Difference between revisions of "GEOS-Chem 13.4.0"

From Geos-chem
Jump to: navigation, search
(Benchmark assessment form)
(Benchmark assessment form)
Line 424: Line 424:
 
     AEIC RCHO          :          0.001384            0.003358      0.001974  142.621
 
     AEIC RCHO          :          0.001384            0.003358      0.001974  142.621
  
<blockquote>However, this update resulted in minimal changes to global quantities, as VOC aircraft emissions are typically less than 1Gg for the simulatio month.  Observed changes from this update were:
+
<blockquote>However, this update resulted in minimal changes to global quantities, as VOC aircraft emissions are typically less than 1Gg for the simulation month.  Observed changes from this update were:
 
* Very small changes in oxidants (too small to show up in ratio plots).   
 
* Very small changes in oxidants (too small to show up in ratio plots).   
 
* Slight change in most primary organics, but % differences are of order 1e-4 or 1e-5.  
 
* Slight change in most primary organics, but % differences are of order 1e-4 or 1e-5.  

Revision as of 21:11, 25 January 2022

GEOS-Chem Versions Page

This page includes information about GEOS-Chem 13.4.0 (currently under development)

Version overview

The following items have been added to GEOS-Chem 13.4.0 since the last GEOS-Chem version GEOS-Chem 13.3.4. Also see the following Github pages:

Feature Contributor(s) Model scope Notes and references
Updates that will affect full-chemistry benchmark simulations
Migrating sulfate chemistry into KPP Mike Long (GCST)
Viral Shah (Harvard)
Bob Yantosca (GCST)
Chemistry &
Structural
Fix bug in Henry law CALC_HEFF routine in hetchem computations Viral Shah (Harvard) Chemistry
Bug fix: Remove double-counting of HCl uptake on acidic sea salt Viral Shah (Harvard) Chemistry
Update surface resistance for ozone deposition to ice/snow Ryan Pound (York)
Shuting Zhai (UW)
Drydep
Use updated definition of PM2.5; Add PM10 diagnostic Bob Yantosca (GCST) Diagnostics &
Benchmarking
Fix scale factors applied to VOC aircraft emissions in AEIC inventory Flavio Quadros (TU Delft)
Seb Eastham (GCST)
Emissions &
Bug Fix
HEMCO bug fix: restore archiving via manual diagnostics (which had been inadvertently disabled in a prior update) Melissa Sulprizio (GCST)
Bob Yantosca (GCST)
Emissions
Use degassing-only volcano emissions in GEOS-Chem benchmark simulations Christoph Keller (GMAO)
Daniel Jacob (Harvard)
Emissions &
Benchmarking
Add option for driving GEOS-Chem with native GEOS meteorological files Liam Bindle (GCST) Structural
Updates that will NOT affect full-chemistry benchmark simulations
Updated Hg chemistry in KPP Viral Shah (Harvard)
Colin Thackray (Harvard)
Mike Long (GCST)
Bob Yantosca (GCST)
Chemistry
Bug fix: Use REAL*8 precision in for CH3O2 + OH reaction definition in fullchem.eqn Bob Yantosca (GCST) Chemistry
Updated Global Inventory of Methane Emissions from Oil/Gas/Coal (GFEI v2) Tia Scarpelli (U. Edinburgh) Emissions
Anthropogenic Methane Emission Inventory for Canada Tia Scarpelli (U. Edinburgh) Emissions
Use EDGARv6 global anthro emissions for CH4 simulation Melissa Sulprizio (GCST) Emissions
Updated Rn-222 emissions for TransportTracers simulations Bo Zhang (NIA) Emissions
Retire HEMCO extension for CH4 wetland emissions Melissa Sulprizio (GCST) Emissions
Updates to enable automated run directory creation Liam Bindle (GCST)
Melissa Sulprizio (GCST)
Run directory
Bug fix: Allow GEOS-Chem Classic History to archive irregular intervals (e.g. a month and a day) Bob Yantosca (GCST) Diagnostics
Updates for running GEOS-Chem within GEOS Christoph Keller (GMAO)
Lizzie Lundgren (GCST)
GEOS
Updates that will only affect GCHP
Add capability to use mass fluxes in advection Sebastian Eastham (MIT)
Liam Bindle (GCST)
Advection
Define a GCHP "Default" diagnostic collection that will archive a small number of commonly-requested diagnostic fields Liam Bindle (GCST) Structural
Add capability to read native meteorological data files Liam Bindle (GCST)
Bob Yantosca (GCST)
Structural

New data directories

The following data directories have been added or updated in this version. You will have to download the directories relevant to your simulation.

