MERRA-2 implementation details

Overview

MERRA2 is the newest reanalysis meteorological data product from NASA/GMAO. Overall, the structure of MERRA2 is very similar to the GMAO "forward processing" data product, which we refer to as GEOS-FP. Therefore, within GEOS-Chem we can (for the most part) treat MERRA2 in the same way as we do for the GEOS_FP.

Comparison between MERRA-2, GEOS-FP, MERRA, and GEOS-5

Feature	MERRA2	GEOS-FP	MERRA	GEOS-5
Native vertical grid	0.5° lat x 0.625° lon	0.25° lat x 0.3125° lon	0.5° lat x 0.667° lon 1° lat x 1.25° lon 1.25° x 1.25° lon	0.5° lat x 0.667° lon
Native horizontal grid	72 hybrid levels	Same grid as MERRA2	Same grid as MERRA2	Same grid as MERRA2
Time-invariant fields	FRLAKE FRLAND FRLANDIC(E) FROCEAN PHIS	Same fields as MERRA2	Same fields as MERRA2	Same fields as MERRA2
Precipitation fields	DQRCU, DQRLSAN, PFICU, PFILSAN, PFLCU, PFLLSAN, REEVAPCN, REEVAPLSAN	Same fields as MERRA2	Same fields as MERRA2	DQIDTMST DQLDTMST DQVDTMST MOISTQ = DQIDTMST + DQLDTMST + DQVDTMST
Time archiving: surface fields	A1 (1-hr time averaged)	Same as MERRA2	Same as MERRA2	A3 (6-hr time average)
Time archiving: 3D fields	A3 (3-hr time averaged) I3 (3-hr instantaneous)	Same as MERRA2 Same as MERRA2	A3 (3-hr time average) I6 (6-hour instantaneous)	A6 (6-hr time average) and I6 (6-hr instantaneous)
Temperature	T [K] I3 (3-hr instantaneous)	Same as MERRA2 Same as MERRA2	Same as MERRA2 A3 (6-hr time average)	Same as MERRA2 A6 (6-hr time average)
Relative humidity	RH [1] A3 (3-hr time average)	Same as MERRA2 Same as MERRA2	Same as MERRA2 I3 (6-hr instantaneous)	Same as MERRA2 A6 (6-hr time-averaged)
Specific humidity	QV [kg kg-1] I3 (3-hr instantaneous)	Same as MERRA2 Same as MERRA2	Same as MERRA2 A3 (3-hr time average)	Same as MERRA2 A6 (6-hr time-averaged)
Tropopause pressure	TROPPT [Pa] A1 (1-hr time average)	TROPPT [hPa] Same as MERRA2	TROPPT [hPa] Same as MERRA2	TROPP [hPa] A3 (3-hr time average)
Ertel potential vorticity	PV [K m2 kg-1 s-1] A1 (1-hr time average)	Same as MERRA2 Same as MERRA2	Same as MERRA2 I6 (6-hr instantaneous)	TROPP (hPa) A6 (3-hr time-average)
GEOS-Chem convection routine	DO_MERRA_CONVECTION (in convection_mod.F)	Same as MERRA2	Same as MERRA2	NFCLDMX (in convection_mod.F)
GEOS-Chem wet deposition routine	WETDEP_MERRA (in wetscav_mod.F)	Same as MERRA2	Same as MERRA2	WETDEP (in wetscav_mod.F)
Wet deposition algorithm	72 hybrid levels	Identical to MERRA2	Identical to MERRA2	Identical to MERRA2

The following aspects of [[MERRA2] are identical to the older GEOS-FP operational data product.

