Regional Climate Modeling over the Glaciated Regions of the Canadian High Arcti

The Canadian Arctic Islands (CAI) contain the largest concentration of terrestrial ice outside of the continental ice sheets. Mass loss from this region has recently increased sharply due to above average summer temperatures. Thus, increasing the understanding of the mechanisms responsible for mass loss from this region is critical. Previously, Regional Climate Models (RCMs) have been utilized to estimate climatic balance over Greenland and Antarctica. This method offers the opportunity to study a full suite of climatic variables over extensive spatially distributed grids. However, there are doubts of the applicability of such models to the CAI, given the relatively complex topography of the CAI. To test RCMs in the CAI, the polar version of the regional climate model MM5 was run at high resolution over Devon Ice Cap. At low altitudes, residuals (computed through comparisons with in situ measurements) in the net radiation budget were driven primarily by residuals in net shortwave (NSW) radiation. Residuals in NSW are largely due to inaccuracies in modeled cloud cover and modeled albedo. Albedo on glaciers and ice sheets is oversimplified in Polar MM5 and its successor, the Polar version of the Weather Research and Forecast model (Polar WRF), and is an obvious place for model improvement. Subsequently, an inline parameterization of albedo for Polar WRF was developed as a function of the depth, temperature and age of snow. The parameterization was able to reproduce elevation gradients of seasonal mean albedo derived from satellite albedo measurements (MODIS MOD10A1 daily albedo), on the western slope of the Greenland Ice Sheet for three years. Feedbacks between modelled albedo and modelled surface energy budget components were identified. The shortwave radiation flux feeds back positively with changes to albedo, whereas the longwave, turbulent and ground energy fluxes all feed back negatively, with a maximum combined magnitude of two thirds of the shortwave feedback magnitude. These strong feedbacks demonstrate that an accurate albedo parameterization must be run inline within an RCM, to accurately quantify the net surface energy budget of an ice sheet. Finally, Polar WRF, with the improved albedo parameterization, was used to simulate climatic balance over the Queen Elizabeth Islands for the summers of 2001 to 2008. Climatic balance was derived from the output using energy balance and temperature index melt models. Regional mass balance was calculated by combining climatic balance with estimates of iceberg discharge. Mass balance estimates from the model agreed, within the bounds of uncertainty, with estimates from previous studies, thus supporting the assertion that mass loss from the QEI accelerated during the first decade of the 21st century. Melt rates on the seven major icecaps of the QEI became more correlated to one another during the period 2001-2008. However, precipitation became less correlated from 2003-2008. These observations are coincident with dramatic increases in melt on all of the ice caps, and it is speculated that both are caused by decreases in the scale of disturbances delivering precipitation to the region over time.