Biomass burning emissions

This page describes the options for biomass burning emissions in GEOS-Chem.

GFED3

GFED3 monthly data

GEOS-Chem v9-01-02 and higher versions will contain the new GFED3 biomass inventory. For more information about this data set, please see the following:

1. GFED web page.
2. Guido van der Werf's GFED3 web page.
3. van der Werf et al (2010). This is the reference paper for the GFED3 inventory.

Prasad Kasibhatla and the GEOS-Chem Support Team implemented GFED3 into GEOS-Chem v9-01-02.

As of October 2011, GFED3 monthly data is available for years 1997-2010. The data is available in bpch format from the following directory structure:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/data/GEOS_1x1/GFED3_201110/ 

--Bob Y. 14:02, 31 March 2011 (EDT)
--Melissa Payer 15:03, 21 February 2012 (EST)

GFED3 daily and 3-hourly emissions

Original implementation

This update was tested in the 1-month benchmark simulation v9-01-03l and approved on 18 May 2012.

Prasad Kasibhatla wrote:

I have updated GEOS-Chem to incorporate the capabilty to use GFED3 daily and 3-hourly fire fractions, and to regrid on the fly from the GFED3 native resolution (0.5x0.5) to the model resolution.

NOTES:

• Daily and 3hourly fractions are only available for 2003-2010. For methodology, see

Mu, M., J.T. Randerson, G.R. van der Werf, L. Giglio, P. Kasibhatla, D. Morton, G.J. Collatz, R.S. DeFries, E.J. Hyer, E.M. Prins, D.W.T. Griffith, D. Wunch, G.C. Toon, V. Sherlock, and P.O. Wennberg. 2011. Daily and 3-hourly variability in global fire emissions and consequences for atmospheric model predictions of carbon monoxide. Journal of Geophysical Research-Atmospheres. 116: D24303. doi:10.1029/2011JD016245.

• Since fire fraction are not available as a function of type of fire type (as is the case for dry matter burnt), the same fractions are applied to all 6 fire types (ag waste, deforestation, extratropical forest, peat, savanna, and woodland).
• I took the opportunity to rewrite gfed3_biomass_mod.F which mainly involved cleaning up code but with one big change - regridding directly from the GFED3 native resolution to the model resolution by adapting the MAP_A2A regriddng package mentioned in Bob's Model Engineer Report.
• As a result of the change to the regridding, the GFED3 data file archive has to be updated to bpch files at the native resolution. I have these ready and will supply to GC support team. Note that these require the creation of new directory structure since they do not belong in the GEOS_1x1 data directory.
• GFED3 for 2011 is likely to be available in late spring-early summer. An added data layer for fire emissions that may be missed by the burnt area algorithm is currently being developed.
• There are occasional cells where the burnt dry matter seems to unrealistically high. For example in Dec 2009, the cell centered at 13.25 degree lat, 123.75 degree lon has DM = 13054 g DM/m²/month. The next highest grid box that month is about 723 g DM/m²/month. The GFED team is aware of this and is investigating.

--Melissa Payer 15:03, 21 February 2012 (EST)

Update 3/9/12:

Prasad Kasibhatla wrote:

I have updated the GFED3 patch file to:

1. revise the regridding-on-the-fly algorithm - the orginal MAP_A2A algorithm results in a bit of incorrect smearing when regridding grid-average fluxes.
2. add code so that the GFED3 code works correctly for nested grid simulations
3. correct a small error - in a few tropical forest boxes, woodland (rather than tropical forest) emissions factros were being applied.

--Melissa Payer 16:13, 9 March 2012 (EST)

Bug fix for 3-hourly GFED3 emissions

This update was tested in the 1-month benchmark simulation v9-02l and approved on 26 Jun 2013.

Shannon Koplitz wrote:

I think there might be a bug in gfed3_biomass_mod.F when trying to run with 3-hourly GFED emissions. Running with 3-hourly emissions from the first day of a month seems to work fine, but trying to start on any other day appears to cause a problem. This didn't happen when running with daily emissions and I think it might have to do with how the monthly and 3-hourly emissions are coupled.

Christoph Keller wrote:

I think I know what's going on. Since the same 3-hourly fractions are used throughout the entire month, the fractions are only read once per month.

To construct the 3-hourly filename, the YYYYMMDD string is used (In gfed3_biomass_mod.F, line 670), which is hardcoded to have the day set to 1 (this happens on line 542). However, the TAU0 value that is passed to READ_BPCH2_GFED3 is constructed using the current day (DD), which can differ from 1 (DD is defined on line 505, the corresponding TAU0 on line 668). So TAU0 does not match the time stamp in the 3-hourly file if the starting day differs from the first of the month, and you get the error in READ_BPCH2_GFED3 that you observe.

