GEOS-Chem 12

From Geos-chem
Revision as of 18:41, 19 July 2018 by Bmy (Talk | contribs) (12.0.0)

Jump to: navigation, search


GEOS-Chem v11-02-final will also carry the designation GEOS-Chem 12.0.0. We are migrating to a purely numeric versioning system in order to adhere more closely to software development best practices. For a complete description of the new versioning system, please see our GEOS-Chem version numbering system wiki page.




Version history

The table below shows the previous, current, and successive major GEOS-Chem versions:

Previous major release Current release candidate Pending major release Next stable version Successive major release
GEOS-Chem v11-01-public v11-02-release-candidate
(aka v11-02-rc)
GEOS-Chem 12.0.0
aka v11-02-final in the
old version number system
GEOS-Chem 12.1.0 GEOS-Chem 13.0.0
RELEASED 01 Feb 2017 RELEASED 22 Jun 2018 RELEASE TBD RELEASE TBD RELEASE TBD 2019
GC v11-01 online manual GC v11-02 benchmark history GC 12 online manual GC 12 benchmark history


The sections below describe each GEOS-Chem feature version and GEOS-Chem patch version belonging to the GEOS-Chem 12 series. Each feature version carries the tag 12.Y.0, which indicates that the version will be (or has been) validated with a 1-month benchmark simulation (and one or more 1-year benchmark simulations if necessary). Each patch version carries the tag 12.Y.Z, which denotes that the version preserves backwards-compatibility with the prior feature version 12.Y.0. For more information, please see our GEOS-Chem version numbering system wiki page.

In the pipeline

Several GEOS-Chem updates have been proposed by the relevant Working Groups, but have not yet been assigned to a specific GEOS-Chem version. For a complete list of these updates, please visit our GEOS-Chem model development priorities wiki page.

12.3.0

The items listed below are slated to be included in GEOS-Chem 12.3.0. This is a feature release which will be validated with a 1-month benchmark simulation.

Feature Submitted by Type Status
Features affecting the full-chemistry simulation:
Update ISORROPIA from v2.0 to v2.2, in order to remove persistent noise in output Seb Eastham (Harvard) Science
  • Delivered to GCST (May 2017)
Update MODIS LAI fields for 2005-2016 using Yuan et al. (2011) Jenny Fisher (Wollongong)
Barron Henderson (US EPA)
Science
  • Will require reworking of LAI regridding in GEOS-Chem Classic and pre-processing land map data
Extend UCX into the mesosphere Sebastian Eastham (MIT) Science
OMI-based Volcanic emissions 2005-2012 Cui Ge (U. Iowa)
Jun Wang (U. Iowa)
Science
  • Working with developers to obtain alternate data for Oct-Dec 2012.

12.2.0

The items listed below are slated to be included in GEOS-Chem 12.2.0. This is a feature release which will be benchmarked with a 1-month benchmark simulation and a 1-year Rn-Pb-Be benchmark simulation.

Feature Submitted by Type Status
Features NOT affecting the full-chemistry simulation:
Common set of tracers in GEOS-Chem and GEOS Transport WG
GCST
Benchmarking
  • A set of passive species will be added to the GEOS-Chem Rn-Pb-Be benchmark simulation. This will facilitate comparisons between GEOS-Chem "Classic" and/or GCHP vs. the NASA GEOS-DAS system.

12.1.0

The items listed below are slated to be included in GEOS-Chem 12.1.0. This is a feature release which will be validated with a 1-month benchmark simulation.

