In the western United States, the seasonal phase of snow storage bridges between winter‐dominant precipitation and summer‐dominant water demand. The critical role of snow in water supply has been frequently quantified using the ratio of snowmelt‐derived runoff to total runoff. However, current estimates of the fraction of annual runoff generated by snowmelt are not based on systematic analyses. Here based on hydrological model simulations and a new snowmelt tracking algorithm, we show that 53% of the total runoff in the western United States originates as snowmelt, despite only 37% of the precipitation falling as snow. In mountainous areas, snowmelt is responsible for 70% of the total runoff. By 2100, the contribution of snowmelt to runoff will decrease by one third for the western U.S. in the Intergovernmental Panel on Climate Change Representative Concentration Pathway 8.5 scenario. Snowmelt‐derived runoff currently makes up two thirds of the inflow to the region's major reservoirs. We argue that substantial impacts on water supply are likely in a warmer climate.

Despite the importance of snow in global water and energy budgets, estimates of global mountain snow water equivalent (SWE) are not well constrained. Two approaches for estimating total range-wide SWE over Sierra Nevada, California, are assessed: 1) global/hemispherical models and remote sensing and models available for continental United States (CONUS) plus southern Canada (CONUS+) available to the scientific community and 2) regional climate model simulations via the Weather Research and Forecasting (WRF) Model run at 3, 9, and 27 km. As no truth dataset provides total mountain range SWE, these two approaches are compared to a “reference” SWE consisting of three published, independent datasets that utilize/validate against in situ SWE measurements. Model outputs are compared with the reference datasets for three water years: 2005 (high snow accumulation), 2009 (average), and 2014 (low). There is a distinctive difference between the reference/WRF datasets and the global/CONUS+ daily estimates of SWE, with the former suggesting up to an order of magnitude more snow. Results are qualitatively similar for peak SWE and 1 April SWE for all three years. Analysis of SWE time series indicates that lower SWE for global and CONUS+ datasets is likely due to precipitation, rain/snow partitioning, and ablation parameterization differences. It is found that WRF produces reasonable (within 50%) estimates of total mountain range SWE in the Sierra Nevada, while the global and CONUS+ datasets underestimate SWE.

Despite the importance of mountain snowpack to understanding the water and energy cycles in North America's montane regions, no reliable mountain snow climatology exists for the entire continent. We present a new estimate of mountain snow water equivalent (SWE) for North America from regional climate model simulations. Climatological peak SWE in North America mountains is 1,006 km3, 2.94 times larger than previous estimates from reanalyses. By combining this mountain SWE value with the best available global product in nonmountain areas, we estimate peak North America SWE of 1,684 km3, 55% greater than previous estimates. In our simulations, the date of maximum SWE varies widely by mountain range, from early March to mid‐April. Though mountains comprise 24% of the continent's land area, we estimate that they contain ~60% of North American SWE. This new estimate is a suitable benchmark for continental‐ and global‐scale water and energy budget studies.

Seasonal climate forecasts are regularly published to provide decision makers with insights on upcoming climate conditions. Precipitation forecasts, in particular, are useful for fields such as agriculture and water resources. Projections frequently cite a single climate oscillation such as El Niño-Southern Oscillation (ENSO) or the North Atlantic Oscillation (NAO) when suggesting whether a region will be wetter or drier than normal. The complex climate system is composed of a multitude of simultaneous oceanic and atmospheric oscillations, however. Through the study of five atmospheric-pressure-based oscillations, their interactions, and associated precipitation values, this research demonstrates the wide variety of precipitation patterns that can arise when different phases of prominent climate modes occur. Results show that incorporating other Northern Hemisphere teleconnections can dampen or shift expected ENSO and NAO impact patterns. These results indicate that seasonal precipitation projections may be improved by incorporating multiple, regionally important teleconnection indices into the forecast.

