Post‐Drought Groundwater Storage Recovery in California's Central Valley

Abstract

Groundwater overdraft during droughts is common in semiarid regions globally (Wada et al., 2010), and climate change is expected to further accelerate groundwater depletion in these regions (Alam, Gebremichael, Li, Dozier, & Lettenmaier, 2019;Wu et al., 2020). Groundwater overdrafts linked to droughts are caused both by reduced groundwater recharge and increased agricultural, industrial, and municipal water demand (Russo & Lall, 2017;Taylor et al., 2013). The effect of drought on groundwater can be especially severe in irrigated agricultural regions with limited surface water supply. There is a critical need to

Subsurface Water Flux in California's Central Valley and Its Source Watershed From Space Geodesy

“…Differences between our and Alam et al. 's (2021) estimate of Central Valley groundwater reflect groundwater fluctuations observed by GPS and GRACE but not accounted for by Alam et al. (2021), as well as uncertainties in the measurements and inversion.…”

“…During harsh drought from October 2011 to October 2015, the Central Valley lost 35 ± 4 km 3 of groundwater, reducing equivalent water thickness by 0.70 ± 0.08 m. If the mean rock porosity of the Central Valley were 0.2, then groundwater well levels would have fallen by a mean of 3.5 m. Heavy precipitation from water years 2016 through 2019 replenished 11 ± 4 km 3 of groundwater, 31% of the loss during the four harsh drought years.Our estimates of recharge of Central Valley groundwater fromGPS and GRACE are 58% (2006) and 31% (2012. The latter 31% recharge is less than the 60% recharge estimated from a water-balance model byAlam et al (2021).…”

“…(2021) water balance elements and can be attributed to water processes not in the water balance model: violet segments are positive differences inferred to be mountain‐block recharge (subsurface flow of groundwater from the mountains to the Valley); and teal green line segments are negative differences inferred to reflect underestimated evapotranspiration associated with groundwater pumping. When we calculate the difference, we remove soil moisture in the Alam et al., 2021 realization of Land Surface Model‐VIC (rather than in NLDAS‐Noah) to be consistent with the soil moisture model subtracted in that study. The 5.9 km 3 peak‐to‐peak seasonal oscillation in soil moisture in LSM‐VIC is smaller than the 7.6 km 3 seasonal oscillation in soil moisture in NLDAS‐Noah.…”

“… Hydrologic character of the Central Valley: (a) Seasonal oscillation in groundwater, (b) long‐term rate of groundwater loss, and (c) change in groundwater from October to October during periods of drought and heavy precipitation. Change in groundwater is estimated from Gravity Recovery and Climate Experiment (GRACE) and a hydrology model following the method of Famiglietti et al., 2011 (pink), from the water‐balance model of Alam et al., 2021 (deep blue), from GPS and GRACE in the joint inversion (gold), and from GPS (green). 95% confidence limits (error bars) are calculated using linear propagation of errors (blue) and Hector spectral analysis (red).…”

“…(2011) estimate change in groundwater in the Central Valley using GRACE and the Global Land Data Assimilation System model (Rodell et al., 2004). Several studies (Ahamed et al., 2022; Alam et al., 2021; Liu et al., 2019; Kim et al., 2020; Ojha et al., 2018, 2019) have followed Famiglietti et al. 's seminal formulation.…”