1. The MERRA2 vertical grid (72 hybrid levels) is identical to that of both MERRA and GEOS-5.2.0.
2. By and large, the same data fields that were in the MERRA product are also contained in GEOS-FP:
   • GEOS-FP provides the same time-invariant fields (FRLAKE, FRLAND, FRLANDIC, FROCEAN, PHIS) as MERRA
   • The same correction that was applied to the MERRA SNOMAS (snow mass) field has also been applied to the GEOS-FP SNOMAS field.
   • GEOS-FP contains the same precipitation fields (DQRCU, DQRLSAN, PFICU, PFILSAN, PFLCU, PFLLSAN, REEVAPCN, REEVAPLSAN) as MERRA. This means that we can use the MERRA wet deposition algorithms with the GEOS-FP data.
3. The GEOS-FP time archiving is identical to that of MERRA ...
   • ... with one exception: GEOS-FP instantaneous data is archived every 3 hours, as opposed to 6 hours in MERRA.

--Bob Y. 09:46, 21 August 2013 (EDT)

Key differences between GEOS-

The following aspects of GEOS-FP differ from the MERRA operational data product:

1. All GEOS-FP data are placed on a horizontal grid of 0.25° x 0.3125°.
   • MERRA data fields are placed on three different horizontal grids (0.5° x 0.666°, 1° x 1.25° and 1.25° x 1.25°).
2. All GEOS-FP data are placed on the 72-level hybrid grid.
   • Some MERRA fields were only available on a 42-level pressure grid (thus requiring interpolation).
3. The following data GEOS-FP data fields differ from the corresponding MERRA fields:
   • RH is now a 3-hour time-averaged field, as opposed to a 6-hour instantaneous field in MERRA.
   • QV and T are now instantaneous 3-hour fields, as opposed to 3-hour time-averaged fields in MERRA.
   • The 3-D precipitation fields now make the (e.g. DQIDTMST, DQVDTMST, DQLDTMST, etc.) obsolete. We no longer archive these.
4. We have chosen to save the processed GEOS-FP data in netCDF format, as opposed to GEOS-Chem binary format.

--Bob Y. 09:47, 21 August 2013 (EDT)

Required coding changes in GEOS-Chem

As noted in the section above, the similarities between MERRA2 and GEOS-FP (especially of the vertical grids) means that in most cases we can just treat MERRA2 in the same way as for GEOS-FP. In many areas of the code, the coding changes are straightforward and simply involve an extension of C-preprocessor statements such as:

#if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP )

to

#if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )

In other areas of GEOS-Chem the changes are more involved. Below we provide a description of the areas of GEOS-Chem that were modified for MERRA2:

Top-level directory

The following modifications for MERRA2 were made in the root GEOS-Chem directory:

File	Modifications made
Makefile_header.mk	If MET=merra2 or MET=merra-2 (case-insensitive), then activate the MERRA2 C-preprocessor switch i.e. Add -DMERRA2 to the Makefile variable USER_DEFS. If GRID=05x0625 is added, then activate the GRID05x0625 C-preprocessor switch i.e. Add -DGRID05x0625 to the Makefile variable USER_DEFS.

--Bob Y. (talk) 18:28, 12 August 2015 (UTC)

Headers directory

The following modifications for MERRA2 were made in the Headers/ directory:

File	Modifications made
CMN_SIZE_mod.F	Added size parameters for GEOS-FP grids at 4° x 5°, 2° x 2.5°, and 0.25° x 0.3125° resolution
gigc_input_opt_mod.F90	Add MERRA2_DIR to the OptInput derived type Set Input_Opt%MERRA2_DIR = './' in routine Init_GIGC_Input_Opt
gigc_state_met_mod.F90	Multiple instances: Now use: #if defined( GEOS_FP ) || defined( MERRA ) || defined( MERRA2 )

--Bob Y. (talk) 18:28, 12 August 2015 (UTC)

GeosUtil directory

The following modifications for MERRA2 were made in the GeosUtil/ directory:

File	Modifications made
bpch2_mod.F	In routine GET_MODELNAME: Add MERRA2_47L and MERRA2 modelnames to the #if block In routine GET_NAME_EXT: Return grid name merra2 for MERRA2 fields In routine GET_NAME_EXT_2D: Now use #if defined( MERRA ) || defined( MERRA2 ) In routine GET_RES_EXT: Return resolution string 05x0625 for 0.5° x 0.625° grids
pressure_mod.F	In routines GET_PEDGE and GET_PEDGE_FULLGRID: Compute PEDGE for MERRA2 in the same way as for GEOS-5, MERRA, and GEOS-FP In routine INIT_PRESSURE: Initialize Ap and Bp for MERRA2 in the same way as for GEOS-5, MERRA, and GEOS-FP
time_mod.F	In routine GET_A1_TIME: Compute the DATE array for MERRA2 the same way as is done for MERRA and GEOS-FP
transfer_mod.F	In routine INIT_TRANSFER: Set L_COPY = 36 for MERRA2 Changed code to #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In several routines: Changed code to #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )

--Bob Y. (talk) 20:33, 12 August 2015 (UTC)

GeosCore directory

The following modifications for MERRA2 were made in the GeosCore/ directory:

File	Modifications made
calcrate.F	Multiple instances: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
carbon_mod.F	Multiple instances: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
cldice_HbrHOBr_mod.F	Multiple instances: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
convection_mod.F	In routine DO_CONVECTION: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routine DO_MERRA_CONVECTION: Changed code to #if defined( GEOS_FP ) || defined( MERRA2 ) where we set PFICU and PFLCU
dao_mod.F	In routine COPY_I6_FIELDS: Changed code to: #if defined( GEOS_FP ) || defined( MERRA2 ) In routine INTERP: Changed code to: #if defined( GEOS_FP ) || defined( MERRA2 )
depo_mercury_mod.F	In routine ADD_Hg2_SNOWPACK Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
dao_mod.F	In routine COPY_I6_FIELDS: Changed code to: #if defined( GEOS_FP ) || defined( MERRA2 ) In routine INTERP: Changed code to: #if defined( GEOS_FP ) || defined( MERRA2 )
diag3.F	Multiple instances: Changed code to: #if defined( GEOS_FP ) || defined( MERRA2 ), etc.
dust_dead_mod.F	In routine DST_MBL: For GEOS-FP, add a new value for parameter FLX_MSS_FDG_FCT. The scale factor ensures that we get the same dust totals (w/in roundoff error) as for GEOS-5.
fast_jx_mod.F	In routine FAST_JX: Do cloud overlapping for GEOS-FP in the same way as for GEOS-5 and MERRA. Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
gamap_mod.F	In routine INIT_TRACERINFO: Changed code to: #elif defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP )
global_ch4_mod.F	In routines WETLAND_EMIS and READ_COPROD: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP )
input_mod.F	In routine READ_SIMULATION_MENU: Now read value for Input_Opt%MERRA2_DIR from input.geos Also set Input_Opt%RES_DIR accordingly for 0.5° x 0.625° nested grids Set LUNZIP=.FALSE. for MERRA2 met In routine READ_CONVECTION_MENU: Change #if statement to allow non-local PBL mixing for MERRA2 met data In routine VALIDATE_DIRECTORIES: Error check MERRA2_DIR for start & end date
land_mercury_mod.F	In routines LAND_MERCURY_FLUX: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routines SOILEMIS: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) Extend #if statement for variable SOIL_EMIS_FAC to get the code to compile w/o error. Team Hg will need to supply an updated value later on.
main.F	In main.F: Changed code to if defined( GEOS_FP ) || defined( MERRA2 ) where we define N_DYN_STEPS Changed code to if defined( GEOS_FP ) || defined( MERRA2 ) where we define Input_Opt%USE_O3_FROM_MET In included routine READ_INITIAL_MET_FIELDS: Added calls to routines from merra2_read_mod.F90 In included routine READ_MET_FIELDS: Added calls to routines from merra2_read_mod.F90
mercury_mod.F	In routine CHEMMERCURY: Add placeholder K_RED_JNO2 parameter for the 0.5° x 0.625° nested grids. For now use the same value as for the 0.5° x 0.666° grid. Team Hg will need to supply an updated value later on. Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
merra2_read_mod.F90	Added this module w/ routines to read GEOS-FP met data (in netCDF format)
ocean_mercury_mod.F	Multiple instances: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
planeflight_mod.F	In routine PLANEFLIGHT Now use: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
strat_chem_mod.F90	Multiple instances: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routine DO_SYNOZ: Add placeholder values for 0.5° x 0.625° grids
sulfate_mod.F	In routine CHEM_SO2: Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routine INIT_SULFATE: Changed code to: #if !defined( GEOS_5 ) && !defined( MERRA ) && !defined( GEOS_FP ) && !defined( MERRA2 ) In routine SRCSFC30 (TOMAS-only): Added a placeholder value for TSCALE for the 0.5° x 0.625° grids
tagged_ox_mod.F	In routine GET_REGIONAL_POX: Now use: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
tpcore_bc_mod.F:	Multiple instances: Added MERRA2 placeholder values for I1_BC_CH, I2_BC_CH, J1_BC_CH, J2_BC_CH Added MERRA2 placeholder values for I1_BC_EU, I2_BC_EU, J1_BC_EU, J2_BC_EU Added MERRA2 placeholder values for I1_BC_NA, I2_BC_NA, J1_BC_NA, J2_BC_NA
vdiff_mod.F90	In routine VDIFFDR: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
wetscav_mod.F	In routine DO_WETDEP: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routine DO_WASHOUT_ONLY:: Changed code to: #if defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 ) In routine MAKE_QQ: Changed code to: #defined( GEOS_FP ) || defined( MERRA2 ) In routine RAINOUT: Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )

--Bob Y. (talk) 18:29, 12 August 2015 (UTC)

HEMCO/Core directory

The following modifications for MERRA2 were made in the HEMCO/Core directory:

File	Modifications made
hco_chartools_mod.F90	At the top of the module: Extend the #ifdef statement to add the parameter DEF_MET = 'merra2' Extend the #ifdef statement to add the parameter DEF_RES = '05x0625'
hco_interp_mod.F90	In routine MODLEV_INTERPOLATE: Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )

--Bob Y. (talk) 18:28, 12 August 2015 (UTC)

HEMCO/Extensions directory

The following modifications for MERRA2 were made in the HEMCO/Extensions directory:

File	Modifications made
hcox_ch4wetland_mod.F90	In routine WETLAND_EMIS: Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
hcox_dustdead_mod.F	In routine HCOX_DUSTDEAD_GETTUN: Add placeholder values of FLX_MSS_FDG_FCT = -999 for the MERRA2 2° x 2.5° and 4° x 5° grids. We will add the correct values once we can run simulations with at least a year of met data.
hcox_lightnox_mod.F	In routine GET_OTD_LIS_SCALE: Add placeholder values of ANN_AVG_FLASHRATE for the MERRA2 0.5° x 0.625° grids. Add placeholder values of BETA for the MERRA2 2° x 2.5° and 4° x 5° grids. We will add the correct values once we can run simulations with at least a year of met data.
hcox_megan_mod.F	In routine GET_GAMMA_SM: Changed code to: #if defined( GEOS_5 ) || defined( MERRA ) || defined( GEOS_FP ) || defined( MERRA2 )
hcox_tomas_dustdead_mod.F	In routine HCOX_TOMAS_DUSTDEAD_GETTUN: Add placeholder values of FLX_MSS_FDG_FCT = -999 for the MERRA2 2° x 2.5° and 4° x 5° grids. We will add the correct values once we can run simulations with at least a year of met data.

--Bob Y. (talk) 18:53, 12 August 2015 (UTC)