Ziyu Wang1, Robert Grant1, Tony Trofymow2, Pierre Bernier3, Werner Kurz2, Graham Stinson2, Luc Guindon3, J. Chen4, Altaf Arain5, Changhui Peng6, Ajit Govind4, Jagadeesh Yeluripati5, and Jianfeng Sun6. (1) University of Alberta, Edmonton, AB T6G 2E3, Canada, (2) Canadian Forest Service, Pacific Forestry Centre, Victoria, BC V8Z 1M5, Canada, (3) Canadian Forest Service, Laurentian Forestry Centre, Quebec City, QC G1V 4C7, Canada, (4) Department of Geography and Program in Planning, University of Toronto, Toronto, ON M5S 3G3, Canada, (5) School of Geography and Geology, McMaster University, Hamilton, ON L8S 4K1, Canada, (6) Institute of Environment Sciences, University of Quebec at Montreal, Montreal, QC H3C 3P8, Canada
The objective of this study is to test various landscape-level process-based modeling approaches for the purpose of tracking the dynamics of ecosystem carbon. The study used multi-decadal datasets of forest properties and disturbance history assembled for two experimental areas located respectively on Vancouver Island (Oyster River) and in the Eastern boreal forest (Chibougamau). Three models, CBM-CFS3 (Kurz, 2006), Ecosys (Grant, 2001) and C-Class (Arain, 2002), were used for Oyster River area and four models, the above three models and INTEC (Chen, 2000), were used for Chibougamau site. The attributes of the combined layers of soil, forest type and disturbance provided input data for all models. CBM-CFS3 model runs were done on a forest polygon basis. Ecosys, C-CLASS and INTEC runs were done on the basis of grid cells. For these models, the GIS vector data layers of soil, vegetation, harvesting, fire and fertilization were converted to 100m × 100m grids. The attributes for the total 2500 grid cells of Oyster River area and 6275 grid cells of Chibougamau site were extracted from their centroids and grouped into different model runs. Above-ground biomass was used as a common comparative variable for all models, and was modeled from 1920 to 2005 for Oyster River site and from 1928 to 2005 for Chibougamau site. The model results showed that the impact of disturbances such as harvesting and fire dominated the dynamics of above ground tree biomass both in Oyster River and Chibougamau sites. For Oyster River site, the above ground biomass of grid cells centered around 10000-15000 gCm-2. For Chibougamau site, it was 2500-5000gCm-2. Comparison of models against measured or empirically estimated changes showed a significant range in the ability of the models to track above ground biomass dynamics. Analysis of the causes of such discrepancies will be used for model improvement.