If you hardcode DD to 1 in GET_TAU0 on line 668 I would expect that the error disappears. So just use

TAU0 = GET_TAU0( MM, 1, YYYY, IT3H )

Instead of

TAU0 = GET_TAU0( MM, DD, YYYY, IT3H )

--Melissa Payer 15:02, 16 April 2013 (EDT)

Update GFED3 emissions to 2011

This update was tested in the 1-month benchmark simulation v9-02m and approved on 30 Jul 2013.

Prasad Kasibhatla wrote:

I have updated the GFED3 database to include data for 2011. Please note that GPCP precip data (one of the input datasets needed for GFED3) was not available for all of 2011 - the ERA-Iterim precip (http://climatedataguide.ucar.edu/reanalysis/era-interim) was therefore used for the second half of 2011.

There were also small isolated changes prior to 2011 in this version of GFED3. I have therefore reprocessed the entire 1997-2011 GFED3 dataset for GEOS-Chem, now stored in at Harvard in the:

  ftp//ftp.as.harvard.edu/gcgrid/data/GEOS_NATIVE/GFED3_201212/  

directory structure.

If you wish to use this updated GFED3 data, please also modify the following line in the module file GeosCore/gfed3_biomass_mod.F, from:

  CALL GFED3_AVAILABLE( YYYY, 1997, 2010 )

to

  CALL GFED3_AVAILABLE( YYYY, 1997, 2011 )

Please note that this is the last update for GFED3. Plans are underway for GFED4, whcih will ultimately replace the pre-2011 GFED3 dataset and go forward in time post-2011.

Bob Yantosca wrote:

Please note, we are placing this data into the GEOS_NATIVE/GFED3_201212 subdirectory so that it will not conflict with the existing GFED3 data in GEOS_NATIVE/GFED3_201203. By not overwriting the data, we will be able to replicate GEOS-Chem results made with prior model versions using the data in GFED3_201203.

We will add support for the updated GFED3 data to the standard GEOS-Chem along with other emissions updates (slated for GEOS-Chem_v9-02), and we will run our customary 1-month benchmark to test the results. In the meantime, if you would like to use the new GFED3 data for 2011, then please download the data to your location and make the above-mentioned update in gfed3_biomass_mod.F in your research code.

--Melissa Payer 18:02, 18 December 2012 (EST)

GFED3 data files and GAMAP

If you want to read the GFED3 dry matter files with GAMAP, please make sure you add the following lines to these files:

In IDL/gamap2/input_files/diaginfo.dat, add the line:

    8000 GFED3-BB                                 GFED3 biomass burning

In IDL/gamap2/input_files/tracerinfo.dat files, add the lines:

#==============================================================================
# OFFSET=8000: GFED3 dry matter emissions 
#==============================================================================
AGW_DM   AGW dry matter emissions            0.e0  1     8091  1.0E+00 g/m2/month
DEF_DM   DEF dry matter emissions            0.e0  1     8092  1.0E+00 g/m2/month
FOR_DM   FOR dry matter emissions            0.e0  1     8093  1.0E+00 g/m2/month
PET_DM   PET dry matter emissions            0.e0  1     8094  1.0E+00 g/m2/month
SAV_DM   SAV dry matter emissions            0.e0  1     8095  1.0E+00 g/m2/month
WDL_DM   WDL dry matter emissions            0.e0  1     8096  1.0E+00 g/m2/month

These modifications will be made in GAMAP v2-16 and higher versions.

--Bob Y. 16:37, 30 January 2012 (EST)

FINNv1

"The Fire INventory from NCAR version 1.0 (FINNv1) provides daily, 1-km resolution, global estimates of the trace gas and particle emissions from open burning of biomass, which includes wildfire, agricultural fires, and pre- scribed burning and does not include biofuel use and trash burning" (Wiedinmyer et al., 2011). For more information about this inventory, see:
Wiedinmyer, C., S. K. Akagi, R .J. Yokelson, L. K. Emmons, J. A. Al-Saadi, J. J. Orlando, and A. J. Soja. The Fire INventory from NCAR (FINN): a high resolution global model to estimate the emissions from open burning. Geosci. Model Dev., 4, 625–641, 2011. [1]

Implementation in GEOS-Chem

Jenny Fisher (U. Wollongong) and Min Huang (JPL) are working with Christine Wiedinmyer to implement FINNv1 into the standard version of GEOS-Chem. This implementation will include a number of features, many of which are improvements over the non-standard implementations currently used (see below):

1. Improved speciation for GEOS-Chem, including specialty simulations like SOA
2. Compatibility with new emissions module (including use of netcdf, high-resolution inputs, online regridding, etc.)
3. GFED-like online computation of emissions for most species based on emissions of a single species for each land type (significantly reducing file storage requirements)
4. Ability to use daily or monthly emissions

If you have questions or suggestions, please contact Jenny Fisher (jennyf@uow.edu.au) and Min Huang (Min.Huang@jpl.nasa.gov).