Inventory Needed for What was added? Directory
ZHANG_Rn222 TransportTracers simulations only Updated Rn222 emissions for TransportTracers simulations. See Bo Zhang et al (2021) HEMCO/ZHANG_Rn222/v2021-11
Updated Hg simulation restart file Hg simulations only Sample restart file for the updated Hg mechanism. See Viral Shah et al (2021) ExtData/GEOSCHEM_RESTARTS/v2021-12
Updated TOMAS simulation restart files TOMAS simulations only Added HSO3m and SO3mm species to TOMAS restart files ExtData/GEOSCHEM_RESTARTS/v2021-12

1-month benchmarks

Model setup for 1-month benchmarks

The following table describes the GEOS-Chem setup for both GEOS-Chem Classic and GCHP 1-month benchmark simulations.

Git tag(s):
Compared to previous benchmark: 13.3.0
Software environment: gcc 10.2.0
OpenMPI 4.1.0
netcdf-fortran 4.5.3
CMake 3.17.3
ESMF 8.0.1
Model resolution:
  • GEOS-Chem Classic: 4° x 5°
  • GCHP: c48
Meteorology fields: MERRA-2
  • GEOS-Chem Classic: 4° x 5°
  • GCHP: 0.5° x 0.625°
Initial conditions: July 1, 2019
  • NOTE: GCHP c48 restart file regridded from GEOS-Chem Classic 4° x 5°
Benchmark month: July 2019
Emissions setup: TBD
Chemistry mechanism: TBD
This version will impact:
(select all that apply with boldface)
  • Advection
  • BL Mixing
  • Convection
  • Meteorology Fields
  • Dry Deposition
  • Wet Deposition
  • Emissions
  • Photolysis
  • Chemistry
  • Other (please specify):
    • Diagnostics
    • CH4 simulation
    • TransportTracers simulation
    • Run directory creation scripts

1-month fullchem: GEOS-Chem Classic 13.4.0-rc.0 vs 13.3.0

NOTES:

  • Enable PDF bookmarks in the file links to navigate between species. If you do not see bookmarks in your browser try using a different browser (not Safari) or installing an Adobe Acrobat plug-in.
  • See the species database for species definitions and details.
  • Species categories and lumped species are defined in YAML files within the GEOS-Chem post-processing software GCPy.

TBD

Benchmark assessment form

Performance GEOS-Chem Classic (rc.0) GCHP (rc.0)
Grid resolution: 4° x 5° C48
# CPUs, nodes: 48 CPUs
1 node (Intel Cascade Lake)
48 CPUs
2 nodes (Intel Cascade Lake)
Wall time: 03:37:13 06:20:26
CPU time / wall time: 31.79 95.0732
% of ideal performance: 66.23% 99.03%
Memory: 11.4128 GB 108.5 GB
Global mean OH:
(% change from previous version)
11.70790814279 x 105 molec/cm3
(-0.329%)
TBD
Methyl chloroform lifetime:
(% change from previous version)
5.294530 years
(+0.557%)
TBD
Methane lifetime:
(% change from previous version)
8.923431 years
(+0.581%)
TBD
Details
Species, J-Values, and COLUMN AOD values that changed by 10% or more at the surface: GEOS-Chem Classic
13.4.0-rc.1 vs. 13.3.0-rc.0
TBD
Species and J-Values that changed by 10% or more at 500 hPa: GEOS-Chem Classic
13.4.0-rc.1 vs. 13.3.0-rc.0
TBD
Species and J-Values that changed by 10% or more in zonal mean: GEOS-Chem Classic
13.4.0-rc.1 vs. 13.3.0-rc.0
TBD
Emissions that changed by 10% or more: GEOS-Chem Classic
13.4.0-rc.1 vs. 13.3.0-rc.0
TBD
Overview
Below we summarize the notable changes caused by specific updates. Supplementary plots may be viewed at http://ftp.as.harvard.edu/gcgrid/geos-chem/validation/InternalBenchmarks/13.4.0/.

