Divergent Sensitivity of Soil Water Stress To Changing Snowmelt Regimes in the Western U.S.

Abstract:

Altered snowmelt regimes from regional warming threaten mountain ecosystems with greater water stress and increased the likelihood of disturbance. The sensitivity of vegetation to changing snowpack regimes is strongly mediated by soil water storage, yet a comprehensive framework to identify areas sensitive to changing snowpack regimes is lacking. In this study we ask two questions: 1) What climatic predictors explain inter-annual variability in the duration of soil water stress (DWS) and length of non-water stress season (NWSS)? and 2) What site characteristics increase the sensitivity of DWS and NWSS to changes in snowmelt dynamics? We compiled soil moisture at 10, 20 and 50 cm depths from 62 SNOTEL stations with >5 years of records. Soil water stress occurred when soil moisture was below the measured wilting point and NWSS was the number of days without water stress after snowmelt began. The day of snow disappearance (DSD) consistently explained the greatest variability in DWS across all site-years and at individual sites. On average, a one day earlier snow disappearance lead to 0.7 days greater DWS, but individual sites ranged from 0.2 to 1.8 days (36 of 62 sites had significant relationships between DSD and DWS). Despite earlier DSD leading to greater DWS at all sites, earlier DSD led to both significant increases (5 of 62) and decreases (7 of 62) in the length of the NWSS. Satellite-derived vegetation greenness confirmed that earlier DSD caused both lower and higher peak annual greenness depending on the site. A simple soil moisture model indicated that areas with finer soil texture, greater potential evapotranspiration, and longer NWSS were most sensitive to reduced NWSS from changing snowpack dynamics. These findings suggest a divergent response across snow-covered forests to earlier snowmelt timing independent of changing precipitation patterns: 1) historically water-stressed sites are most at risk for reduced vegetation productivity and 2) sites with low historical water stress may see enhanced vegetation productivity. These results have important implications for future alterations to the location and magnitude of vegetation-mediated hydrologic fluxes from altered snowpack regimes.