Consistency between MERIS and SPOT derived fAPAR to optimize ecosystem model parameters
Cédric Bacour(1), Philippe Peylin(2), Fréderic Baret(3), Marie Weiss(3), Jérôme Demarty(3), Peter Rayner(1), Frédéric Chevallier(1) and Pascal Prunet(4)
(1) LSCE, L'Orme des Merisiers, Bâtiment 701, 91191 Gif-Sur-Yvette Cedex, France
(2) INRA, Bâtiment EGER; Aile B/C 1er étage, 78850 THIVERVAL-GRIGNON, France
(3) INRA, Site Agroparc - domaine Saint-Paul, Avignon, France
(4) NOVELTIS, Parc Technologique du Canal, 31520 Ramonville-Saint-Agne, France
Significant knowledge gaps remain in our understanding of the role of the Earth's biosphere in the carbon cycle. The modeling of the biosphere functioning with process-based models for determining the distribution and probable evolution of terrestrial CO2 sources and sinks is still subject to large uncertainties. Assimilation of satellite-derived biophysical products into terrestrial ecosystem models should help improving their parameterizations more systematically and with an enhanced spatial representativeness, assuming that satellite products and model outputs are actually compatible.
We evaluate here 1) the consistency between remotely sensed estimates of fAPAR (fraction of absorbed photosynthetically active radiation) derived from medium (MERIS) and high (SPOT) spatial resolution sensors, and 2) their compatibility with the outputs of the ORCHIDEE vegetation model in terms of simulated time series of fAPAR as well as mass and energy fluxes. ORCHIDEE is a state of the art mechanistic vegetation model that can be run at local or global scales, depending on the meteorological forcing and the biome composition of the considered ecosystem. Most of ORCHIDEE parameters usually derived from biome-specific measurements are still associated with large uncertainties. A four dimensional variational data (4D-var) assimilation system has been developped in order to assimilate into ORCHIDEE the temporal variation of such ecosystem flux measurements and such satellite fAPAR estimates, alone or together.
The consistency of the model simulations with the estimated fAPAR is first assessed over an agricultural area of 10x10 km², considering weekly MERIS fAPAR (at 1km resolution) and monthly SPOT (at 40m) products. The study analyzes changes in ORCHIDEE's parameters and in optimized fAPAR time series when assimilating the various satellite products, either considering an heterogeneous pixel (resulting from the aggregation of the fAPAR data at a coarse 10x10 km² spatial resolution) or homogeneous SPOT pixels.
We then investigate the complementarity of these remotely sensed fAPAR products with in situ flux measurements at a local scale to optimize ORCHIDEE parameters. Special attention is given to a temperate broadleaf forest site (Fontainebleau, France) and a temperate needleleaf forest site (Le Bray, France) for which half-hourly eddy covariance measurements of net CO2 flux (NEE), sensitive heat flux (H), and latent heat flux (LE) were available. The compatibility between both satellite and in situ informations is appraised, especially for the seasonal evolution of these constraints and for their spatial footprint. Finally, uncertainty on the estimated model parameters is evaluated as a function of the different type of data considered.
Keywords: ESA European
Space Agency - Agence spatiale europeenne,
observation de la terre, earth observation,
satellite remote sensing,
teledetection, geophysique, altimetrie, radar,
chimique atmospherique, geophysics, altimetry, radar,