The impact of the selective logging in the energy-water and carbon exchange processes using optimization algorithms on the SiB2 model over a tropical forest.

HR: 1340h
AN: B43A-0128 [Abstracts]
TI: The impact of the selective logging in the energy-water and carbon exchange processes using optimization algorithms on the SiB2 model over a tropical forest.
AU: Rosolem, R
EM: rosolem@model.iag.usp.br
AF: Departamento de Ciˆncias Atmosf‚ricas, Instituto de Astronomia, Geof¡sica e Ciˆncias Atmosf‚ricas, Universidade de Sao Paulo, Rua do Matao 1226, Sao Paulo, SP 05508-900 Brazil
AU: Bastidas, L A
EM: luis.bastidas@usu.edu
AF: Department of Civil and Environmental Engineering, Utah State University, 4110 Old Main Hill, Logan, UT 84322-4110 United States
AU: * de Gon‡alves, L G
EM: gustavo@hwr.arizona.edu
AF: Department of Hydrology and Water Resources, University of Arizona, Tucson, AZ 85721 United States
AU: Shuttleworth, W J
EM: shuttle@hwr.arizona.edu
AF: Department of Hydrology and Water Resources, University of Arizona, Tucson, AZ 85721 United States
AB: Several studies have demonstrated that the deforestation in the Amazonian rainforest could lead to a significant impact in the regional and even global climate. For instance, it could change the energy, water, and CO2 cycles. In this study, we evaluate the impacts of the selective logging over an undisturbed forest using the second generation Simple Biosphere Model (SiB2). Parameter estimation and calibration plays an important role in model performance and recent studies have shown that even manually calibration can result in a better performance of the model. Therefore, automatic procedures have been applied widely within the scientific community to calibrate land surface models (LSS) such as SiB2, BATS2, etc. The MultiObjective Generalized Sensitivity Analysis (MOGSA - University of Arizona) and the MultiObjective Shuffled Complex Evolution Metropolis (MOSCEM - University of Amsterdam and University of Arizona) have been used to evaluate the differences in the sensitive parameters before and after the selective logging. Understanding this difference can provide background information on changes in some of the (physiological, soil physical, morphological) properties of the ecosystem. We have used observations from meteorological towers in the Brazilian Amazonia to constrain the model and establish preferred parameter sets that improve the model ability to simulate the water, energy, and carbon fluxes. We perform an extensive sensitivity analysis (MOGSA) and optimization (MOSCEM) of the model using multi-criteria techniques to evaluate and improve the model performance and test the transferability of the preferred parameters to different hydrological conditions (such as logging). The data used was from the site named km 83, located in Santar‚m (Par ), one of the sites where data where collected during to the Large-scale Biosphere Atmosphere (LBA) Experiment in the Amazonia, lead by Brazil.
DE: 3322 Land/atmosphere interactions
DE: 0400 Biogeosciences
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting