Skip to main content
U.S. flag

An official website of the United States government

Publication Date
11 September 2018

Assessing Mountains as Natural Reservoirs with a Multi‐Metric Framework

Subtitle
Improving the Simulation of Snowpack: Elucidating Error in Modeled Snow Processes.
Print / PDF
Powerpoint Slide
Science

Developed a multi-metric framework, the snow water equivalent (SWE) triangle, that helps to describe snowpack characteristics associated with total water volume build‐up, peak water availability, and the rate of water release. This approach can be applied across a wide-range of available snow products and highlights compensating errors that would not be reflected in conventional large‐scale spatiotemporal analysis.

Impact

All snow products analyzed show variation across the various SWE triangle metrics, even within observationally constrained snow products. This spread was especially shown in spring season melt rates. Due to the importance of snowmelt in reservoir management, addressing melt rate biases in climate models will be critical to maximize climate simulation utility for water stakeholders.

Summary

Mountain snowpack is a key natural water reservoir. Due to anthropogenic climate change, this natural reservoir will likely decline over the next century. This decline will be nonlinear in both space and time. As such, climate model estimates will be key in constraining uncertainty surrounding this decline. These virtual laboratories allow us to understand what could be and not just what has been. The utility of climate models is apparent, however, uncertainties surrounding climate model estimates of mountain snowpack need to be understood. This article assesses uncertainties surrounding both observationally based and climate model estimates of mountain snowpack in the California Sierra Nevada. We use a new multi‐metric evaluation framework called the snow water equivalent (SWE) triangle. This framework elucidates agreement/disagreement in estimates of peak water volume and timing, accumulation and melt rates, and the lengths of the accumulation and melt seasons. We found that spread across snowpack datasets were partly driven by differences in the accumulation to peak timing phase of the winter season but was dominated by melt season differences. To expand climate model utility for water management applications more research is needed to understand why models tended to have earlier peak timing and abrupt snowmelt.

Point of Contact
Andrew D. Jones
Institution(s)
Lawrence Berkeley National Laboratory (LBNL)
Funding Program Area(s)
Publication