Alaska Science Center
End-of-winter snow depth variability on glaciers in Alaska
Start a New Search | Return to existing search by pressing your browser's back arrow
Full Publication: http://dx.doi.org/10.1002/2015JF003539
Product Type: Journal Article
Ground Penetrating Radar on North American Glaciers
A quantitative understanding of snow thickness and snow water equivalent (SWE) on glaciers is essential to a wide range of scientific and resource management topics. However, robust SWE estimates are observationally challenging, in part because SWE can vary abruptly over short distances in complex terrain due to interactions between topography and meteorological processes. In spring 2013, we measured snow accumulation on several glaciers around the Gulf of Alaska using both ground- and helicopter-based ground penetrating radar surveys, complemented by extensive ground-truth observations. We found that SWE can be highly variable (40% difference) over short spatial scales (10s to 100s of meters), especially in the ablation zone where the underlying ice surfaces are typically rough. Elevation provides the dominant basin-scale influence on SWE, with gradients ranging from 115–400 mm /100 m. Regionally, total accumulation and the accumulation gradient are strongly controlled by a glacier's distance from the coastal moisture source. Multiple linear regressions, used to calculate distributed SWE fields, show that robust results require adequate sampling of the true distribution of multiple terrain parameters. Final SWE estimates (comparable to winter balances) show reasonable agreement with both the PRISM climate dataset (9–36% difference) and the USGS Alaska Benchmark Glaciers (6–36% difference). All the glaciers in our study exhibit substantial sensitivity to changing snow-rain fractions, regardless of their location in a coastal or continental climate. While process-based SWE projections remain elusive, the collection of GPR derived datasets provides a greatly enhanced perspective on the spatial distribution of SWE and will pave the way for future work that may eventually allow such projections.