Feature Submitted by Type Status
Features affecting the full-chemistry simulation:
FlexGrid: Use HEMCO to read met fields and enable nested simulations over any custom domains Jiawei Zhuang (Harvard)
Jintai Lin (PKU)
Structural
  • NOTE: FlexGrid might result in slight differences in met fields w/r/t the current regridding method. This could have an impact on the full-chemistry simulations.
Anthropogenic PM2.5 dust source Sajeev Philip (Dalhousie) Science
  • Update 29 Jun 2018: Fixed an initialization issue that caused much more dust to be emitted than the expected 13.1 Tg/yr.
  • Code updates have been added to GC git repository, but are commented out until we officially release this feature in 12.1.0.
NetCDF Output Phase 3b: Updates to eliminate differences between single and multi-segmented GEOS-Chem runs GCST
Christoph Keller
Bug Fix &
Structural
Features NOT affecting the full-chemistry simulation:
Updated flux diagnostics Chris Holmes (Florida State)
GCST
Structural/
Benchmarking

12.0.0



GEOS-Chem v11-02-final will also carry the designation GEOS-Chem 12.0.0. We are migrating to a purely numeric versioning system in order to adhere more closely to software development best practices. For a complete description of the new versioning system, please see our GEOS-Chem version numbering system wiki page.




The following updates and fixes are slated to be added to GEOS-Chem 12.0.0 (aka v11-02-final). Many of these updates fix issues that were identified during the GEOS-Chem v11-02 release candidate testing period.

Feature Submitted by Type Status
Features affecting the full-chemistry (benchmark) simulation:
Bug fix for CH4 boundary conditions Lee Murray (Rochester) Bug fix TBD
Fixed typo in call to GAMMA_HOBr_AER Xuan Wang
GCST
Bug fix Implemented 18 Jul 2018
  • NOTE: This update produced only very small differences w/r/t internal run v11-02-rc-DICEnewrc
    • Mean OH decreased by -0.058%
    • MCF lifetime increased by 0.082%
    • CH4 lifetime increased by 0.084%
    • O3 STE flux decreased by 0.005%
Correct issues when specifying DICE-Africa emissions: Eloise Marais (U. Birmingham)
GCST
Bug Fix Implemented 16 Jul 2018.
  • Results approved by Eloise Marais.
  • NOTE: This update produced only very small differences w/r/t internal run v11-02-rc-arctasship:
    • Mean OH increased by 0.07%
    • MCF lifetime decreased by 0.12%
    • CH4 lifetime decreased by 0.13%
    • O3 STE flux decreased by -0.0023%
Add missing Criegee intermediate reaction Xin Chen (UMN) Bug fix Implemented 11 Jul 2018, pending validation
  • NOTE: This update produced only very small differences w/r/t internal run v11-02-rc-arctasship:
    • Mean OH decreased by 0.00002%
    • MCF lifetime was unchanged
    • CH4 lifetime increased by 0.001%
    • O3 STE flux increased by 0.0001%
Bug fix for ARCTAS_SHIP in HEMCO_Config.rc files Barron Henderson (EPA) Bug Fix Implemented 03 Jul 2018
  • NOTE: This update only produced very small differences w/r/t internal run v11-02-rc-gfed2016:
    • Mean OH decreased by 0.0004%
    • O3 STE flux decreased by 0.0007%
    • MCF and CH4 lifetimes were unchanged
Avoid double counting CEDS and GEIA NH3 biofuel emissions Christoph Keller (NASA/GMAO)
GCST
Bug fix Implemented 29 Jun 2018
  • NOTE: The default HEMCO behavior is to prioritize CEDS over GEIA NH3 anthro and biofuel emissions. But to avoid confusion, we will explicitly request this behavior in the HEMCO_Config.rc files that ship with the GEOS-Chem Unit Tester.
Bug fix in FlexChem: do not zero out ACTA, EOH, HCOOH concentrations Katie Travis (MIT) Bug Fix Implemented 27 Jun 2018
  • NOTE: This update only produced very small differences w/r/t v11-02-rc:
    • Mean OH decreased by 0.083 %
    • MCF lifefime increased by 0.112 %
    • CH4 lifetime increased by 0.112 %
    • O3 STE flux decreased by 0.011 %
Add GFED4.1 data for 2015 and 2016; Use GFED4.1 2016 in benchmarks Kate O'Dell (CSU)
GCST
Science Data files downloaded 27 Jun 2018
Awaiting approval of results by developers
Features NOT affecting the full-chemistry simulation:
Update to HEMCO v2.1.007, which includes minor updates:
  1. Bug fix for distributing emissions in the vertical dimension
  2. Add extra error checks in HEMCO standalone module
  3. Remove null string character from netCDF unit string
  4. Bug fix for HEMCO soil NOx error with ifort 17
GCST
Christoph Keller
Jenny Fisher
Structural &
Bug fix
Update history:
  1. Implemented 19 Jul 2018
  2. Implemented 18 Jul 2018
  3. Implemented 17 Jul 2018
  4. Implemented 26 Jun 2018
Bug fix for reference time in netCDF diagnostic files GCST Bug fix Implemented 12 Jul 2018
Updates for QFED biomass burning emissions:
  1. Update to QFED v2.5r1
  2. Fix QFED files for 2016/10/13
Christoph Keller (NASA GMAO) Bug fix Update history:
  1. In progress
  2. Implemented 11 Jul 2018
Fix issues in complexSOA and complexSOA_SVPOA restart files Eloise Marais (U. Birmingham) Bug fix Implemented 05 Jul 2018
Bug fixes for v11-02 simulations using the RRTMG radiative transfer model Chris Holmes (U. Florida) Bug Fix Implemented 26 Jun 2018
  • NOTE: This only affects the RRTMG simulations. Full-chemistry difference-tests yielded identical results w/r/t v11-02-rc.
Now always declare the netCDF time variable with an unlimited dimension Chris Holmes (U. Florida) Structural Implemented 26 Jun 2018
  • NOTE: Full-chemistry difference-tests yielded identical results w/r/t v11-02-rc.

