Dry fields and bare trees at Panoche Road, looking west, on Wednesday February 5, 2014, near San Joaquin, CA. Photo by Gregory Urquiaga/UC Davis 2014

by Brad Arnold¹, Alvar Escriva-Bou², and Jay Lund¹

¹ UC Davis Center for Watershed Sciences

² Public Policy Institute of California

Passage of the Sustainable Groundwater Management Act (SGMA) and the recent drought have brought attention to chronic shortages of water in the San Joaquin Valley. Although the portfolio of water flows available to the Valley is diverse, several major inflows – including groundwater use, Delta imports, and local streamflows – are unsustainable or threatened by climate change and environmental demands. Here we examine long-term balances for San Joaquin Valley’s water supplies and demands that we discussed in a prior blog post.

In addition to the San Joaquin Valley’s substantial long-term water imbalance, many of its individual water supplies are highly variable and involve substantial operational, regulatory, and planning uncertainties. Not only is there growing concern for water scarcity for the San Joaquin Valley, the character and causes of this scarcity are highly and unavoidably variable and uncertain. We look at how major flows into and from the Valley vary and the uncertainty in such water balance estimates, with some policy and management implications.  

Large Variations in Valley Water Supplies and Use

The San Joaquin Valley’s water supply portfolio varies each year. Precipitation and river runoff from the central and southern Sierra Nevada and Coastal Range all vary greatly each year. Changing water conditions, such as those in recent droughts and wet years, provide a stark reminder of their immense variability.

The San Joaquin Valley. Map prepared by USGS.

A typical year’s ‘natural availability’ of river inflows averages about 10.1 MAF/year, roughly 56% of all Valley inflows from 1986 to 2015. These total river flows vary substantially each year, from 3.0 MAF in a dry year (2015, minimum) to 21.9 MAF in a wet year (1998, maximum); typical variation of approximately 5.6 MAF from the average in any given year. Other Valley inflows also vary annually – specifically SWP and CVP Delta Imports – and this erraticism of inflows can cause major impacts to the water balance in any single year (especially affecting San Joaquin River flows and groundwater pumping).This variability brings planning and management challenges and uncertainties for the Valley’s water resources.

When surface water inflows, outflows, and uses are “out of balance,” the difference typically comes from drawdown or refill from groundwater or surface reservoirs. Reservoir storage generally follows a seasonal refill and draw-down pattern – with additional refilling in wetter years balancing drawdowns in drier years. Groundwater storage shows a seasonal drawdown-refill cycle, and much larger imbalanced drawdowns in drier years (averaging about 1.8 MAF/year of San Joaquin Valley overdraft). The plot below compares annual natural availability data to other annual outflow and inflow for the San Joaquin Valley.

Natural nvailability data plotted with outflows and uses for 30-year period.

From linear regression lines for these scatter plots, an average annual decrease of 500 thousand acre-feet in Valley-wide natural availability (decreased Sierra Nevada, Coastal Range, and precipitation flows) generally coincides with the changes presented here:

Approximate changes to San Joaquin Valley outflows and uses
resulting from changes in natural availability (for the period 1986-2015).

These statistical regression slopes – constrained so slopes close the water balance – illustrate the relative trends and magnitudes of water flows and management responses to wet and dry years in the San Joaquin Valley. In drier years the major response is much greater groundwater pumping, reduced San Joaquin River outflows, somewhat decreased agricultural and urban net water use, and more withdrawals from surface reservoirs. In wetter years, many of these responses reverse.

Balanced water management in California must prepare to operate across diverse wet and dry year conditions. Large outflow and water use changes, and long-term groundwater overdraft, are clearly seen, including recent droughts – where natural availability averaged 2.3 MAF/year less for 2007 to 2015 than its 30-year average. If natural availability becomes more variable, as predicted with climate change, the result will likely be larger fluctuations in the Valley’s water balance. Valley inflows, outflows, and uses vary greatly, and vary together.