1. Migrating sulfate chemistry into KPP

TBD

2. Fix bug in Henry law CALC_HEFF routine in hetchem computations

TBD

3. Fix scale factors applied to VOC aircraft emissions in AEIC inventory

The updated scale factors, which corrected the conversion from atoms C to kg species, resulted in significant decreases in emissions for VOC species. This was observed mostly in the northern hemispher and up to the lower stratosphere (where most aircraft emissions occur).
   ###################################################################################
   ### Emissions totals for inventory AEIC [Tg]                                    ###
   ### Ref = GCC_13.3.0-rc.1; Dev = GCC_13.4.0-rc.0                                ###
   ###################################################################################
                                     Ref                 Dev     Dev - Ref    % diff
    AEIC ACET          :           0.000139            0.000086     -0.000053   -38.027
    AEIC ALD2          :           0.001829            0.000996     -0.000833   -45.529
    AEIC ALK4          :           0.005617            0.004274     -0.001343   -23.912
    AEIC C2H6          :           0.000152            0.000122     -0.000031   -20.213
    AEIC C3H8          :           0.000022            0.000018     -0.000004   -18.386
    AEIC MACR          :           0.001827            0.000100     -0.001727   -94.522
    AEIC PRPE          :           0.004855            0.003779     -0.001076   -22.168
    AEIC RCHO          :           0.001384            0.003358      0.001974   142.621
However, this update resulted in minimal changes to global quantities, as VOC aircraft emissions are typically less than 1Gg for the simulation month. Observed changes from this update were:
  • Very small changes in oxidants (too small to show up in ratio plots).
  • Slight change in most primary organics, but % differences are of order 1e-4 or 1e-5.
  • Some zonal mean plots show diffs at top of atm where conc's are already low.
  • Mean OH decreased by -0.005%, CH4 and MCF increasd by +0.006%.
  • Very small changes in mass for most species.

4. Bug fix: Remove double-counting of HCl uptake on acidic sea salt

The following changes from this update were observed:

  • As expected, chlorine species concentrations decrease in the southern latitudes (esp. at about 40S), where there is more sea salt (and now less removal of HCl due to the bug fix).
  • Some increase of Cl species concentrations at northern latitudes are observed.
  • Cly concentrations generally increase.
  • Very small chagnes in OH metrics as well (mean OH -0.03% change).

5. Update surface resistance for ozone deposition to ice/snow

Barten et.al 2021 (https://doi.org/10.5194/acp-21-10229-2021) increased the surface resistance of ozone dry deposition to snow and ice to 1E4 s m−1, this results in a reduction of ozone deposition velocity to ice of <= 0.01 cm s−1 (~ 66% reduction) which has better agreement with observations and results in improved comparison between Ozone observations in the arctic.

Observed changes from this update include:

  • Increase in O3 concentration at the summer polar latitudes (i.e. deposition to ice/snow is suppressed);
  • Minor change in other oxidants;
  • Neglible effect on OH metrics;
  • Neglible change in species mass (O3 only increases by 0.01%).

6. Use updated definition of PM2.5; Add PM10 diagnostic

TBD

7. HEMCO bug fix: restore archiving via manual diagnostics (which had been inadvertently disabled in a prior update)

TBD

8. Use degassing-only volcano emissions in GEOS-Chem benchmark simulations

TBD

9. Add option for driving GEOS-Chem & GCHP with native GEOS meteorological files

TBD

The following updates introduced differences on the order of numerical noise:

TBD


The following were zero-difference updates:

  1. GCHP-specific updates:
  2. CH4 simulation updates:
  3. Hg simulation updates:
  4. TransportTracers simulation updates:
  5. Other updates:
Approval
Requires further investigation:
Approved by:
Date of approval: