RADIATION ABSORPTION AND USE BY HUMID SAVANNA GRASSLAND - ASSESSMENT USING REMOTE-SENSING AND MODELING

Citation
X. Leroux et al., RADIATION ABSORPTION AND USE BY HUMID SAVANNA GRASSLAND - ASSESSMENT USING REMOTE-SENSING AND MODELING, Agricultural and forest meteorology, 85(1-2), 1997, pp. 117-132
Citations number
58
Categorie Soggetti
Metereology & Atmospheric Sciences",Agriculture,Forestry
ISSN journal
01681923
Volume
85
Issue
1-2
Year of publication
1997
Pages
117 - 132
Database
ISI
SICI code
0168-1923(1997)85:1-2<117:RAAUBH>2.0.ZU;2-R
Abstract
The components of the canopy radiation balance in photosynthetically a ctive radiation (PAR), phytomass and leaf area index (LAI) were measur ed during a complete annual cycle in an annually burned African humid savanna, Directional reflectances measured by a hand-held radiometer w ere used to compute the canopy normalized difference vegetation index (NDVI). The fraction f(APAR) of PAR absorbed by the canopy (APAR) and canopy reflectances were simulated by the scattering from arbitrarily inclined leaves (SAIL) and the radiation interception in row intercrop ping (RIRI) models. The daily PAR to solar radiation ratio was linearl y related to the daily fraction of diffuse solar radiation with an ann ual value around 0.47, The observed f(APAR) was non-linearly related t o NDVI. The SAIL model simulated reasonably well directional reflectan ces but noticeably overestimated f(APAR) during most of the growing se ason. Comparison of simulations performed with the 1D and 3D versions of the RIRI model highlighted the weak influence of the heterogeneous structure of the canopy after fire and of the vertical distribution of dead and green leaves on total f(APAR). Daily f(APAR) values simulate d by the 3D-RIRI model were linearly related to and 9.8% higher than o bserved values. For sufficient soil water availability, the net produc tion efficiency epsilon(n) of the savanna grass canopy was 1.92 and 1. 28 g DM MJ(-1) APAR (where DM stands for dry matter) during early regr owth and mature stage, respectively. In conclusion, the linear relatio nship between NDVI and f(APAR) used in most primary production models operating at large scales may slightly overestimate f(APAR) by green l eaves for the humid savanna biome. Moreover, the net production effici ency of humid savannas is close to or higher than values reported for the other major natural biomes. (C) 1997 Elsevier Science B.V.