sustainability would cost roughly six times more in lost crop revenues. Much longer compliance horizons lessen the tradeoff between farm sector costs and reliability of attaining sustainability.

Figure 3: Higher likelihood of achieving groundwater sustainability reduces farm revenues (example for sub-region 15)

An important caveat is that we only account for agricultural revenue losses from reducing overdraft in this study. Plans would also need to consider costs of not reducing overdraft, including infrastructure damage related to land subsidence, capital costs of stranded wells, higher pumping costs, dry drinking water wells, and harm to groundwater-dependent ecosystems.

Some lessons

1. Initial overdraft estimates will almost certainly be wrong. Uncertainties in data and models should be considered when developing overdraft estimates for Groundwater Sustainability Plans, and the plans should provide ranges of estimates. For instance, overdraft estimates for different historical periods could show the potential range of hydrologic variability, and the implications of this variability for attaining sustainability.
2. Groundwater pumping restrictions will need to vary across years. To accommodate future droughts within the compliance horizon for attaining sustainability, required reductions in water use will need to vary across years.  Groundwater demand is generally higher during droughts, when less surface water is available. As droughts progress and groundwater is drawn down, pumping shares should be adjusted to avoid reaching the minimum threshold for groundwater levels too early, potentially causing significant investment losses in perennial crops.
3. Long-term pumping and recharge efforts will likely need to adjust over time. Anticipating and building such adjustments into local groundwater plans should make it easier to implement plans and to anticipate changes in groundwater availability, as well as make it easier for state agencies to evaluate plans.
4. Successful groundwater sustainability plans and institutions will analyze uncertainty to prepare to adapt to variable hydrology and new estimates of overdraft. Adaptive methods to account for California’s hydrologic variability and the inherent uncertainties are essential for a dynamic response to changing conditions. SGMA’s requirement that GSPs update their plans at least every five years provides an important opportunity to incorporate new information and lessons, and adapt to changing conditions.
5. State regulators can help by providing common regional templates, data, and models for analysis and institutional response. To facilitate the transition to stronger sustainability accounting—incorporating uncertainties—state technical assistance should focus on standardizing and improving the quality of the data, documentation, modeling, and other key tools for GSP planning and implementation. DWR’s Draft Handbook for Water Budget Development (2020) may be a step in this direction.

Data availability and replicability

The data and code to replicate all the analyses included in this paper are available at https://github.com/alesbou/GWSustainabilityUncertainty-Costs

Further reading

Escriva-Bou, A., R. Hui, S. Maples, J. Medellín-Azuara, T. Harter, and J. Lund (2020) “Planning for Groundwater Sustainability Accounting for Uncertainty and Costs: an Application to California’s Central Valley,” Journal of Environmental Management, Volume 264, 110426, https://doi.org/10.1016/j.jenvman.2020.110426.

Brush, C.F., Dogrul, E.C., Kadir, T.N., 2013. Development and Calibration of the California Central Valley Groundwater-Surface Water Simulation Model (C2VSim), Version 3.02-CG. California Department of Water Resources, Bay-Delta Office.

DWR (California Department of Water Resources) 2018 Water Available for Replenishment Report (Sacramento CA: California Department of Water Resources) (Now seemingly unavailable from the DWR website)

DWR (California Department of Water Resources), 2020. Draft Handbook for Water Budget Development. With or Without Models.

Escriva-Bou, A., and Hanak, E. (2018), Appendix A: Update of the San Joaquin Valley’s Water Balance and Estimate of Water Available for Recharge in 2017. In Hanak et al (2018) Replenishing Groundwater in the San Joaquin Valley. PPIC.

Escriva-Bou, A. (2019), Appendix A: Update Assessment of the San Joaquin Valley’s Water Balance. In Hanak et al (2019) Water and the Future of the San Joaquin Valley, PPIC

Famiglietti J S, Lo M, Ho S L, Bethune J, Anderson K J, Syed T H, Swenson S C, de Linage C R and Rodell M 2011, “Satellites measure recent rates of groundwater depletion in California’s Central Valley,” Geophy. Res. Lett. 38

Faunt C C 2009 Groundwater availability of the central valley aquifer, California, USGS Professional paper No. 1766. https://pubs.usgs.gov/pp/1766/

Kocis, T.N. and H.E. Dahlke, 2017. Availability of high-magnitude streamflow for groundwater banking in the Central Valley, California. Environmental Research Letters. https://doi.org/10.1088/1748-9326/aa7b1b

Orth, D. “SGMA IMPLEMENTATION: David Orth gives his observations on how sustainable groundwater management is playing out in the San Joaquin Valley,” summarized by Maven March 19, 2020 https://mavensnotebook.com/2020/03/19/sgma-implementation-david-orth-gives-his-observations-on-how-sgma-implementation-is-playing-out-in-the-san-joaquin-valley/

Scanlon B R, Faunt C C, Longuevergne L, Reedy R C, Alley W M, McGuire V L and McMahon P B 2012 Groundwater depletion and sustainability of irrigation in the US high plains and central valley Proc. Natl Acad. Sci. 109 9320–25

Xiao M, Koppa A, Mekonnen Z, Pagán B R, Zhan S, Cao Q, Aierken A, Lee H and Lettenmaier D P 2017 How much groundwater did California’s Central Valley lose during the 2012–2016 drought? Geophy. Res. Lett. 44 1–8

Alvar Escriva-Bou is a Research Fellow with the Public Policy Institute of California. Jay Lund is the Director of the Center for Watershed Sciences and a Professor of Civil and Environmental Engineering at the University of California, Davis. Josue Medellin-Azuara is an Associate Professor of Environmental Engineering at the University of California, Merced. Thomas Harter is a Professor of Hydrology and Cooperative Extension Specialist at the University of California, Davis.