--Bob Yantosca (talk) 20:51, 18 July 2018 (UTC)

New data directories

The following HEMCO data directories have been added or updated for GEOS-Chem 12. You will have to download the directories relevant to your simulation.

Inventory What was added? When added? Directory
GFED4 biomass inventory GFED4.1 data files for 2015 and 2016 have been added 12.0.0 HEMCO/GFED4/v2015-10/2015
HEMCO/GFED4/v2015-10/2016

--Bob Yantosca (talk) 20:02, 29 June 2018 (UTC)

Validation

In this section we provide information about the benchmarks and tests that we have done to validate GEOS-Chem v11-03.

1-month and 1-year benchmarks

For complete information about the benchmark simulations used to validate GEOS-Chem v11-03, please see our GEOS-Chem 12 benchmark history wiki page.

Unit tests for GEOS-Chem 12

We perform a full set of unit tests on GEOS-Chem on an almost daily basis. These tests ensure that each new feature being added does not break any existing functionality.

Here are the unit tests that we ran before submitting 1-month benchmark simulation for v11-03. Click on the links below to see the results from each set of unit tests:

Unit test Date
TBD

Previous issues now resolved in GEOS-Chem 12

The following bugs and/or technical issues have been resolved in v11-03.

Outstanding issues not yet resolved in GEOS-Chem 12

The following known issues are slated to be fixed in GEOS-Chem 12:

Bug fix for HEMCO soil NOx error with ifort 17

This fix will be included in GEOS-Chem 12.0.0.

For complete information about this issue and how it was resolved, please see this post on our Implementation of HEMCO in GEOS-Chem wiki page..

--Bob Yantosca (talk) 13:54, 27 June 2018 (UTC)

FlexChem bug fix: do not zero ACTA, EOH, HCOOH

This fix will be included in GEOS-Chem 12.0.0.

For complete information about this bug and how it was fixed, please see this post on our FlexChem wiki page.

--Bob Yantosca (talk) 14:42, 27 June 2018 (UTC)

Avoid double counting CEDS and GEIA NH3 biofuel emissions

This fix will be included GEOS-Chem 12.0.0.

Eloise Marais wrote:

If CEDS includes NH3 emissions, then this will be double counting. I recommend using CEDS, as GEIA is very out of date (1998 according to the HEMCO configuration file I have).

Daniel Jacob replied:

We should use the GEIA NH3 natural emission inventory (1° x 1° file from Eleanor Morris) but not the GEIA NH3 biofuel inventory.

