| Abstract | Carbon absorption by vegetation, commonly referred as the net primary productivity (NPP), is one of important components in terrestrial carbon budget which affects climate change. Among different
approaches determining NPP, ecophysiologically-based models have advantages of (1) being scientifically sound and therefore potentially reliable, (2) the ability to handle interactions and feedbacks of different processes, (3) the flexibility to
describe details of biological processes under variety of conditions. However, application of models of this type has been hindered by the data availability and the need for temporal and spatial scalings, especially over large areas and for moderate
or high resolutions. At Canada Centre for Remote Sensing, efforts are made to develop an ecophysiologically-based model to simulate NPP over Canadian landmass at moderate resolutions (~1 km). An instantaneous leaf level photosynthesis model is
scaled up temporally and spatially to the entire canopy at daily step. Remote sensing techniques provide two of most important driving variables to the model: land cover type, and leaf area index (LAI). The information on land cover type is critical
in determining different functionlities of the various species, while LAI strongly affects almost all components of the model, including radiation absorption, photosynthesis, respiration, transpiration, rainfall interception and soil water balance.
In the model, dynamic change in vegetation detected by remote sensing is immediately considered. Timely remotely-sensed data are critical to accurate NPP calculations. Multiple-year satellite-derived data and ancillary climate and soil data are
compiled to simulate interannual carbon absorption by Canadian boreal forests (from 1994-96). The results are validated with biomass data and tower flux measurements in several locations in Canada. The interannual variation of carbon absorption is
analyzed for different vegetation type, weather and soil conditions. |