Climate change and storage response in alpine geologic endmember catchments using integrated modeling and baseflow recession analysis

Abstract:

Runoff generation in snowmelt-dominated alpine systems predominantly occurs in subsurface, be it in the soil, saprolite, or fractured bedrock zone, and shifts in timing and amount of runoff due to climate change remains an open topic of research. Furthermore, the degree to which subsurface storage offsets the loss of snow storage in porous and fractured alpine terrains, i.e., the hydrogeologic buffering capacity, is still largely unknown. The snowmelt-dominated alpine watersheds in California and Chile are particularly vulnerable to climate change due to their Mediterranean climate, where winter snowpack sustains the demand of urban and agricultural needs during the dry summers. The streams draining the western slope of the Sierra Nevada and Andes mountains show a decline in snowmelt runoff, with an earlier shift in spring pulse and center of mass timing over the past 50 years. Following the snowmelt period, summer low flows are sustained by groundwater, and interbasin baseflow trends have been shown to correlate with geology, and to some extent, soil thickness in less permeable basins. However, the interannual (intrabasin) baseflow trends have not been explored with respect to climate change impacts to storage-discharge relationships. Here we estimate long-term groundwater storage trends via baseflow recession analysis for two geologically distinct alpine basins: the granitic Middle Fork Kaweah in the southern Sierra Nevada, California (640 masl, 264.2 km² with daily data back to 1949) and the volcanic Diguillín in the central Andes, Bío Bío Region, Chile (670 masl and 334 km² with daily data back to 1959). We employ a simple linear reservoir model for estimating storage from baseflow, and investigate the sensitivity to watershed characteristics, such as depth of groundwater circulation and storage on the results. We supplement these results with numerical experiments conducted using ParFlow-CLM, a fully-integrated hydrologic model coupled to a land surface model for alpine endmember hillslopes. By combining the data-driven analysis and process-driven modeling, we can not only glean trends from the observational record, but test hypotheses such as a snow to rain transition and increased evapotranspiration impacting recharge and thus runoff generation in these distinct hydrogeologic settings.