Mountain snow has a fundamental role in regional water budgets through its seasonal accumulation, storage, and melt. However, characterizing snow accumulation over large regions remains difficult because of limited observational networks and the inability of available satellite instruments to remotely sense snow depth or water equivalent in mountains. Models offer some ability to estimate snow water storage (SWS) on mountain range to continental scales. Here we compare four commonly used global data sets to understand whether there is a consensus regarding mountain SWS estimates among them. The data sets—European Centre for Medium‐Range Weather Forecasts Reanalysis‐Interim, Global Land Data Assimilation System, Modern‐Era Retrospective Analysis for Research and Applications version 2, and Variable Infiltration Capacity—agree to within ±36% of the four–data set average for total global SWS. When mountain areas are extracted using a new seasonal mountain snow classification data set, the four data products have more agreement, where all are within ±21% of the seasonal SWS for mountain regions. However, when compared to high‐resolution (9 km) simulations of SWS from the Weather Research and Forecasting (WRF) regional model, the four global products differ from WRF‐estimated North American mountain snow accumulation by 40–66%, with a negative bias up to 651 km3, comparable to the annual streamflow of the Mississippi River. If we extend the North America SWS bias to global mountains, the global data sets may miss as much as 1,500 km3 of SWS, equivalent to 4% of the flow in all the world's rivers. The potential difference of SWS suggests more work must be done to characterize water resources in snow‐dominated regions, particularly in mountains.

Abstract. The Snow Ensemble Uncertainty Project (SEUP) is an effort to establish a baseline characterization of snow water equivalent (SWE) uncertainty across North America with the goal of informing global snow observational needs. An ensemble-based modeling approach, encompassing a suite of current operational models is used to assess the uncertainty in SWE and total snow storage (SWS) estimation over North America during the 2009–2017 period. The highest modeled SWE uncertainty is observed in mountainous regions, likely due to the relatively deep snow, forcing uncertainties, and variability between the different models in resolving the snow processes over complex terrain. This highlights a need for high-resolution observations in mountains to capture the high spatial SWE variability. The greatest SWS is found in Tundra regions where, even though the spatiotemporal variability in modeled SWE is low, there is considerable uncertainty in the SWS estimates due to the large areal extent over which those estimates are spread. This highlights the need for high accuracy in snow estimations across the Tundra. In midlatitude boreal forests, large uncertainties in both SWE and SWS indicate that vegetation–snow impacts are a critical area where focused improvements to modeled snow estimation efforts need to be made. Finally, the SEUP results indicate that SWE uncertainty is driving runoff uncertainty, and measurements may be beneficial in reducing uncertainty in SWE and runoff, during the melt season at high latitudes (e.g., Tundra and Taiga regions) and in the western mountain regions, whereas observations at (or near) peak SWE accumulation are more helpful over the midlatitudes.

Mountain snow cover plays an important role in regional climate due to its high albedo, its effects on atmospheric convection, and its influence on runoff. Snowpack water storage is also a critical water resource and understanding how it varies is of great social value. Models are often employed to reconstruct snowpack and explore and understand snow cover variability. Here, we use a new, accurate satellite-derived snow product to evaluate the ability of the Weather Research and Forecasting (WRF) regional climate model, combined with the Noah land surface model with multi-parameterization options (Noah-MP), to simulate snow cover fraction (SCF) and snow water equivalent (SWE) in a 3-km domain over the central Sierra Nevada. WRF/Noah-MP SWE simulations improve on previous versions of the Noah land surface model by removing an early bias in snow melt, though a 2-day positive melt bias in SWE timing remains significant at the 90% confidence level. In addition, WRF/Noah-MP identifies the areas where snow is present to within 94.3% and captures large-scale variability in SCF. Temporal root mean squared error (RMSE) of the domain-average SCF was 1938.6 km 2 (24%). However, our study shows that WRF/Noah-MP struggles to simulate SCF at finer spatial scales. The parameterization for SCF fails to produce temporal variations in grid-scale SCF, and depletion occurs too rapidly. As a result, the WRF/Noah-MP SCF parameterization reduces to a binary function in mountain environments. Sensitivity tests show that adjustment of the parameterization may improve simulation of SCF during accumulation or melt but does not remove the bias for the entire snow season. Although WRF/Noah-MP accurately simulates the presence or absence of snow, high-resolution, reliable SCF estimates may only be attainable if snow depletion parameterizations are designed specifically for complex topographical areas.

Shugong (2021) Snow Ensemble Uncertainty Project (SEUP): quantification of snow water equivalent uncertainty across North America via ensemble land surface modeling. The Cryosphere, 15 (2). pp. 771-791.