Anthropogenic emissions

Aircraft NOx

02-Oct-2007

Due to the common block error described below by Lee Murray, in GEOS-Chem v7-04-13 we have added a new module array (EMIS_AC_NOx) to "aircraft_nox_mod.f". The EMIS_AC_NOx array will be referenced in routine "setemis.f", which will pass the aircraft NOx emissions to the SMVGEAR solver. The old common-block array "GEMISNOX" will be removed.

EPA/NEI99

04-Jan-2008

The EPA/NEI99 anthropogenic emissions files have been reprocessed from scratch in order to fix the lack of emissions from the transport sector in California. This was a bug in the original EPA dataset.

The new files are located in the EPA_NEI_200708 directory. Only the anthropogenic files were reprocessed.

These new EPA/NEI99 anthropogenic emissions have not been implemented into the standard code yet. They are slated to be added shortly after v8-01-01, the next public release.

Here is the README file by Philippe Le Sager that describes the procedure.

MODIFICATION TO EPA-NEI99 emissions

10 Aug 2007

Philippe Le Sager

phs@io.as.harvard.edu

Filenames are the same as the previous NEI99 anthropogenic emissions files.

The new emissions have been updated to account for transport emissions in California, which were missing from the 1999 EPA inventory.

The monthly_average_2001_ai_mb.us36b.*.ncf files are anthropogenic emissions due to transport from 2001 inventory. They are 1x1 data set over USA. Details are in the netCDF attributes.

wk*_avg_an.1999*.geos.1x1 are the original anthropogenic emissions files used in GEOS-Chem (see original README hereafter).

Procedure to update the emissions includes

1. A mask was created to keep only the California transport emissions from the 2001 inventory.
2. The speciation of PARaffin, a 2001 category that is separated into C2H6, C3H8, ALK4+, and ACET in the 1999 inventory has been estimated from the original NEI99 inventory for consistency.
3. Data have been regridded to conform with GEOS-Chem 1x1 grid.
4. Units conversion (2001 inventory is in moles/s/box).
5. Once masked and separated into their corresponding 1999 species, the 2001 mobile emissions are added to the original NEI99 data.
6. Further regridding is done to get 4x5 and 2x2.5 data.
7. Finally, emissions are cut to get 1x1 North American grid data.

IDL routines to process 2001 emissions and add them to the original files can be found in ~phs/IDL/dvpt/EPA.

Results

• The following emissions in California are increased: NOx, CO, ALK4, PRPE, C3H8, C2H6, SO2, SO4, and NH3

01 Nov 2004

Here is the original README file by Rynda Hudman that describes the creation of the EPA/NEI99 emissions files (both anthropogenic and biofuel) that are stored in the EPA_NEI_200411 directory:

NEI99 Emission Inventory README (rch 11/1/04)

1) These files were created from NEI99 regridded emissions originally at .25X.25 resolution. These files can be found /data/epa_data. Please see README file in this directory for original file info.

The following emission types are included in the NEI99 inventory. Anthropogenic__ cow/ household_solvents/ power/ aircraft/ fertilizer/ industrial/ fowl/ industrial_solvents/ residential_fossil/ hogs/ other/ transport/

__Biofuel__ residential_biomass/

__Not updated but available in inventory__ >biogenic/ >biomass/ > >Procedures Document for Emission Inventory can be found in the group library > >2) The files are large and as such were preprocessed into hourly files using extract_epa.pro of the format: e.g., temp_cow_average.1999jul.wkend_avg.24.gz. These files can be found at /data/epa_data/EPA_TEMP/ > >3) These files were then summed over a day, summed over each source type and converted into molec/cm2/s --> to get anthropogenic and biofuel emissions for each month and regridded to 1X1 using read_regrid_epa.pro > >4) These files were then scaled. All emissions were kept as is except for the following anthropogenice emissions (not biofuel): > >;Propane(19), Ethane(21), and Acetone(9) are scaled by X6.4, X2.5, and X3 >;respectively to match prior emissions > if tracerin(i) eq 19 then scalefactor = 6.4d0 > if tracerin(i) eq 21 then scalefactor = 2.5d0 > if tracerin(i) eq 9 then scalefactor = 3.0d0 > >It is well-known that this inventory likely underestimates many VOCs. >This scaling was done using scale_to_geos_anthro.pro and scale_to_geos_bf_new.pro > >5) These emissions were then regridded using regrid_biofuel.pro and regrid_anthro.pro to create the files in 2X25 and 4X5 > >6) A mask over the U.S.(1=in U.S. and 0=not in U.S.) is used to implement these emissions since we do not have original global emission files. NEI99 is used in boxes exclusively U.S., those boxes which have Canadian and Mexican emissions are left as previous 1998 emissions. > >7) NEI99 includes both weekend and weekday emissions and those are implemented here. Diurnal scaling is done as before in GEOS-CHEM.