Skip to content

README.MetForcing_and_BCS.md

Latest commit

 

History

History
506 lines (383 loc) · 19.9 KB

README.MetForcing_and_BCS.md

File metadata and controls

506 lines (383 loc) · 19.9 KB

README.metforcing_and_bcs

Description:

Information about surface meteorological forcing and boundary conditions for GEOSldas.

Author:

  • reichle (5 Oct 2017) - first version; updated regularly
  • jperket (10 Dec 2019) - converted to markdown

Surface Meteorological Forcing

The forcing time step is controlled with the configurable resource variable FORCE_DTSTEP and must match the frequency of the input forcing files. Specify FORCE_DTSTEP=3600 [seconds] for most GEOS products, incl. MERRA, MERRA-2, FP, FP-IT/RP-IT, and GEOS-IT, which are 1-hourly datasets.

The spatial (horizontal) interpolation method for the met forcing is controlled by MET_HINTERP (see optional parameters in exeinp input file to ldas_setup).

MET_PATH and MET_TAG must be consistent:

  • MET_PATH is the full path to the forcing data set.

  • MET_TAG is an identifier for the forcing data set.

  • Available non-GEOS forcing data sets are pre-defined in subroutine get_forcing().

  • For GEOS forcing datasets, see special MET_TAG parsing conventions in subroutines parse_MERRA_met_tag(), parse_MERRA2_met_tag() and parse_G5DAS_met_tag().

  • For details on corrected precipitation data see also (https://gmao.gsfc.nasa.gov/pubs/):

    • Reichle and Liu (2014), Observation-Corrected Precipitation Estimates in GEOS-5, NASA Technical Report Series on Global Modeling and Data Assimilation, NASA/TM-2014-104606, Vol. 35, National Aeronautics and Space Administration, Goddard Space Flight Center, Greenbelt, Maryland, USA, 18pp.
    • Reichle, R. H., and Q. Liu (2021), Observation-Corrected Precipitation for the SMAP Level 4 Soil Moisture (Version 6) Product and the GEOS R21C Reanalysis, NASA Technical Report Series on Global Modeling and Data Assimilation, NASA/TM-2021-104606, Vol. 59, National Aeronautics and Space Administration, Goddard Space Flight Center, Greenbelt, Maryland, USA, 28pp.

COMMONLY USED values for MET_PATH:

Legacy datasets

  MET_PATH : [XXX]/l_data/ECMWF/GRID/CORRECTED/netcdf/
  MET_PATH : [XXX]/l_data/GLDAS/netcdf/
  MET_PATH : [XXX]/l_data/GSWP-2/Baseline_Forcings/
  MET_PATH : [XXX]/l_data/RedArk/RedArk_subbasin_forcing/red_ark_forc/

ERA5 (LDAS-Monde via NASA LIS group)

  MET_PATH : /discover/nobackup/projects/lis/MET_FORCING/ERA5/

GEOS datasets

MERRA forcing (including precip-corrected MERRA forcing)

  MET_PATH : /discover/nobackup/projects/gmao/merra/iau/merra_land/MERRA_land_forcing/

MERRA2 forcing (including precip-corrected MERRA2 forcing)

  MET_PATH : /discover/nobackup/projects/gmao/merra/iau/merra_land/MERRA2_land_forcing/

MERRA-21C forcing

  MET_PATH : [TBD]

Note: Used for SMAP Nature Run v05, v7.2, v8.1, v8.3, v9.1, v10.0 (before 1/1/2015).

GEOS FP forcing with "seamless" file names, for use with MET_TAG=GEOS.fp.asm[__prec*] (PREFERRED)

  MET_PATH : /discover/nobackup/projects/gmao/smap/SMAP_L4/GEOS/FP/

Note: Used for SMAP Nature Run v8.1, v8.3, v9.1, v10.0 (after 1/1/2015); SMAP L4_SM Version 5, 6, 7.

GEOS forcing with experiment-specific file names, incl. FP (DEPRECATED), FP-IT/RP-IT, and precip-corrected GEOS forcing

  MET_PATH : /discover/nobackup/projects/gmao/merra/iau/merra_land/GEOS5_land_forcing/

Note: Used for SMAP Nature Run v03, v04, v04.1, v05, v7.2 (after 1/1/2015); SMAP L4_SM Version 4.

