Land-Cover Change and Climate Change Analysis of the Amur River Basin Using Remote Sensing Data

Abstract

Leaf area index (LAI) and vegetation type are two ecological variables that influence atmosphere-biosphere exchange of CO2 and that can be estimated from remote sensing techniques. A forest ecosystem process model was used to examine the importance of LAI and species-dependent physiology when estimating photosynthesis in 21 black spruce, white spruce, quaking aspen, paper birch, and balsam poplar forests near Fairbanks, Alaska. Model sensitivity analyses for these 21 stands showed that uncertainty in LAI and species composition caused errors in net canopy assimilation of as much as 42-70% and 14-36%, respectively, depending on forest type. The sensitivity of net canopy assimilation to species-dependent physiology was greater between needleleaf coniferous and broadleaf deciduous life forms than among species within life forms. A simple regression model that recognized stand differences in LAI and life-form type (needleleaf coniferous, broadleaf deciduous) accounted for 94% of the variation in simulated net canopy assimilation for the 21 stands. Expanding the model to include species composition rather than life-form type only accounted for an additional 1% of the variation in simulated net canopy assimilation. The statistical model was applied to a synthetic aperture radar scene that discriminatted black spruce, white spruce, balsam poplar, and alder forests in a 76.8 km2 region near Fairbanks. This analysis showed that regional net canopy assimilation was as sensitive to the area of each forest type as it was to the LAI of each forest type. The analyses reported in this article highlight the importance of recognizing physiological differences among needleleaf coniferous and broadleaf deciduous forest types when estimating regional net canopy assimilation in boreal forests. © 1993.