UPDATE 29 Jun 2018:

We discovered that HEMCO's default behavior is to ignore the GEIA NH3 anthro and biofuel emissions if CEDS is turned on. This is shown in the HEMCO.log file when VERBOSE = 3:

 Register_Base: Checking GEIA_NH3_ANTH
HEMCO: Entering Get_targetID (hco_config_mod.F90) ( 4)
 Skip container GEIA_NH3_ANTH because of CEDS_NH3_WST
. . .
 Register_Base: Checking GEIA_NH3_BIOF
HEMCO: Entering Get_targetID (hco_config_mod.F90) ( 4)
 Skip container GEIA_NH3_BIOF because of CEDS_NH3_WST_Cat00002

But for the sake of clarity, we will modify the HEMCO_Config.rc files to explicitly request this behavior:

(((.not.CEDS.or.EDGARv43

# GEIA NH3 anthro (skip if CEDS or EDGARv43 is on)
0 GEIA_NH3_ANTH $ROOT/NH3/v2014-07/NH3_geos.4x5.nc         NH3_ANTH     1990/1-12/1/0 C xy  kg/m2/s NH3 - 1 1 

# GEIA NH3 biofuel (skip if CEDS or EDGARv43 is on)
(((BIOFUEL
0 GEIA_NH3_BIOF $ROOT/NH3/v2014-07/NH3_biofuel.geos.1x1.nc NH3_BIOFUEL  1998/1-12/1/0 C xy  kg/m2/s NH3 - 2 1
)))BIOFUEL

))).not.CEDS.or.EDGARv43

--Bob Yantosca (talk) 21:02, 29 June 2018 (UTC)

Fix bug in ARCTAS_SHIP_SO2 scale factors

This fix will be included GEOS-Chem 12.0.0.

For a complete description of this issue and how it was resolved, please see this post on our Ship Emissions wiki page.

--Bob Yantosca (talk) 15:54, 3 July 2018 (UTC)

Fix issues in complexSOA and complexSOA_SVPOA restart files

This fix will be included GEOS-Chem 12.0.0.

Eloise Marais wrote:

I just wanted to draw your attention to a few issues with the restart files provided on the ftp site (http://ftp.as.harvard.edu/gcgrid/data/ExtData/SPC_RESTARTS/). The complexSOA and complexSOA_SVPOA restart files (4° x 5° and 2° x 2.5°) both have a species entry for IEPOX and also IEPOXD, IEPOXB, and IEPOXA, but should only have the latter 3. This is the same for RIP. The number of species in the 4° x 5° and 2° x 2.5° files for complexSOA is also different.

To fix these issues, we have generated new restart files for the complexSOA and complexSOA_SVPOA simulations. You can download these to your SPC_RESTARTS folder with the following commands:

 wget "ftp://ftp.as.harvard.edu/gcgrid/data/ExtData/SPC_RESTARTS/initial_GEOSChem_rst.2x25_complexSOA.nc"
 wget "ftp://ftp.as.harvard.edu/gcgrid/data/ExtData/SPC_RESTARTS/initial_GEOSChem_rst.2x25_complexSOA_SVPOA.nc"
 wget "ftp://ftp.as.harvard.edu/gcgrid/data/ExtData/SPC_RESTARTS/initial_GEOSChem_rst.4x5_complexSOA.nc"
 wget "ftp://ftp.as.harvard.edu/gcgrid/data/ExtData/SPC_RESTARTS/initial_GEOSChem_rst.4x5_complexSOA_SVPOA.nc"

--Bob Yantosca (talk) 18:01, 5 July 2018 (UTC)

Bug fix for reference time in netCDF diagnostic files

This fix will be included GEOS-Chem 12.0.0.

