Scale effects in a distributed snow water equivalence and snowmelt model for mountain basins

Distributed energy-balance modeling of snow-cover evolution and melt in rugged terrain: Tobacco Root Mountains, Montana, USA

“…These calculated values are consistent with those reported in a related paper (Olyphant, 1986a). Cline et al (1998b) noted the importance of correctly identifying the initial snow distribution before attempting to make calculations of snowmelt, and their work has been refined and reinforced by others (e.g., Balk and Elder, 2000;Mittaz et al, 2002;Marks et al, 2002;Winstral and Marks, 2002;Anderton et al, 2003). In this study, the snow-transport model SnowTran-3D of Liston and Sturm (1998) was used to generate a spatial distribution of snow-water equivalent for the beginning of the snowmelt season.…”

How do tradeoffs in satellite spatial and temporal resolution impact snow water equivalent reconstruction?

“…Winstral et al (2014) examined scale in a snow energy balance model at a range of spatial resolutions and found that 100 m spatial resolution is needed to accurately simulate snowmelt. Contrary to Cline et al (1998), Schlögl et al (2016) report that SWE increases with DEM resolution in two alpine basins. Similarly, Baba et al (2019) used an energy balance model with a DEM at 8-1000 m and report good agreement with fine-resolution snow cover maps up to 250 m but a loss in agreement at coarser resolution likely due to excessive smoothing of topographic effects.…”

“…For example, in the UTM 11N projection, the shapefile obtained from ASO Inc. for the Tuolumne River basin above Hetch Hetchy Reservoir has an area of 1175 km 2 ; a raster of the basin at 120 m has an area of 1153 km 2 (−1.8 %), while a raster at 2000 m has an area of 1132 km 2 (−3.6 %). Even when using vector basin outlines, as suggested by Cline et al (1998), these artifacts are inherent in the discretization of geospatial data and cannot be eliminated.…”

“…In Durand et al (2008) and Molotch and Margulis (2008), the finer-resolution Landsat ETM+ snow cover used a spectral unmixing technique (Painter et al, 2003), but the MODIS snow cover was based on the normalized snow difference technique, which only uses two bands versus all available for spectral unmixing and is shown to have higher MAE and bias (Stillinger et al, 2023). In Cline et al (1998), the only other study to specifically examine spatial scale with SWE reconstruction, a spectral mixture technique was used on 30 m Landsat ETM+ to produce snow cover estimates (Rosenthal and Dozier, 1996). In that study, the coarsened results produced basin-wide SWE above and below the control simulation used as validation, suggesting that coarsening components of the energy balance did not show a clear trend in error.…”

Trends in spring snowpack over a half-century of climate warming in California, USA

“…The former presents the steepest gradient of the western side of the Sierras, and the latter lies directly within the winter Pacific storm track. The future efficacy of the ‘snow gun’ in offsetting increased melt is unknown, largely due to the coarse spatial resolution of regional climate models which are unable to simulate this finer-scale orographic effect (Cline and others, 1998; Kim, 2001). The importance of the snow gun effect in the overall SWE budget in California may have already found expression in the relatively small change in seasonal run-off despite snowpack loss at low elevations (Dettinger and Cayan, 1995).…”