Regional Carbon Flux Simulated using the Simple Biosphere Model (SiB3)
Baker, Colorado State University, email@example.com
Prihodko, CSU, firstname.lastname@example.org
Denning, CSU, email@example.com
Goulden, UC Irvine, firstname.lastname@example.org
Miller, SUNY Albany, email@example.com
Rocha, IAG, firstname.lastname@example.org
Manzi, INPA, email@example.com
Nobre, INPE, firstname.lastname@example.org
dos Santos, INPA, email@example.com
We simulate Net Ecosystem Exchange of CO2 (NEE) using the Simple Biosphere Model (SiB3; Baker et al 2007) on a regional scale in the Amazon basin. Historically, Landsurface models have had difficulty reproducing the observed annual cycle of NEE (Saleska et al, 2003), wherein there is biospheric uptake during the dry season and efflux during seasonal rains. A number of observational and modeling studies have weighed in on the mechanisms invoked to produce the observed annual cycle, and we have previously found that by including these mechanisms into SiB3 we can obtain a reasonably accurate simulation of the annual cycle of NEE at the Kilometer 83 site on the Tapajos River (Prihodko et al, 2007).
In the current study, we extend the analysis to eight flux tower sites throughout the Amazon basin as part of an LBA Model Intercomparison Program (LBA-MIP). The sites include tropical rainforest, savannah and cropland, and annual precipitation ranges from around 900mm/year to over 2000mm/year. These simulations represent Phase 1 of LBA-MIP, and as such we do not have access to validation data other than at the Santarem Km 34 Tower and Tapajos River Km 83 Tower (from our previous work). However, even ‘blind’ runs have value in that we can evaluate differences in model performance across moisture and vegetation gradients. Ultimately, the goal is to have the ability to model both the stress-free situation in the deep tropics as well as the areas that experience seasonal stress on the dry side of the ecotone. These initial simulations can assist our efforts to determine how model physics interact with the simulation of seasonal stress, which will then lead us toward model mechanisms and parameter values necessary to improve our understanding of the biomechanics of this region of the planet.