Evaluating the performance of the Canadian Land Surface Scheme (CLASS) in three different winter landscapes.

摘要

This thesis tests the Canadian Land Surface Scheme (CLASS) [Version 2.6], a land surface scheme (LSS) designed for use in the Canadian Climate Centre's Global Climate Model (MSC GCM). Specifically this thesis tests CLASS's ability to model snow and winter-time fluxes of energy and moisture in three distinct northern hemisphere landscapes: a bog located near Ottawa, Ontario, a subarctic open woodland near Churchill, Manitoba and a hardwood forest in the Turkey Lakes Watershed (TLW) near Sault Ste Marie, Ontario.; The datasets were optimized and then used to test the effectiveness of proposed snow depth and density modifications. The optimization process involved sensitivity tests to determine which of CLASS's four vegetation types most suited the vegetation at each site and choosing the best longwave downward (L↓) formulation for driving the Churchill and TLW sites. Then CLASS's minimum complete snow depth, which is the depth at which CLASS's snowcover completely covers the grid cell, was analyzed using a series of sensitivity tests with various minimum depth values. This was followed by a comparison of three new density algorithms to CLASS's original algorithm [Version 2.6]. (Abstract shortened by UMI.)