Modeling BRF and radiation regime of boreal and tropical forest: II. PAR regime

Monitoring functioning of forest ecosystems with remote sensing depends on canopy BRF (bidirectional reflectance function) sensitivity to biophysical parameters and PAR (photosynthetically active radiation) regime. Here, we s tudied the 3-D PAR regime of tropical (Sumatra) and boreal (Canada) forests , with the DART (dis crete anisotropic radiative transfer) model. We consid ered rc;ide ranges of Sun off-nadir angles (theta(s)), leaf area index (LAI ), and leaf clumping. The BRF of these forests is analyzed in a companion a rticle. Here, we also investigated the possibility to derive simple analyti cal expressions of PAR vertical profiles: We fitted DART simulated PAR (abs orbed PAR) profiles with a modified Goudriaan late; (1977) the extinction c oefficient of which is multiplied by a factor a that accounts for canopy ar chitecture. Similarly to BRF, the PAR regime is very influenced by canopy s tructure: for theta(s)=50 degrees, a approximate to 0.40 for tropical fores t, a approximate to 0.56 for coniferous boreal forest (OBS), and a approxim ate to 0.86 for deciduous boreal forest (OA). Moreover, a strongly depends on theta(s) and LAI for example, for tropical forest a decreases from 0.44 to 0.12 if theta(s) varies from 0 degrees to 80 degrees, and from 0.70 to 0 .38 if LAI increases from 3 to 10. a decreases slightly with the increase o f leaf clumpiness. The NDVI of tropical and boreal forest sites was rather related to the LAI and fAPAR of the upper canopy than to those of total can opy. Finally, we studied the impact of forest architecture on canopy photos ynthesis with the coupling of DART with a leaf functioning model. Neglect o f architecture can lend to errors as large as 25% in relative on forest CO2 assimilation. (C) Elsevier Science Inc., 1999.