GEOS-IT forcing (including precip-corrected GEOS-IT forcing)

  MET_PATH : /discover/nobackup/projects/gmao/merra/iau/merra_land/GEOSIT_land_forcing/

Forcing from post-processed output of the GEOS S2S system (FCST, AODAS)

  MET_PATH : [check with GMAO S2S group]

COMMONLY USED values for MET_TAG:

Legacy datasets

  MET_TAG  : Berg_netcdf
  MET_TAG  : GLDAS_2x2_5_netcdf
  MET_TAG  : GSWP2_1x1_netcdf
  MET_TAG  : RedArk_ASCII

ERA5 (LDAS-Monde via NASA LIS group)

  MET_TAG  : ERA5_LIS

GEOS datasets

MERRA

  MET_TAG  : d5_merra_cross__GEOSdas-2_1_4

MERRA-Land

  MET_TAG  : d5_merra_cross__GEOSdas-2_1_4__precCPCU

MERRA-2

  • All MERRA-2 options use native MERRA-2 "lfo" files for all surface met forcing fields except precipitation.

  • Option 1a: Use corrected precip seen by the land w/in the MERRA-2 system (i.e., native MERRA-2 "lfo" files). Precip is as corrected within the MERRA-2 system. Closest to land-only MERRA-2 replay.

    MET_TAG  : M2COR_cross                                       ! all streams
MET_TAG  : M2COR_100                                         ! stream 1 only
MET_TAG  : M2COR_200                                         ! stream 2 only
MET_TAG  : M2COR_300                                         ! stream 3 only
MET_TAG  : M2COR_400                                         ! stream 4 only

  • Option 1b: Use corrected precip forcing constructed in post-processing using MERRA-2 as background. Background precip is typically from MERRA-2 "int" data, but corrected precip is stored in files that look like MERRA-2 "lfo" files.

    For example, select corrected precip version "CPCUGPCP22clim_MERRA2_BMTXS" as follows:

    MET_TAG  : M2COR_cross__precCPCUGPCP22clim_MERRA2_BMTXS      ! all streams
MET_TAG  : M2COR_100__precCPCUGPCP22clim_MERRA2_BMTXS        ! stream 1 only
MET_TAG  : M2COR_200__precCPCUGPCP22clim_MERRA2_BMTXS        ! stream 2 only
MET_TAG  : M2COR_300__precCPCUGPCP22clim_MERRA2_BMTXS        ! stream 3 only
MET_TAG  : M2COR_400__precCPCUGPCP22clim_MERRA2_BMTXS        ! stream 4 only

    This particular version uses as background the MERRA-2 model ("int") precip, rescaled to the GPCPv2.2 climatology (indicated by "S" in "BMTXS"). Outside of Africa and the high latitudes (tapered between 42.5 and 62.5 deg lat), this background precipitation is corrected with CPCU data (also rescaled to the GPCPv2.2 climatology). See Reichle and Liu (2014) GMAO Tech Memo #35 for more information on methods "BMTX".

  • Option 2: Use uncorrected precip generated by the AGCM w/in the MERRA-2 system (i.e., native MERRA-2 "int" files).

    MET_TAG  : M2INT_cross                                       ! all streams
MET_TAG  : M2INT_100                                         ! stream 1 only
MET_TAG  : M2INT_200                                         ! stream 2 only
MET_TAG  : M2INT_300                                         ! stream 3 only
MET_TAG  : M2INT_400                                         ! stream 4 only

MERRA-21C (M21C)

  • IMPORTANT: M21C forcing option is experimental. As of Jan 2025, M21C is in production. Publication is planned for 2026/2027.

  • All M21C options use native M21C "lfo" files for all surface met forcing fields except precipitation.

  • Option 1a: Use corrected precip seen by the land w/in the M21C system (i.e., native M21C "lfo" files). Precip is as corrected within the M21C system. Closest to land-only M21C replay.

    Use 1152-by-721 lat/lon gridded forcing:

    MET_TAG  : M21CCOR_cross                                     ! all streams
MET_TAG  : M21CCOR_jan98                                     ! stream 1 only
MET_TAG  : M21CCOR_jan08                                     ! stream 2 only
MET_TAG  : M21CCOR_jan18                                     ! stream 3 only

    Use c360 cube-sphere gridded forcing (works only if using matching c360 tile space for simulation):

    MET_TAG  : M21CCSCOR_cross                                   ! all streams
MET_TAG  : M21CCSCOR_jan98                                   ! stream 1 only
MET_TAG  : M21CCSCOR_jan08                                   ! stream 2 only
MET_TAG  : M21CCSCOR_jan18                                   ! stream 3 only

  • Option 1b: Use corrected precip forcing [xxx] constructed in post-processing using M21C as background. Background precip is typically from M21C "int" data, but corrected precip is stored in files that look like M21C "lfo" files.