Large Statistical Uncertainty in Major Inflows

Estimates of individual inflows or outflows are more straightforward, but always include some error from measurement inaccuracy, modeling uncertainty, and hydrologic variability. The “standard deviation of the average” is an index of uncertainty in average flows based on statistical chance the “actual” average flow differs from the historical sample average. Long-term averages are statistically better-behaved than estimates for a single year. However, if climate and underlying hydrologic processes are changing, errors may become less well-behaved.

Average flow estimates, such as those for the San Joaquin Valley water balance, are the basis for many water policy, planning, management, and regulation decisions. Averages are central for implementing the water budgets required by SGMA, estimating impacts for an environmental flow policy, and decisions to invest in water infrastructure or to trade water. But such averages often mask great variability and uncertainty. Given the economic value of this water, and how economic valuations ($/acre-ft) change from dry to wet periods, errors and variability from averages are important.  

Here we look at uncertainty in annual inflows from the Central and Southern Sierra Nevada – the origins of most natural availability for the San Joaquin Valley. Confidence in these flows is important to analyzing water demands and uses (especially for calculating groundwater overdraft as a water balance “closure term”). Technical Appendix A of the recent PPIC report details the river/inflow selection process and data sources.

Uncertainty range of flows for major San Joaquin Valley surface inflows. Average plus and minus standard deviation of annual flows and (smaller) standard deviation of average annual flows.
Second columns do not include recent drought years 2007 to 2015 data.

These data show important results: total standard deviation in average central Sierra Nevada inflows is about 1.74 MAF/yr, and southern Sierra average inflows have a standard deviation of 600 TAF/yr. This means there is a 33% chance the ‘actual’ average flows may deviate more than these amounts from the historical averages, either more or less. Neglecting recent drought years from 2007 to 2015 doesn’t necessarily reduce this uncertainty. There is substantial uncertainty in published “Full Natural Flow” estimates, both in single year and long term average values. If climate model projections are correct, climate change could make this long-term uncertainty even greater than uncertainty estimates based only on historical statistics alone.

Regional water balances have large uncertainties and intricate internal variabilities. Planning around water balances often overlooks these uncertainties. Three important implications from these unavoidable uncertainties and variability are:

1. Given California’s natural hydrologic variability, and the inherent uncertainty of our models, water and groundwater plans need to be prepared for simple long-term water balances to be substantially wrong. Plans must support adjustments and adaptations into the future. This is especially relevant for SGMA-required local Groundwater Sustainability Plans.
2. Water plans and operations also need to prepare for substantial variability in sources of water available across years. Source and sustainability planning should also account for uncertainty estimates and try to reduce them over time to improve the accuracy of their water budgets estimations.
3. To reduce and better understand water uncertainties and variability, and improve collaboration among local and state interests, more solid regional water accounting and measurement is needed.

Further Reading:

Arnold, Brad, Alvar Escriva-Bou, Jay Lund, and Ellen Hanak (2017). Accounting Water for the San Joaquin Valley. California WaterBlog.

Arnold, Brad, and Alvar Escriva-Bou (2017). Water Stress and a Changing San Joaquin Valley. Technical Appendix A: The San Joaquin Valley’s Water Balance. 13 pp. Public Policy Institute of California, San Francisco, CA.

Escriva-Bou, Alvar, Henry McCann, Ellen Hanak, Jay Lund, Brian Gray. 2016. Accounting for California’s Water. Public Policy Institute of California.

Hanak, Ellen, Jay Lund, Brad Arnold, Alvar Escriva-Bou, Brian Gray, Sarge Green, Thomas Harter, Richard Howitt, Duncan MacEwan, Josue Medellin-Azuara, Peter Moyle, and Nathaniel Seavy (2017). Water Stress and a Changing San Joaquin Valley. 48 pp. Public Policy Institute of California, San Francisco, CA.