Here is a quick overview of the emissions inventories that are currently available in GEOS-Chem.
- 1 Anthropogenic emissions
- 2 Biofuel emissions
- 3 Biogenic emissions
- 4 Biomass burning
- 5 Other NOx emissions sources
- 6 Aerosol emissions
- 7 Emissions regridding issues
EDGAR (Default inventory)
Starting with v9-01-02, the default anthropogenic emissions in GEOS-Chem come from the global EDGAR emissions inventory. In previous versions, the GEIA/Piccot inventory was the default. EDGAR can be overwritten by one or more regional emissions inventories in various locations of the globe (e.g. N. America, Asia, Europe, etc.).
RETRO anthropogenic VOC emissions are available beginning with version 9-01-02, and are the recommended default VOC emission inventories. The RETRO inventory has improved temporal and spatial resolution and a more inclusive suite of speciated NMVOCs than many of the other optional inventories.
Anthropogenic emissions of aerosols
Several aerosol species have anthropogenic emissions. Please see our aerosol emissions page for more information about these inventories.
Annual scale factors
Several of the emission inventories are for a particular year or range of years. To obtain emissions for years outside of the given range, interannual scale factors must be applied.
The default biofuel inventory is that of Yevich and Logan . This inventory can be overwritten by:
Please see our biofuel emissions wiki page for more detailed information.
Biogenic species (Isoprene, Monoterpenes, Methyl Butenol) are emitted following the MEGAN model inventory. Please see our MEGAN biogenic emissions wiki page for more information. Also see our Isoprene Emission Estimates in the Literature page for Isoprene Emission Estimates in the literature.
Originally, GEOS-Chem emitted biogenic species from the GEIA biogenic emissions inventory. This option was removed in version 9-01-03, but remains available as an option in previous versions.
GFED is the recommended biomass burning emission inventory. GFED3 will be used for model versions 9-01-02 and later, while GFED2 is available for previous versions. GFED includes emission factors for gas-phase species (NOx, CO, ALK4, ACET, MEK, ALD2, PRPE, C3H8, CH2O, C2H6), aerosol species (SO2, NH3, BC, OC), and CO2.
Most of these are also anthropogenic emissions species.
--Mcooper 12:59, 9 August 2011 (EDT)
Other NOx emissions sources
Please see the following wiki pages for information about other sources of NOx emissions:
--Bob Y. 14:40, 17 March 2010 (EDT)
Please see our Aerosol emissions wiki page for a full description of the inventories which contain aerosol species (e.g. SO2, SO4, NH3, NH4, etc.)
--Bob Y. 14:40, 17 March 2010 (EDT)
Emissions regridding issues
Different regridding approaches have been tried in order to prepare for model simulations at the GEOS-5 native resolution of 0.5x0.667, or other high resolution possibilities.
With each step in regridding, information is reorganized in such a way that it cannot be recovered by later regridding in the reverse direction, therefore, optimal regridding is done in a single step. This means that when input emissions data are provided on a given grid, the optimal approach (to avoid losing information) is to leave them on that grid, read them into the model and regrid in the code to the desired resolution. This is the approach currently used for GFEDv2 which comes in on a generic 1x1 grid (360x180).
However, at present, the model does a 2-step regridding using functions from regrid_1x1_mod.f in the following way:
1) function DO_REGRID_G2G_1x1 regrids generic_1x1 --> GEOS_1x1
2) function DO_REGRID_1x1 regrids GEOS_1x1 --> desired resolution
The above is not equivalent to the single step regridding carried out in IDL (regridH.pro).
An alternative is to use the Fortran regridding code map_a2a (developed by S.-J. Lin and refined by Bob Yantosca). I have experimented with this in the model and have it working for some situations but the code has a single option for half-polar grids. This permits direct regridding from an input half-polar grid to an output half-polar grid, or an input full-polar grid to an output full-polar grid, but does not properly regrid between full-polar and half-polar.
Figures depicting CO2 differences in the net terrestrial exchange field which was on a generic 1x1 grid then regridded to 2x2.5 for a model simulation using different approaches (IDL regridh.pro, regrid_1x1_mod, and map_a2a) are shown below, to emphasize that a single-step regridding approach is the best universal approach.
-- Ray Nassar 16:22, 10 May 2011 (EDT)