README -- Contents of /pub/geos-chem/data/GEOS_1x1/MEGAN_200909/ 09 Dec 2011 GEOS-Chem Support Team geos-chem-support@g.harvard.edu Data Download: =============================================================================== The files in this directory may be accessed by FTP download: ftp ftp.as.harvard.edu cd pub/geos-chem/data/GEOS_1x1/MEGAN_200909 Or by "wget" download: wget -r -nH "ftp://ftp.as.harvard.edu/pub/geos-chem/data/GEOS_1x1/MEGAN_200909/" Files: =============================================================================== MEGANv2.1_AEF_ALPHA_PINENE.geos.1x1 -- Annual emission factor (AEF) MEGANv2.1_AEF_BETA_PINENE.geos.1x1 files from MEGAN 2.1 inventory MEGANv2.1_AEF_CARENE.geos.1x1 for the various biogenic MEGANv2.1_AEF_ISOPRENE.geos.1x1 hydrocarbon species. MEGANv2.1_AEF_LIMONENE.geos.1x1 MEGANv2.1_AEF_MBO.geos.1x1 Data files are on the GEOS 1x1 MEGANv2.1_AEF_METHANOL.geos.1x1 grid, in binary punch (bpch) format. MEGANv2.1_AEF_MONOTERPENES.geos.1x1 MEGANv2.1_AEF_MYRCENE.geos.1x1 Files created by MEGANv2.1_AEF_NITRIC_OXIDE.geos.1x1 Mike Barkley @ U. Edinburgh MEGANv2.1_AEF_OCIMENE.geos.1x1 mailto:michael.barkley@ed.ac.uk MEGANv2.1_AEF_SABINENE.geos.1x1 MEGANv2.1_AEF_ACET_geos.1x1 -- Annual emission factor (AEF) for Acetone. This was originally created for MEGAN v2.0, but Dylan Millet says that the data has not changed and that it is appropriate to use that for MEGAN v2.1. NOTES: (1) The GEOS-Chem support team keeps a copy of the IDL programs that were used to create the above files. Contact geos-chem-support@g.harvard.edu for more information. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%% Original file "ReadMe.Overview" from Mike Barkley follows %%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% *** There should be 9 other items in this directory *** ============================================================================== 1) "emission_factors" this contains the raw & regridded MEGAN emission factors & PFT distrubtion. 2) "code_for_emission_factors" This contains the code I used to make the emission factor maps for GEOS-Chem. - Use program: make_revised_efs.pro 3) "MEGANv2.1_revised" This directory contains the MEGAN v2.1 emission factors in bpch format which should be put in to the GEOS-Chem 1x1 emission directory. ============================================================================== 3) "lai" This contains MODIS collection 5 LAI product which has been regridded to the GEOS-Chem resolutions. 4) "code_for_lai" This contains the code I used to make the MODIS LAI maps for GEOS-Chem. To re-create the LAI maps, one has to first go to the ftp site to download the raw monthly averaged data files: ftp://primavera.bu.edu/pub/datasets/MODIS/MOD15_BU/C5/LAI/data/monthly/ The programs have to be run in this order to make the LAI maps: i) modis_prep_v5.pro ii) modis2_halfdeg_v5.pro iii) modis2_halfdeg_v5_to_laiv.pro iv) do_modis_regridding_v5.pro Note: (a) modis2_halfdeg_v5_to_laiv.pro weights the LAI by the fraction of the grid cell covered by vegetation. This 'fraction' is determined by summing the MEGAN PFTs (broadleaf, needleaf, grasses, crops & shrubs). If this is greater or equal to 1.0 then we assume cell is totally covered and set the fraction = 1. The weighted LAI denoted: LAIv = ( LAI / fraction vegetated ) BUT it is set not to exceed the maximum LAI of that given month (Alex Guenther set the max LAIv = 6 which is slightly different to I've done). The program checkplot_raw_pfts.pro is included to show how I summed the PFTs. (b) some of the paths may be hardwired - check before using! (c) To read in the bpch files using gamap (I used v2.12) the following lines need to be added: ; Add in to the tracerinfo.dat file: AVHRR AVHRR LAI 0.e0 1 6901 1.0E+00 cm2/cm2 MODIS MODIS LAI 0.e0 1 6902 1.0E+00 cm2/cm2 ; Add in to the diaginfo.dat file: 6900 MODIS MODIS LAI ============================================================================== 5) "ReadMe.Modifications" This a readme file which documents the changes I've made to run using the new MODIS LAI and MEGAN v2.1 6) "Code.v8-02-01.bvoc" The modified GEOS-Chem source code 7) "run_directory" The GEOS-Chem run directory with a modifed input.geos & includes a restart file which I used for my 10yr simulation test runs. ============================================================================== 8) "example_plots" This contains some plots comparing the two BVOC emission models which are now installed. Note the file MAPS.hybrid.vs.pecca.ps shows the two BVOC emission models compared to one another and with the isoprene emissions of Muller et al., ACP, 2008. 9) "references" Directory with the journal papers & megan user guide. ============================================================================== ============================================================================== ============================================================================== ============================================================================== Here are the results of 4x5 global simulations using GEOS-4 for ISOPRENE. Code was compiled using the ifort compiler. SUed ND46 diagnostic. Hybrid model & GEOS-4 : 1996 :: 474 Tg C = 537 Tg isoprene 1997 :: 478 Tg C = 542 Tg isoprene 1998 :: 493 Tg C = 559 Tg isoprene 1999 :: 419 Tg C = 475 Tg isoprene 2000 :: 440 Tg C = 499 Tg isoprene 2001 :: 441 Tg C = 500 Tg isoprene 2002 :: 456 Tg C = 517 Tg isoprene 2003 :: 444 Tg C = 503 Tg isoprene 2004 :: 463 Tg C = 525 Tg isoprene 2005 :: 463 Tg C = 525 Tg isoprene -------------------------------------- Mean :: 457 Tg C = 518 Tg isoprene SD :: 22 Tg C = 25 Tg isoprene For 2004: Using the GEOS-4 hybrid model yields: 463 Tg C = 525 Tg isoprene. Using the GEOS-5 hybrid model yields: 461 Tg C = 522 Tg isoprene. PECCA model & GEOS-4 : : 1996 :: 385 Tg C = 436 Tg isoprene 1997 :: 380 Tg C = 431 Tg isoprene 1998 :: 390 Tg C = 442 Tg isoprene 1999 :: 324 Tg C = 367 Tg isoprene 2000 :: 342 Tg C = 388 Tg isoprene 2001 :: 348 Tg C = 394 Tg isoprene 2002 :: 356 Tg C = 403 Tg isoprene 2003 :: 347 Tg C = 393 Tg isoprene 2004 :: 372 Tg C = 422 Tg isoprene 2005 :: 372 Tg C = 422 Tg isoprene -------------------------------------- Mean :: 362 Tg C = 410 Tg isoprene SD :: 21 Tg C = 24 Tg isoprene For 2004: Using the GEOS-4 PECCA model yields: 372 Tg C = 422 Tg isoprene. Using the GEOS-5 PECCA model yields: 349 Tg C = 396 Tg isoprene. -------------------------------------------- Note for Muller's isoprene emissions estimates -------------------------------------------- 1995 = 385 Tg of isoprene per year 1996 = 374 Tg of isoprene per year 1997 = 399 Tg of isoprene per year 1998 = 449 Tg of isoprene per year 1999 = 387 Tg of isoprene per year 2000 = 414 Tg of isoprene per year 2001 = 397 Tg of isoprene per year 2002 = 397 Tg of isoprene per year 2003 = 412 Tg of isoprene per year 2004 = 421 Tg of isoprene per year 2005 = 449 Tg of isoprene per year 2006 = 424 Tg of isoprene per year Average = ~ 410 Tg of isoprene per year BUT; Muller includes soil moisture activity factor which reduces global emissions by ~20%. We dont use soil moisture activity factor. By same analogy, if we apply a 20% reduction to the 'hybrid' model the mean comes to ~415 Tg isoprene. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %%% Original README file "ReadMe.Modifications" by Mike Barkley follows %%% %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Modifications to upgrade GEOS-Chem v8-02-01 biogenic VOC emissions: _______________________________________________ Section 1: Updgrading to MODIS LAI collection 5 _______________________________________________ Programs that were modified: (1) findmon.f Now includes the current simulation & LAI years. (2) rdlai.f Now includes the current simulation & LAI years in the calls to: findmon.f & readlai.f (3) readlai.f Now includes the correct filenames for reading MODIS LAI, which needs the current year as input. (4) lai_mod.f Add a sub-routine: READIOSLAI_MODIS. Renamed sub-routine READIOSLAI to READISOLAI_AVHRR. In RDISOLAI, added year as input argument & include option to read in MODIS or AVHRR LAI. (5) logical_mod.f Include flag for MODIS. (6) input_mod.f (& therefore input.geos) Added logical switch to choose MODIS LAI. Module subroutines: read_emissions_menu & init_input (7) main.f Included the year when reading in the LAI data. _______________________________________________ Section 2: Updgrading to MEGAN v2.1 _______________________________________________ *** Overview *** MEGAN (v2.1) is based on the following formula: Emissions = Basal Emission Factors (EF) x Gamma_for_leaf_age x Gamma_for_soil_moisture x Gamma_for_Canopy_environement x Rho_canopy_prod_loss Here we will not consider the impacts of soil moisture and production and losses of the BVOC within the canopy. That is: Gamma_for_soil_moisture = 1.0 Rho_canopy_prod_loss = 1.0 To calculate the Gamma_for_Canopy_environement factor, two different approaches can be used: (i) A canopy model that calculates the light (PAR) & temp dependence at each level within the canopy. Here: Gamma_for_Canopy_environement factor = Gamma_T x Gamma_PAR x LAI (ii) A simplified parameterized scheme can be used. Here: Gamma_for_Canopy_environement factor = Gamma_T x Gamma_PAR x Gamma_LAI There are two aspects to consider: (1) Note the subtle difference in the application of the LAI. (2) In the current 5-layer canopy model only the light dependence at each leaf-level is calcalated & then summed; for a temperature the gamma factor is calculated for the entire canopy. That is in scheme (i): Gamma_for_Canopy_environement factor = [Gamma_T]_entire_canopy x SUM{ [Gamma_PAR x LAI]_leaf_level } Here we will implement both schemes: Scheme (i) will be termed the 'hybrid' canopy model. - This will be the DEFAULT option. Scheme (ii) will be termed the 'PECCA' model. - The user will be able to turn this on using a flag in input.geos Confused, eh? Well let's complicate things a little further. For isoprene and MBO both species are light-dependent so the above schemes are standard - easy to implement. For monoterpenes, these BVOCs were originally thought to show no light-dependence, however, new evidence suggests a certain light dependent fraction (LDF) might be influenced by changes in light. To account for this we have to introduce the following: Emissions = [Emissions from LDF] + [Emissions from non-LDF] where (non-LDF) = 1 - LDF In the PECCA model (scheme ii) this is really easy to implement: Emissions = [ Basal Emission Factors (EF) x Gamma_for_leaf_age x Gamma_T x Gamma_LAI ] x [ ( 1 - LDF ) + ( Gamma_PAR x LDF ) ] For the 'hybrid' canopy model (scheme