Distribution of snow accumulation on the Svartisen ice cap, Norway, assessed by a model of orographic precipitation

Abstract

Abstract:We apply a linear model of orographic precipitation (LT model) to estimate snow accumulation on the western Svartisen ice cap (220 km 2 ) in northern Norway. This model combines 3D airflow dynamics with simple parameterizations of cloud physics. The model is forced by large-scale atmospheric input variables taken from the ECMWF Re-analysis (ERA-40) of the European Center for Medium Range Weather Forecast (ECMWF), and the model parameters are kept constant for the entire simulation period, after optimization. The domain covers a 120 ð 125 km area surrounding the ice cap. The model is run using a 1-km resolution digital elevation model, and 6-h time steps over the period from 1958 to 2002. Precipitation data from surrounding meteorological stations and winter glacier mass balance measurements on several glaciers within the region are used to evaluate the model results. Precipitation obtained from the LT model agrees well with observations from precipitation gauges and there is also fair agreement between model results and specific winter mass balance observations on the ice cap if these are corrected for winter rain. The LT model reproduces well the spatial pattern of winter accumulation across the ice cap as well as the area-averaged winter mass balances of several other glaciers in the region. This indicates that it is a useful tool in providing high-resolution, deterministic estimates of precipitation in complex terrain as required for distributed hydrological and/or glacier mass balance modelling of unmeasured areas.

Modeling Winter Precipitation Over the Juneau Icefield, Alaska, Using a Linear Model of Orographic Precipitation

“…where q c (x, y) and q h (x, y) are the vertically integrated cloud water density and hydrometeor density respectively, and U = Ux + Vy is the advecting wind vector with northward and eastward components V and U. τ c and τ f generally range from 200 to 2,000 s Barstad and Smith, 2005;Crochet et al, 2007;Schuler et al, 2008) and their determination is discussed in section 3.2.2. S(x, y) is the vertically integrated (from z = 0 to z = ∞) condensation rate to cloud water as a response to uplift over terrain, referred to as the source term.…”

“…Early applications of the LT model required the optimization of the parameters N m , τ c , and τ f that were assumed constant in time and space Barstad and Smith, 2005;Crochet et al, 2007;Jóhannesson et al, 2007;Schuler et al, 2008). show typical values for these parameters.…”

“…The WRF precipitation data include some orographic enhancement from the WRF topography that must be removed prior to applying the LT model to the finer resolution DEM to avoid doublecounting of the orographic effect. We followed the procedure outlined by Schuler et al (2008) and apply the LT model to the coarse WRF topography using the WRF meteorological variables. The resulting orographic precipitation was subtracted from the WRF precipitation data to yield P ∞ .…”

“…Glacier mass balance models have often applied simple empirical, mostly elevation-dependent, relations to distribute point precipitation data across glacier surfaces (e.g., Hock and Holmgren, 2005;Radić et al, 2014) or directly used gridded coarse-scale climate data (e.g., Ziemen et al, 2016). Both approaches typically fail to resolve the complex precipitation patterns in glacierized mountainous terrain (Schuler et al, 2008;Jarosch et al, 2012).…”

“…While the LT model includes many assumptions and is not able to represent nonlinear processes, it is highly adaptable and has been explored and applied in numerous mountainous regions (e.g., Barstad and Smith, 2005;Anders et al, 2007Anders et al, , 2008Minder, 2010). However, it has only been used for direct glaciological applications in Iceland Jóhannesson et al, 2007), Norway (Schuler et al, 2008), Western Canada (Jarosch et al, 2012), and Patagonia (Weidemann et al, 2013).…”