The Simple Biosphere Model, Version 3: Model Structure and Evaluation
Ian
T
Baker, Atmospheric Science, Colorado State University, kevin@atmos.colostate.edu
Scott
Denning, Atmospheric Science, Colorado State University, denning@atmos.colostate.edu
(Presenting)
Joseph
A
Berry, Carnegie Institution of Washington, joberry@globalecology.stanford.edu
James
Collatz, NASA GSFC, jcollatz@biome.gsfc.nasa.gov
Lara
Prihodko, Atmospheric Science, Colorado State University, lara@atmos.colostate.edu
Kevin
M
Schaefer, Atmospheric Science, Colorado State University, kevin@atmos.colostate.edu
Niall
Hanan, Natural Resource Ecology Laboratory, CSU, niall@nrel.colostate.edu
Neil
Suits, Atmospheric Science, Colorado State University, nsuits@atmos.colostate.edu
Modeled fluxes of energy, moisture, momentum between the atmosphere and terrestrial biosphere are critical to simulations of weather and climate. Simulated photosynthesis/CO2 flux can provide a further constraint on heat and moisture flux (as a result of stomatal resistance constaint on Bowen ratio) as well as provide essential information to chemical transport or mesoscale models.
We have modified the Simple Biosphere Model (SiB) to provide a higher level of biophysical realism; modeled surface fluxes, when confronted with eddy covariance tower observations of energy, moisture and CO2 have improved as well. Model upgrades have taken two predominant forms: 1) separation of the vegetative canopy into explicit sunlit and shaded fractions for radiative transfer and photosynthesis calculations, as opposed to a continuous representation of all vegetation, and 2) higher resolution of soil processes along with a modified calculation of soil water stress on photosynthesis.
We ran the modified SiB3 code at a variety of sites for comparison with eddy covariance observations, and results from 3 sites are presented: a midlatitude mixed forest (WLEF tall tower site, Wisconsin USA), midlatitude grassland (ARM Southern Great Plains site, Oklahoma, USA) and a tropical evergreen forest (Tapajos River, km 83 tower, Brazil). At all sites the model changes result in an improvement in the comparison of fluxes to observations, on both diurnal and annual scales.