i), it is slightly more 'tricky' since the light dependent part will need the canopy model (which takes into account the LAI) & the non-light dependent part will require a seasonality factor (GAMMA_LAI) to take into account changes in LAI (i.e. leaf fall and new growth). So [Emissions from LDF] = [ Basal Emission Factors (EF) x Gamma_for_leaf_age x [Gamma_T]_entire_canopy x SUM{ [Gamma_PAR x LAI]_leaf_level } x LDF ] [Emissions from non-LDF] = [ Basal Emission Factors (EF) x Gamma_for_leaf_age x [Gamma_T]_entire_canopy x Gamma_LAI x (1-LDF) ] ...which can be written more simply as: Emissions = [ Basal Emission Factors (EF) x Gamma_for_leaf_age x Gamma_T ] x [ Gamma_LAI x ( 1 - LDF ) + ( SUM{ [Gamma_PAR x LAI]_leaf_level } x LDF ) ] To implement these new emission algorithms it has been necessary to restructure the whole of the megan_module.f; there are too many changes to record in detail. ** Please closely check the module ! ** To include these NEW biogenic emissions we also have to update the following: 1) CMN_DIAG - Set new # of parameters for the diagnostic array ND46. PD46 = 13 (was orignaly 6) 2) PROGRAM: gamap_mod.f - Updated biogenic species listing for ND46. Isoprene = 1 Acetone = 2 Propene = 3 Total Monoterpenes = 4 MBO = 5 C2H4 = 6 Alpha-Pinene = 7 Beta-Pinene = 8 Limonene = 9 Sabinene = 10 Mycrene = 11 3-Carene = 12 Ocimene = 13 3) PROGRAM: emissdr.f - Now assign the BVOC species to the AD46 array & included the calls for individual monoterpene species. 4) PROGRAM: logical_mod.f & input_mod.f - Included flag for PECCA model 5) PROGRAM: carbon_mod.f - Sub-routine: BIOGENIC_OC - Modified the call to the new MEGAN monoterpenes to include the light dependence. 6) PROGRAMS: xlpardf.f & xlpardr.f - These subroutines are similar to xltmmp.f but just for PAR diffuse & direct. - I only updated the Makefile.ifort file though. _______________________________________________ Section 3: Addition of extra diagnostics _______________________________________________ Here I've added extra diagnostic output to ND49 & ND51. 1) PROGRAM: CMN_O3 - Added common array to hold BVOC emissions. Array name = EMISS_BVOC Species | Order ---------------------------- Isoprene = 1 Total Monoterpenes = 2 MBO = 3 Alpha-Pinene = 4 Beta-Pinene = 5 Limonene = 6 Sabinene = 7 Mycrene = 8 3-Carene = 9 Ocimene = 10 2) PROGRAM: emissdr.f - Added the calculated (above) emissions into the EMISS_BVOC array. 3) PROGRAM: diag51_mod.f - Added the emissions into this diagnostic Diagnostic | # ---------------------------- PAR Direct = 100 PAR Diffuse = 101 Daily LAI = 102 TS (T @ 2m) = 103 Isoprene = 104 Total Monoterpenes = 105 MBO = 106 Alpha-Pinene = 107 Beta-Pinene = 108 Limonene = 109 Sabinene = 110 Mycrene = 111 3-Carene = 112 Ocimene = 113 - Note, increased # tracers to 120 in the module subroutine: READ_ND51_MENU (in input_mod.f) 4) PROGRAM: gamap_mod.f - Add diagnostics for the activity factors (not used here) but more critcally for the Daily LAI. These are under TIME-SER Diagnostic | # ---------------------------- Gamma T = 27 Gamma_PAR = 28 Gamma_AGE = 29 Gamma_LAI = 30 Gamma_SM = 31 Daily LAI = 32 5) PROGRAM: diag49_mod.f - Added the BVOC emissions into this diagnostic - Note, increased # tracers to 120 in the module subroutine: READ_ND49_MENU (in input_mod.f) _______________________________________________ Section 4: Final comments _______________________________________________ If you struggle to find where I've changed the code do a: grep -in mpb,2008 * or grep -in mpb,2009 * This 'should' highlight my additions.