Causes of Missing Snowmelt Following Drought

Abstract

Water management in snowy mountainous regions hinges on forecasting snowmelt runoff. However, droughts are altering snowpack‐runoff relationships with ongoing debate about the driving mechanisms. For example, in 2021 in California, less than half of predicted streamflow arrived. Mechanisms proposed for this “missing” streamflow included changes in evapotranspiration (ET), rainfall, and subsurface moisture conditions. Here, we demonstrate that ET in drought years generates dry subsurface conditions that reduce runoff in subsequent years. A model including this legacy of depleted moisture storage reduced median error in 2021 forecasts from 60% to 20% at 15 minimally disturbed basins and from 18% to 2% at 6 water supply basins in the Sierra Nevada (basins range in area from 5 to 23,051 km2 and mean annual precipitation from 814 to 1,549 mm). Our findings indicate that the relationship between snowpack and runoff will evolve as plant ecosystems respond to climate change and alter subsurface water storage dynamics.

Relationships between snowpack, low flows and stream temperature in mountain watersheds of the US west coast

“…The summer of 2021 provided a good example of multiple dry years in a row causing exceptionally low flows due to an accumulated deficit in subsurface water storage (Lapides et al, 2022). Water Year (WY) 2021 (denoted with an X in Figure 5) had the lowest Qmin in our study period for Sagehen and Merced.…”

“…The summer of 2021 provided a good example of multiple dry years in a row causing exceptionally low flows due to an accumulated deficit in subsurface water storage (Lapides et al, 2022). Water Year Hatchett, 2023).…”

“…The storage-only model (Figure4A) predicted 2021 Qmin almost perfectly in both of those watersheds (overpredicted by <0.01 mm/day). High model accuracy for 2021 suggests that the record low flows were due to exceptional multi-year drought conditions (and the resulting storage deficits, as suggested byLapides et al, [2022]) rather than by a new emerging behaviour. Peak SWE and total precipitation for WY 2021 were low in these watersheds (bottom 25th and 10th percentiles for SWE and WY P, respectively), but not the lowest on record.…”

Geologic Controls on Apparent Root‐Zone Storage Capacity

“…Nevertheless, storage capacity calculated via deficit‐style approaches has many theoretical and pragmatic advantages. S R results in improved hydrological model performance when used as an input parameter (Lapides et al, 2023; Wang‐Erlandsson et al., 2016) and can explain continental‐scale patterns in water partitioning (Cheng et al., 2022) and storage dynamics (Trautmann et al., 2022); deficit calculations have also proven essential in the accurate prediction of snowmelt contributions to streamflow following droughts (Lapides et al., 2022). Importantly, deficit‐calculated S R does not require a priori assumptions regarding porosity or rooting profiles, and distributed hydrologic flux datasets make it feasible to estimate S R at large spatial scales in cloud‐based analysis platforms like Google Earth Engine.…”

The Increasing Role of Seasonal Rainfall in Western U.S. Summer Streamflow

“…However, rainfall in winter and the following warm season (April-September) can make nontrivial contributions to summer streamflow by replenishing soil moisture, a mechanism that generally has received less attention than snowmelt. For example, Lapides et al (2022) found that warm season runoff generation is strongly tied to root-zone water storage deficits, which must be replenished by snowmelt or rainfall before substantial runoff occur. Moreover, Cooper et al (2018) found a larger percentage change of summer low flows per percentage change of winter precipitation for rain-dominated catchments as compared with snow-dominated catchments in maritime WUS mountains.…”