HR: 16:00h
AN: B54B-01    [Abstracts]
TI: Calibration of a land surface carbon flux model against data from two nearby sites in Amazonia: what explains results so different?
AU: Imbuzeiro, H
EM: hew@vicosa.ufv.br
AF: Federal University of Vicosa, Department of Agricultural Engineering Av. P.H. Rolfs, s/n, Vicosa, MG 36570-000 Brazil
AU: * Costa, M H
EM: mhcosta@ufv.br
AF: Federal University of Vicosa, Department of Agricultural Engineering Av. P.H. Rolfs, s/n, Vicosa, MG 36570-000 Brazil
AU: Baleeiro, G C
EM: harles_botelho@yahoo.com
AF: Federal University of Vicosa, Department of Agricultural Engineering Av. P.H. Rolfs, s/n, Vicosa, MG 36570-000 Brazil
AU: Rocha, H R
EM: humberto@model.iag.usp.br
AF: University of Sao Paulo, Department of Atmospheric Sciences Rua do Matao 1226, Sao Paulo, SP 05508-900 Brazil
AU: Saleska, S R
EM: saleska@email.arizona.edu
AF: University of Arizona, Dept. of Ecology and Evolutionary Biology 1041 E. Lowell Street, Tucson, AZ 85721 United States
AU: Goulden, M
EM: mgoulden@uci.edu
AF: University of California Irvine, Dept. of Earth System Science, Irvine, CA 92697-3100 United States
AU: Hutira, L
EM: lhutyra@fas.harvard.edu
AF: Harvard University, Department of Earth & Planetary Sciences 24 Oxford Street, Cambridge, MA 02138 United States
AU: Miller, S D
EM: sdmiller@uci.edu
AF: University of California Irvine, Dept. of Earth System Science, Irvine, CA 92697-3100 United States
AB: Carbon cycle models (CCMs), properly calibrated against field data, are one of the techniques used to estimate the regional carbon cycle. Traditionally in the field, CCMs are usually calibrated against data collected in one single site per ecosystem. Given the availability of LBA data for Amazonia, in this study, however, we challenge this tradition and calibrate the CCM IBIS against data from two nearby sites in Amazonia. Both sites are close to the city of Santarem, state of Para, Brazil, one close to km67 and the other close to km83 of the Cuiaba-Santarem highway. Sites are about 15 km apart, and both have primary tropical rainforest vegetation with very similar canopy structure. The process of calibration aims at minimizing the RMSE for carbon fluxes, as well as latent and sensible heat fluxes. Results for both sites are significantly different: the parameters obtained for km67 indicate a canopy with a low capacity to fix carbon and low canopy conductance, while the parameters obtained for km83 indicate the inverse, a canopy with a very high capacity to fix carbon and high canopy conductance. The lower daily peaks of CO2 flux, and the more uniform night carbon loss at km67, when compared to km83, apparently constrains the physiological parameters that control carbon assimilation and ecosystem respiration. These results are unexpected, considering the proximity and similarity of the two sites. Although it is not clear why two different sets of parameters are obtained, the possibilities include errors in the representation of soil physical parameters, and differences in the design, installation and operation of the experimental sites, as well as the post-processing of the data collected in each site.
DE: 0426 Biosphere/atmosphere interactions (0315)
DE: 0428 Carbon cycling (4806)
DE: 0466 Modeling
SC: Biogeosciences [B]
MN: Fall Meeting 2005