Current non-standard implementations

As of IGC6 (May 2013), FINNv1 has been implemented in non-standard versions of GEOS-Chem by at least 3 groups. These implementations allow users to use the ASCII emissions files (or a converted gridded netcdf equivalent) available directly from http://bai.acd.ucar.edu/Data/fire/. If you are interested in working with FINNv1 emissions immediately, please contact Jingqiu Mao (Jingqiu.Mao@noaa.gov), Min Huang (Min.Huang@jpl.nasa.gov), or Jenny Fisher (jennyf@uow.edu.au).

GFED2

NOTE: Once the GFED3 daily and hourly biomass burning emissions are installed into GEOS-Chem v9-01-03, then GFED2 will be removed.

GFED2 monthly data thru 2008

As of July 2009, the GFED2 monthly data is now available for years 1997-2008. The data is available in bpch format from the following directory structure:

ftp ftp.as.harvard.edu
cd pub/geos-chem/data/GEOS_1x1/GFED2_200601/

For more information about the data, please see the following README file: GEOS_1x1/GFED2_200601/README

You will need to make a small modification in routine GFED2_COMPUTE_BIOMASS (in gfed2_biomass_mod.f) in order to update the maximum available year of data. Look for the lines:

     ! Availability of MONTHLY data
     !-------------------------------
     ELSE IF ( LGFED2BB ) THEN
        
        CALL GFED2_AVAILABLE( YYYY, 1997, 2007 )

and change the 2007 to 2008:

     ! Availability of MONTHLY data
     !-------------------------------
     ELSE IF ( LGFED2BB ) THEN
        
        !-----------------------------------------------------------
        ! Prior to 7/8/09:
        ! GFED2 2008 monthly data is now available (bmy, 7/8/09)
        !CALL GFED2_AVAILABLE( YYYY, 1997, 2007 )
        !-----------------------------------------------------------
        CALL GFED2_AVAILABLE( YYYY, 1997, 2008 )

You may also download the updated version of gfed2_biomass_mod.f from:

ftp ftp.as.harvard.edu
cd pub/geos-chem/patches/v8-02-02
get gfed2_biomass_mod.f

--Bob Y. 10:32, 24 February 2010 (EST)

GFED2 8-day emissions

The GFED2 8-day emissions are now included in the standard model as a user option as of GEOS-Chem v8-02-01. The data files for years 2001-2007 may be downloaded from:

ftp ftp.as.harvard.edu
cd pub/geos-chem/data/GEOS_1x1/GFED2_8day_200712/

For more information on the GFED2 8-day emissions data, please see the README file: GEOS_1x1/GFED2_8day_200712/README

Ray Nassar wrote:

The [GFED2 8-day emissions] code is working and I am using it for my work. Overall, the results I get for CO and ozone are very similar to my monthly GFED runs but this changes if one focuses on a specific region over with a short time scale.

{Here are} ... Some of {my comparisons}. This is by no means a complete analysis but more of an initial assessment.

In my opinion, the move to 8-day emissions is a definite improvement, although as I said, the differences are fairly localized and really only matter for time scales shorter than one month.

--Bob Y. 10:32, 24 February 2010 (EST)

FLAMBE biomass emissions

NOTE: At this time FLAMBE has not been implemented into the mainline standard GEOS-Chem, but it has been used for the NRT-ARCTAS codes. Therefore probably only a small fraction of GEOS-Chem users will be concerned with these emissions at this time.

Data issues

Jenny Fisher wrote:

Sorry to revisit the old FLAMBE issues, but we are seeing some strange results in the fire emissions, and I am trying to verify with Ed Hyer that we are processing the data correctly.

To get a handle on how we are processing the file, I have been looking in Philippe's directory ~phs/IDL/dvpt/flambe/

I took a look at some of the data files in the data/ directory in there. In the column that corresponds to carbon emissions (column 11, or 10 in IDL accounting), I see values that range from 4.95e4 to 765e4. Hyer tells me that in his version of the files, these range from 4.95 to 7650 (i.e. 4 orders of magnitude different). We are assuming these emissions are in g/m2, which they clearly aren't at our values.

