Glacier retreat is ubiquitous in the world’s mountain and polar-regions, reshaping the landscape, impacting regional hydrology (e.g., Huss, 2011), and contributing to global sea level rise (Radic and Hock, 2011). Areal changes in glacier extent can be measured by satellite and are well documented (e.g., Paul et al., 2004; Bolch et al., 2010), but mass balance and volume changes need to be measured in the field and modelled through glaciological and regional climate models (RCMs). Despite widespread interest and concern about glacier retreat, the intrinsic scale of glacier processes and the meteorological complexity of mountain regions have limited attempts to include a dynamic representation of glaciers in RCMs. Innovation and model development are needed to improve the representation of glaciers/icefields, glacier-hydrological processes, and atmospheric feedbacks associated with changing glacier cover and surface characteristics within climate models. Kotlarski et al. (2010) present important advances on this front, and we propose to adapt and build on this within CRCM5. Our aim is to dynamically represent the ensemble of glaciers in western and Arctic Canada. This will allow examination of glacier sensitivity to climate variability and change, improving understanding of glacier contributions to streamflow in glacierized catchments (cf. project C.2) and broader climate system feedbacks.
Subgrid representation of mountain glaciers can be introduced through a glacier land surface class in the land-surface model, with adaptive area and discrete elevation ‘bins’ within the RCM grid cell. Elevation bins are essential to allow high-elevation accumulation areas and low-elevation valleys to be represented in RCM, as these dictate the location and extent of glaciers on the landscape (Marshall and Clarke, 1999). The subgrid snow model parameterization developed in project C.2 will be applied for this purpose, for simulation of snow accumulation on the glaciers. Snow and ice ablation models will also be adapted from the improved snow models developed in project C.2. This will involve an extension for glacier-specific energy balance processes such as albedo models for firn and ice, turbulent exchanges in the stable glacier boundary layer, and impacts of katabatic winds on temperature lapse rates and glacier melt (e.g., Shea and Moore, 2010). Numerical schemes to represent these processes need to be fully coupled with CRCM5 and Canadian LAnd Surface Scheme (CLASS; Verseghy, 2008) to enforce conservation with respect to exchanges of energy and moisture between glaciers, the atmosphere and the land surface.
In addition to models of glacier mass balance, a treatment of glacier dynamics needs to be added to CRCM5/CLASS to allow simulation of evolving glacier area in transient runs, enabling a representation of this important feedback in future climate change and hydrological impact analyses. Simplified subgrid representations of flow dynamics are possible (e.g., Marshall et al., 2011), and are preferable to static glacier cover or volume-area scaling techniques to estimate future glacier extent (e.g., Paul and Kotlarski, 2010). Parameterizations of ice flow, along with other glacier-specific hydrological processes, need to be developed in the context of the subgrid land surface model developments within Theme C, such as project C2.