    Use 1152-by-721 lat/lon gridded forcing:

    MET_TAG  : M21CCOR_cross__prec[xxx]                          ! all streams
MET_TAG  : M21CCOR_jan98__prec[xxx]                          ! stream 1 only
MET_TAG  : M21CCOR_jan08__prec[xxx]                          ! stream 2 only
MET_TAG  : M21CCOR_jan18__prec[xxx]                          ! stream 3 only

    Use c360 cube-sphere gridded forcing (works only if using matching c360 tile space for simulation):

    MET_TAG  : M21CCSCOR_cross__prec[xxx]                        ! all streams
MET_TAG  : M21CCSCOR_jan98__prec[xxx]                        ! stream 1 only
MET_TAG  : M21CCSCOR_jan08__prec[xxx]                        ! stream 2 only
MET_TAG  : M21CCSCOR_jan18__prec[xxx]                        ! stream 3 only

  • Option 2: Use uncorrected precip generated by the AGCM w/in the M21C system (i.e., native M21C "int" files).

    Use 1152-by-721 lat/lon gridded forcing:

    MET_TAG  : M21CINT_cross                                     ! all streams
MET_TAG  : M21CINT_jan98                                     ! stream 1 only
MET_TAG  : M21CINT_jan08                                     ! stream 2 only
MET_TAG  : M21CINT_jan18                                     ! stream 3 only

    Use c360 cube-sphere gridded forcing (works only if using matching c360 tile space for simulation):

    MET_TAG  : M21CCSINT_cross                                   ! all streams
MET_TAG  : M21CCSINT_jan98                                   ! stream 1 only
MET_TAG  : M21CCSINT_jan08                                   ! stream 2 only
MET_TAG  : M21CCSINT_jan18                                   ! stream 3 only

RP-IT/FP-IT (d591)

MET_TAG  : cross_d591_RPFPIT                                  ! all streams
MET_TAG  : d591_rpit1_jan00                                   ! stream 1 only
MET_TAG  : d591_rpit2_jun06                                   ! stream 2 only
MET_TAG  : d591_rpit3_jan11                                   ! stream 3 only
MET_TAG  : d591_fpit                                          ! stream 4 only

RP-IT/FP-IT (d5124)

MET_TAG  : cross_d5124_RPFPIT                                 ! all streams   - uses "late-look" through present
MET_TAG  : d5124_rpit1_jan00                                  ! stream 1 only
MET_TAG  : d5124_rpit2_jun04                                  ! stream 2 only
MET_TAG  : d5124_rpit3_jan12                                  ! stream 3 only - updated through present

GEOS FP

MET_TAG  : GEOS.fp.asm   ! PREFERRED:  "seamless" FP files (published/generic file names, ~same result as cross_FP)

MET_TAG  : cross_FP      ! DEPRECATED: stitch FP experiment names across years
MET_TAG  : e5110_fp      ! starting 11 Jun 2013
MET_TAG  : e5130_fp      ! starting 20 Aug 2014
MET_TAG  : e5131_fp      ! starting  1 May 2015
MET_TAG  : f516_fp       ! starting 24 Jan 2017
MET_TAG  : f517_fp       ! starting  1 Nov 2017
MET_TAG  : f521_fp       ! starting 11 Jul 2018
MET_TAG  : f522_fp       ! starting 13 Mar 2019
MET_TAG  : f525_fp       ! starting 30 Jan 2020
MET_TAG  : f525_p5_fp    ! starting  7 Apr 2020
...

GEOS-IT

MET_TAG  : cross_d5294_GEOSIT

Forcing from post-processed output of the GEOS S2S system

  • Forcing derived through post-processing of daily average output from the GEOS S2S system, including S2S hindcasts/forecasts ("FCST") and the "AODAS" used for S2S initialization.

    S2S output is from the geosgcm_vis2d and geosgcm_surf Collections for FCST and from the geosgcm_rad and geosgcm_surf Collections for AODAS (see GMAO Office Note No. 16).

    For FCST, post-processing includes a monthly bias correction to the MERRA-2 climatology.

    Daily data are disaggregated to 6-hourly (FCST) or 1-hourly (AODAS) using the MERRA-2 climatological diurnal cycle.