Do either of you know if I am looking at old files, and if what we have looks more like what he has? Or are we processing things 4 orders of magnitude too large?? Or, are we assuming a different unit on the emissions when we actually process them? I can't seem to find raw flambe data files anywhere besides Philippe's development directory...

Philippe Le Sager wrote:

We had a couple of problems with the data. The one that gave me headache was the difference between forecast and analysis data, which can be mixed in some files. There was the issue of double counting fires seen by both GOES and MODIS, and we found a problem with unit. It was Kg instead of g as advertised in the ppt [that was given to us] (the only documentation for the data). Ed did correct the files and back processed the data on its server.

I have a "new" directory in my test data. You probably look at the old set of data. The new one is in the "new" subdirectory:

   /home/phs/IDL/dvpt/flambe/data/new/

You can also still get the data at the following website:

   http://www.nrlmry.navy.mil/aerosol_web/arctas_flambe/data_hourly/

Finally here are some totals we check w/ Maria and Jingqiu:

   Data for one set of satellites was still given Kg.  Now fixed, I have the following
   total for one day all over the world, assuming 3 land types. Seems reasonable, no?

   NOx (Tg N) assuming ALL is savanna / trop forest / extraTrop forest:
     0.080809369     0.063615890      0.10316091

   CO (Tg) assuming ALL is savanna / trop forest / extraTrop forest:
     4.6422410       7.5160097       7.8107557

   SMOKE (Tg) assuming ALL is savanna / trop forest / extraTrop forest:
     0.47039473      0.47039473      0.47039473

--Bob Y. 15:07, 18 February 2009 (EST)

Obtaining the FLAMBE data

Raw data

The FLAMBE biomass burning data files are available at the following archive: http://www.nrlmry.navy.mil/aerosol_web/arctas_flambe/data_hourly/.

Bob Yantosca has some scripts that can be used to download the data (contact him for more info). These are:

sleepFlambe

Perl script which issues a Unix wget command to download one day (24 hourly files) of FLAMBE data to disk.

wrap_flambe.pro

IDL batch file which acts as a driver for looping over all 24 hours of FLAMBE data for a given date.

read_flambe.pro

IDL program which reads each FLAMBE "raw" data file and saves it to GEOS-Chem bpch format.

Processed data

The processed FLAMBE data (for input to GEOS-Chem) are kept on the Harvard data archive in the directories:

ftp ftp.as.harvard.edu
cd /pub/geos-chem/NRT-ARCTAS/flambe/YYYY/MM

where YYYY/MM are the year and month of the data. Currently we have the full data archive from 2008 and 2009, with some months of 2010.

--Bob Y. 14:32, 10 February 2010 (EST)

Duncan et al "Seasonal" Biomass Emissions

See Bey et al [2001].

References

1. Duncan, B.N., et al., Interannual and Seasonal Variability of Biomass Burning Emissions Constrained by Satellite Observations, J. Geophys. Res., 108(D2), 4040, doi:10.1029/2002JD002378, 2003. PDF
2. Hyer, E., FLAMBE Biomass Burning emissions for ARCTAS, 2008. PDF
3. Lobert, J. M., W. C. Keene, J. A. Logan, and R. Yevich, Global chlorine emissions from biomass burning: the reactive chlorine emissions inventory, J. Geophys. Res., 8, 2999-3014, 2008.
4. Mu, M., J.T. Randerson, G.R. van der Werf, L. Giglio, P. Kasibhatla, D. Morton, G.J. Collatz, R.S. DeFries, E.J. Hyer, E.M. Prins, D.W.T. Griffith, D. Wunch, G.C. Toon, V. Sherlock, and P.O. Wennberg, Daily and 3-hourly variability in global fire emissions and consequences for atmospheric model predictions of carbon monoxide, Journal of Geophysical Research-Atmospheres, 116, D24303, doi:10.1029/2011JD016245, 2011.
5. Nassar, R., J. A. Logan, I. A. Megretskaia, L. T. Murray, L. Zhang, and D. B. A. Jones, Analysis of tropical tropospheric ozone, carbon monoxide and water vapor during the 2006 El Niño using TES observations and the GEOS-Chem model, J. Geophys. Res., 114, D17304, doi:10.1029/2009JD011760, 2009. PDF
6. van der Werf, G., J.T. Randerson, L. Giglio, G.J. Collatz, M. Mu, P.S. Kasibhatla, D.C. Morton, R.S. DeFries, Y. Jin, and T. T. van Leeuwen, Global fire emissions and the contribution of deforestation, savanna, forest, agricultural, and peat fires (1997–2009), Atmos. Chem. Phys. Discuss., 10, 16153-16230, 2010. PDF

--Melissa Payer 15:10, 21 February 2012 (EST)