2500, boul. de l'Université
Université de Sherbrooke
819.821.8000 extension 62506
Climate change not only affects the entire globe but is also more pronounced in polar regions where the rate of warming is nearly twice of what is observed on average on the planet and is caused by the arctic amplification. This effect is primarily due to a significant loss of sea ice and snow cover that leads to a positive albedo feedback. The snow cover is therefore a key variable governing climate processes in polar regions, however, a lot of uncertainties emerge from spatial and temporal variability in snowpack properties which leads to more uncertainties in climate prediction. It is therefore crucial to improve and continue to monitor arctic/sub-arctic snow cover during this climate crisis. Satellite remote sensing is used to monitor at large scale proprieties of the snow cover and is the principal tool for monitoring the vast regions of Québec and Canada.
The global objective of my project is to analyze spatial variability of snow using remote sensing tools to reduce uncertainties linked to snow spatial heterogeneity for large scale SWE retrievals, but also to improve our understanding of the processes governing snow depth and snowpack properties distribution.
The specific objectives of my project are:
The main study site is located in Cambridge Bay, Nunavut, and will allow a proper characterization of snow state variables for an open tundra environment. More sites will potentially be necessary but are still unknown to this date.
First, to quantify snow spatial variability, snow depth maps derived from UAV imagery will be analyzed with orthomosaics of normalized difference vegetation index (NDVI) and an ecotype map to quantify the impact of topography and vegetation on total depth. Snow trenches combined with infrared imagery as well as measurements from a FMCW 24 GHz radar will be used to analyze the ratio of wind slab and depth hoar. Then, these relations will be used in the assimilation process to restrain the initial assumption on snowpack layers. Finally, an algorithm to detect ratio of depth hoar and wind slab will be developed with passive microwave using different frequencies (11, 19, 37, 89 GHz).
The expected results will be an improvement in our understanding of local processes by correlating variability in snow depth and ratios of the main layers with topography, land cover and vegetation. Secondly, these relationships will allow an improvement of arctic SWE estimations given their linkages with products that can be scaled up to satellite, which includes land cover type, NDVI product and DEM that will work with the assimilation process. Finally, the detection algorithm of the ratios of depth hoar will allow a temporal analysis of the evolution of depth hoar affecting the permafrost isolation in a climate change context.