Modelling water, carbon, and nitrogen dynamics in CLASS: Canadian Land Surface Scheme.

摘要

Land surface schemes in atmospheric General Circulation Models (GCMs) significantly affect the predicted surface climate. Over the past decade, several “second-generation” land surface schemes have emerged and dominated the modelling studies of climate by GCMs; CLASS, the Canadian Land Surface Scheme which was developed at the Canadian Climate Center for the Canadian GCM, is one of them. While it greatly improved the evaluations of land surface processes over its earlier version of the “first-generation” land surface scheme, it was realized recently that improperly prescribed vegetation parameters were the largest source of error in climate modelling.; These limitations were addressed in this thesis research by developing three modules in the current version of CLASS V2.6: SVATC—a carbon-coupled water transfer module in the soil-vegetation-atmosphere system; PLANTC&barbelow;—a dynamic plant module designed to simulate plant carbon and nitrogen processes including photosynthesis, respiration, growth and litterfall, etc.; and SOILC—a soil carbon and nitrogen module designed to simulate organic matter transformation processes in and on soil. This new version of CLASS physiologically couples plant water and carbon dynamics, implements plant litter and soil carbon biogeochemical cycles, emphasizes the role of nitrogen in land surface processes, and feeds back dynamically based vegetation parameters to the GCM. The CLASS has been improved by including carbon dioxide (CO₂) flux between land surfaces and the atmosphere, thus making the predictions of climate change more realistic.; Simulations were implemented on deciduous trees. Data from the Old Aspen (Populus tremuloides) site in the Southern Study Area (SSA-OA) of the Boreal Ecosystem-Atmosphere Study (BOREAS) were used to initialize and drive the model. Comparisons show that annual root mean square error and correlation coefficient between model output and measurements for daily evapotranspiration were 0.71 mm H₂O d−1 and 0.87, and for carbon exchange were 1.10 g C m−2 d−1 and 0.93. The model predicted this aspen ecosystem was a net carbon sink of 163.6 g C m−2 y−1 and 203.2 g C m −2 y−1 for 1994 and 1996, respectively. It accounted for about 16.7% of the total gross primary production (GPP) on average for the two years.