HR: 17:45h
AN: A34A-08    [Abstracts]
TI: Sensitivity of biogenic emission estimates to uncertainties in land-cover datasets: An example from Texas, USA
AU: Yang, Z
EM: liang@mail.utexas.edu
AF: Department of Geological Sciences, John R. and Katherine G. Jackson School of Geosciences, University of Texas at Austin, 1 University Station #C1100, Austin, TX 78712-0254 United States
AU: * Gulden, L
EM: gulden@mail.utexas.edu
AF: Department of Geological Sciences, John R. and Katherine G. Jackson School of Geosciences, University of Texas at Austin, 1 University Station #C1100, Austin, TX 78712-0254 United States
AU: Niu, G
EM: niu@speer.geo.utexas.edu
AF: Department of Geological Sciences, John R. and Katherine G. Jackson School of Geosciences, University of Texas at Austin, 1 University Station #C1100, Austin, TX 78712-0254 United States
AB: We use an offline land-surface model to simulate biogenic volatile organic compound (BVOC) emissions at 0.1§ resolution on a regional scale. We analyze the sensitivity of modeled biogenic emissions to the vegetation dataset used in the land-surface model. The Community Land Model version 3.0 (CLM3) is driven from 1979 to 2004 using bilinearly interpolated North American Regional Reanalysis (NARR) meteorological forcing data. We use two distinct land-cover datasets in CLM. The first run uses Moderate Resolution Imaging Spectroradiometer (MODIS)-derived vegetation plant functional types (PFTs); the second run uses a vegetation PFT dataset developed with ground survey data. Both runs employ MODIS-derived monthly leaf and stem area index values. BVOC emission capacities that are based on the emission capacities of plant species native to Texas are assigned to each PFT. The relative magnitude of the variation in the runs' biogenic emissions estimates is assessed, and the underlying reasons for the discrepancies between the two datasets are analyzed. In most locations, the ground-referenced dataset contains a higher density of trees than the satellite-derived dataset. Consequently, for a specific location, the inherent BVOC flux rate (defined as the product of the BVOC emission capacity and the leaf biomass density) of the ground-referenced dataset tends to be higher than the inherent BVOC flux rate of the satellite-derived dataset. In both datasets, BVOC flux increases west-to-east, coincident with the general trend of west-to-east increase in tree biomass density that is observed in Texas. Consistent with the work of other researchers, when forced with the same meteorological input data, using two different vegetation datasets in CLM results in modeled environments that have distinct state variables (e.g., soil moisture, leaf-surface temperature) and surface-to-atmosphere fluxes. We evaluate the extent to which these differences contribute to differences in biogenic emissions. Uncertainty of biogenic emissions estimates derived using regional climate and weather models is also discussed, and broader applicability of this study to other regional modeling applications is addressed.
DE: 0315 Biosphere/atmosphere interactions (0426, 1610)
DE: 0365 Troposphere: composition and chemistry
DE: 0416 Biogeophysics
DE: 0466 Modeling
SC: Atmospheric Sciences [A]
MN: Fall Meeting 2005