    For FCST, MET_TAG must specify S2S ensemble member ('ensX'; currently: 'ens1', 'ens2', 'ens3', or 'ens4') and month/day of forecast initialization ('MMMDD'; e.g., 'jan01'), separated by double underscores.

    As of 14 Jun 2021:

    • Preparation of S2S forcing data ignores the 3-hour offset between S2S daily averages (21z-21z) and the MERRA-2 daily averages (0z-0z) used for the temporal disaggregration.
    • The processing of the S2S output incorrectly partitioned total precipitation into snowfall and convective precipitation. Therefore, rainfall and snowfall are determined in the S2S forcing reader from total precipitation and air temperature. Convective rainfall is set to 0. (As of now, only total rainfall is used by Catchment.)
    MET_TAG  : GEOSs2sFCST__[ensX]__[MMMDD]
MET_TAG  : GEOSs2sAODAS

SMAP L4_SM

Pre-beta SMAP L4_SM

  MET_TAG  : cross_FP__precCPCUG5FPv2

SMAP Nature Run v03

  MET_TAG  : cross_RPFPIT__precCPCUG5RPFPITv1                ! before 1/1/2014
  MET_TAG  : cross_FP__precCPCUG5FPv1                        ! after  1/1/2014

SMAP Nature Run v04

  MET_TAG  : cross_d591_RPFPIT__precCPCUG5RPFPITv2           ! before 1/1/2014
  MET_TAG  : cross_FP__precCPCUG5FPv2                        ! after  1/1/2014

SMAP Nature Run v04.1

  MET_TAG  : cross_d5124_RPFPIT__precCPCUG5RPFPITv2.1        ! before 1/1/2015
  MET_TAG  : cross_FP__precCPCUG5FPv2                        ! after  1/1/2015

SMAP Nature Run v05, v7.2, v8.1, v8.3; SMAP L4_SM Version 4

  MET_TAG  : M2COR_cross__precCPCUGPCP22clim_MERRA2_BMTXS    ! before 1/1/2015
  MET_TAG  : cross_FP__precCPCUG5FPv3                        ! after  1/1/2015

SMAP L4_SM Version 5

  MET_TAG  : GEOS.fp.asm__precCPCULLKG5FPv3                  ! (precip corr with late-look CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLKG5FPv3                  ! (precip corr with first-look CPCU)

SMAP Nature Run v9.1; SMAP L4_SM Version 6

  MET_TAG  : M2COR_cross__precSMAPv6_CGIM2                   ! before 01/01/2015
  MET_TAG  : GEOS.fp.asm__precCPCU_IMGFinal_IMGFclim_G5FP    ! before 06/30/2021 (precip corr with IMERG-Final and late-look  CPCU) 
  MET_TAG  : GEOS.fp.asm__precCPCULLK_IMERGLateV06b_fp_v1    ! before 10/27/2021 (precip corr with IMERG-Late  and late-look  CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06b_fp_v1    ! before 05/09/2022 (precip corr with IMERG-Late  and first-look CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06c_fp_v1    !                   (precip corr with IMERG-Late  and first-look CPCU)

SMAP Nature Run v10.0*; SMAP L4_SM Version 7

  MET_TAG  : M2COR_cross__precSMAPv6_CGIM2                   ! before 01/01/2015 
  MET_TAG  : GEOS.fp.asm__precCPCU_IMGFinal_IMGFclim_G5FP    ! before 09/30/2021 (precip corr with IMERG-Final and late-look  CPCU) 
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06b_fp_v1    ! before 05/09/2022 (precip corr with IMERG-Late  and first-look CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06c_fp_v1    ! before 07/02/2023 (precip corr with IMERG-Late  and first-look CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06d_fp_v1    ! before 11/07/2023 (precip corr with IMERG-Late  and first-look CPCU)
  MET_TAG  : GEOS.fp.asm__precCPCUFLK_IMERGLateV06e_fp_v1    !                   (precip corr with IMERG-Late  and first-look CPCU)

*Transitions dates are for L4_SM Version 7. Transition dates for NRv10.0 may differ somewhat.

Boundary Conditions

Boundary conditions (bcs) are tile-space model parameters that are provided in a set of files located in BCS_PATH for a given BCS_RESOLUTION.

For "land" tiles, the discretization (tile-space) is constructed in one of two different ways:

  1. By intersecting watersheds with a regular grid (typically that used in the atmospheric model), e.g., DC0576xPC0361, where "DC" indicates that the grid cells are centered on the date line and "PC" indicates that the grid cells are centered on the poles. Another example is the 0.5-degree ("c180") cube-sphere grid used by the atmospheric model in the MERRA-2 reanalysis: CF0180x6C_DE1440xPE0720.