Maria Zatko wrote:

For the sake of run time, we would like to run GEOS-Chem with the old (15/30 minute timesteps) instead of the recommended timesteps (10/20 minutes). I noticed that the output netcdf files have different timestamps depending on the timestep (e.g., 10/20 minutes, 15/30 minutes) used. For example, when 10/20 minute timesteps are used, the time units in the following netcdf output files range from 0 to 23 UTC:
     GEOSChem3hrInst.SpeciesConcWithMet.20150701_0000z.nc4 (file size: 4.6 GB)
     GEOSChem3hrInst.SpeciesConcWithMet.20150702_0000z.nc4 (file size: 5.3 GB)
     time:units = "minutes since 2015-07-01 00:00:00 UTC"
When 15/30 minute timesteps are used, the time units in the following output files include hour 24:
     GEOSChem3hrInst.SpeciesConcWithMet.20150701_0000z.nc4 (file size: 4.6 GB)
     GEOSChem3hrInst.SpeciesConcWithMet.20150701_2400z.nc4 (file size: 5.3 GB)
     GEOSChem3hrInst.SpeciesConcWithMet.20150702_2400z.nc4 (file size: 5.3 GB)
     time:units = "minutes since 2015-07-01 24:00:00 UTC"

I wanted to check in to see if others have run into this issue as well.

Bob Yantosca replied:

This issue is caused by floating-point roundoff error in converting from the Julian date to the calendar date. I have created a more robust fix for this issue, which now computes the correct netCDF file reference time independent of the model "heartbeat" timestep.

To apply the fix:

  1. Copy the patch file ftp://ftp.as.harvard.edu/pub/exchange/bmy/nc_ref_time_fix.txt to your computational server.
  2. Change to your GEOS-Chem v11-02 release candidate code directory and check out the "master" branch.
  3. Type: git am < nc_ref_time_fix.txt
The fix will be standardized in the GEOS-Chem 12.0.0 release.

--Bob Yantosca (talk) 15:10, 12 July 2018 (UTC)

Bug fix for CH4 boundary conditions

This fix will be included GEOS-Chem 12.0.0.

Lee Murray wrote:

I’ve been testing the release candidate, and believe that we need to change the default input.geos options for the Standard and Benchmark simulations in the unit tester to :
    %%% EMISSIONS MENU %%%  : 
    Turn on emissions?      : T  
    HEMCO Input file        : HEMCO_Config.rc
     => 1ppt MBL BRO Sim.?  : F
    Switches for UCX        :---
     => Use CH4 emissions?  : F
     => Turn on surface BCs :---
        => CH4?             : F
        => OCS?             : T
        => CFCs?            : T
        => Cl species?      : T
        => Br species?      : F
        => N2O?             : T
     => Set initial glob MRs:---
        => strat. H2O?      : T
     => CFC emission year   : 0
By default in the unit tester, => CH4 is still enabled. This appears to set up conflict between the older UCX surface layer and the newer HEMCO methane PBL boundary conditions, causing air to be a mix of preindustrial and present-day values (see below for Jul 1, 2013 surface methane in a benchmark run). Setting it to F and/or removing LCH4SBC causes the methane to behave as expected, i.e., being set by HEMCO in all simulations, with the expected spatial pattern. Also, we can slate Online CH4 chemistry?  : T in the Chemistry Menu and Input_Opt%LCH4CHEM for future removal, as it no longer does anything.

--Melissa Sulprizio (talk) 18:16, 16 July 2018 (UTC)

Fixed typo in call to GAMMA_HOBr_AER

This fix will be included GEOS-Chem 12.0.0.

Xuan Wang wrote:

When looking back at the standard GC codes of v11-02d, I found a bug in the function HETHOBr_SS_JS (in module KPP/Standard/gckpp_HetRates.F90). When calling the gamma calculation, in v11-02d, around line 3514, we have:
      CALL Gamma_HOBr_AER(rAer, denAir, 2, TK, clConc, brConc, &
                          hConc, GAM_HOBr, r_gp)
This will cause both HOBr+Cl- (X=1) and HOBr+Br- (X=2) on sea salt use the gamma of HOBr + Br- (X=2). It should be:
      CALL Gamma_HOBr_AER(rAer, denAir, X, TK, clConc, brConc, &
                          hConc, GAM_HOBr, r_gp)
It looks like a typo which could lead to quite different result for halogen chemistry.

--Bob Yantosca (talk) 20:26, 17 July 2018 (UTC)