  2. Directly on a regular grid, e.g., EASEv2_M09.

IMPORTANT: Beginning with GEOSldas release v18.0.0, bcs must be provided in a revised directory layout and naming convention. On NCCS/Discover, use:

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/fvInput/ExtData/esm/tiles/

COMMONLY USED boundary conditions (bcs):

MERRA-Land (and MERRA)

  BCS_PATH : /discover/nobackup/projects/gmao/ssd/land/l_data/geos5/bcs/SiB2_V2_bad_lon_onDL/DC/

Ganymed-4_0 (GM4): MERRA2

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Ganymed-4_0/Ganymed-4_0_MERRA-2/

Note: The same land bcs (for cube-shere only) are also in /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Icarus/Icarus_MERRA-2/

SMAP Nature Run v03

  BCS_PATH : /discover/nobackup/projects/gmao/ssd/land/l_data/geos5/bcs/CLSM_params/mkCatchParam_v15/

SMAP Nature Run v04, v04.1

  BCS_PATH : /discover/nobackup/projects/gmao/ssd/land/l_data/geos5/bcs/CLSM_params/mkCatchParam_SMAP_L4SM_v001/

SMAP Nature Run v05

  BCS_PATH : /discover/nobackup/projects/gmao/ssd/land/l_data/geos5/bcs/CLSM_params/mkCatchParam_SMAP_L4SM_v002/

Icarus-NL ("New Land"): SMAP Nature Run v7.2

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Icarus-NL/Icarus-NL_[XXXX]/

Notes:

  • DO NOT USE unless replicating previous experiments. There is "missing" data in green*.data, nirdf*.dat, and visdf*.dat files.
  • This path remains in place to permit recreating experiments that have used this path.
  • The sub-directory "Icarus-NL_MERRA-2/" contains the "new land" bcs. The string "MERRA-2" in this sub-directory name refers to ocean bcs that are not relevant for GEOSldas.

Icarus-NLv2: SMAP L4_SM Version 4

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Icarus-NLv2/Icarus-NLv2_[XXXX]/

Notes:

  • Icarus-NLv2 is a update to Icarus-NL bcs. A patch has been applied to files green*.data, nirdf*.dat, and visdf*.dat.

Icarus-NLv3 (NL3): SMAP Nature Run v8.1, GEOS-FP 5.25, 5.27, 5.29, GEOS-IT

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/legacy_bcs/Icarus-NLv3/Icarus-NLv3_[XXXX]/

Notes:

  • Soil parameters for a small fraction (< 0.05%) of tiles changed to correct "Mali" bug.

  • Vegdyn.data now netcdf4; reverts to using Dorman/Sellers veg heights (abandons JPL/Simard et al. 2011 Lidar data).

  • Some underlying ASCII data files are now grouped into a netcdf4 file. I.e., data in ar.new, bf.dat, ts.dat, etc are now in:

    • clsm/catch_params.nc4
    • clsm/catchcn_params.nc4 (additional parameters for CatchCN)

  • Generated with cvs tag Jason-3_0_LANDBCS

Icarus-NLv4 (NL4): SMAP Nature Run v8.3, v9.1, SMAP L4_SM Version 5, 6

  BCS_PATH : /discover/nobackup/projects/gmao/smap/SMAP_L4/L4_SM/bcs/CLSM_params/Icarus-NLv4_EASE/

Notes:

  • Icarus-NLv4 is identical to Icarus-NLv3 except that NLv4 reinstates veg heights from JPL/Simard et al. 2011 Lidar data.
  • Generated with GEOSldas tag v17.9.0-beta.7 under SLES11 O/S.

Icarus-NLv5 (NL5): SMAP Nature Run v10.0, SMAP L4_SM Version 7

  BCS_PATH : /discover/nobackup/projects/gmao/smap/SMAP_L4/L4_SM/bcs/CLSM_params/Icarus-NLv5_EASE/

Notes:

  • Icarus-NLv5 is identical to Icarus-NLv4 except for parameters associated with PEATCLSM.

v12: SMAP Nature Run v11.3, SMAP L4_SM Version 8

  BCS_PATH : /discover/nobackup/projects/gmao/bcs_shared/fvInput/ExtData/esm/tiles/v12/

Notes:

  • v12 is identical to Icarus-NLv5 (within roundoff) except for new MODIS-based snow albedo (v2) parameters and the Argentina peatland fix.