GEOS-Chem 12 benchmark history

From Geos-chem
Jump to: navigation, search

On this page we have posted complete information about all benchmark simulations (both 1-month and 1-year) for GEOS-Chem 12.

1-year full-chemistry benchmarks

12.0.0-Run0

Three GEOS-Chem model versions were compared to each other:

Color Model Version Met Type Year Updates affecting the benchmark simulation Annual Mean OH
[105 molec/cm3]
O3 STE
[Tg a-1]
Black Observations          
Green 12.0.0-Run0 GEOS-FP,
72L, 4x5
2016 Updates introduced in this benchmark: 11.648 606.580
Blue v11-02f-Run1 GEOS-FP,
72L, 4x5
2016 Updates introduced in this benchmark: 11.532 593.857
Red v11-02e-Run1 GEOS-FP,
72L, 4x5
2016 Updates introduced in this benchmark:
  • Use GEOS-FP 4x5 met fields for 2016
11.541 603.246

The output plots for this run may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/GC_12/12.0.0/Run0/output
mget *

You may also view the PDF files online by pointing your browser to

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/GC_12/12.0.0/Run0/output/

Please also view the following pages comparing this version to past 1-year benchmarks:

--Melissa Sulprizio (talk) 11:44, 10 August 2018 (UTC)

1-year transport tracers benchmarks

12.2.0-TransportTracers

A 1-year Transport Tracers benchmark simulation was performed using GEOS-Chem 12.2.0. The simulation utilized 4° x 5° GEOS-FP met fields for the year 2016, with a 10-year spinup.

For comparison of the Pb-210 and Be-7 budgets to previous versions, please see the following posts on the Rn-Pb-Be simulation wiki page:

  1. Budget of Pb210 from 1-year benchmark simulations
  2. Budget of Be7 from 1-year benchmark simulations

You may view the benchmark plots for the simulation by pointing your browser to the following link:

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/GC_12/12.2.0/TransportTracers/output/

Benchmark comments:

  • **New in 12.2.0** A suite of transport tracers was added to the Rn-Pb-Be simulation to create a new Transport Tracers simulation to facilitate comparisons of transport schemes between GEOS-Chem "Classic" and/or GCHP with the NASA GEOS-DAS system. A new AgeofAir field was also added to State_Met and saved out via the Transport diagnostic collection for diagnosing the age of air.
  • The passive species (PassiveTracer) included in this benchmark and past Rn-Pb-Be benchmark simulations is used to test for mass conservation in GEOS-Chem transport. In past GEOS-Chem versions, we found that mass was not conserved when using the non-local PBL mixing scheme (VDIFF). To track the issue, we have been running the 1-year Rn-Pb-Be benchmark with and without non-local PBL mixing. However, a fix for mass conservation when using the NLPBL mixing scheme was added in GEOS-Chem 12.1.0 and the output from the 12.2.0-TransportTracers benchmark confirms that mass is now conserved for both mixing schemes.
PassiveTracer Summary for TURBDAY mixing
----------------------------------------
Max mass =  17.6750032410721 Tg
Min mass =  17.6750032410641 Tg
Abs diff =               8.0 g
Pct diff =            0.0000 %

PassiveTracer Summary for NLPBL mixing
---------------------------------------
Max mass =  17.6750032384831 Tg
Min mass =  17.6750032384776 Tg
Abs diff =               5.5 g
Pct diff =            0.0000 %

--Melissa Sulprizio (talk) 14:26, 30 January 2019 (UTC)

1-month benchmarks

12.3.0

Please see our GEOS-Chem 12.3.0 benchmark wiki page.

--Melissa Sulprizio (talk) 17:08, 15 March 2019 (UTC)

12.2.0

This section includes the assessment form for the 1-month benchmark simulation of 12.2.0.

Description
New features added into GEOS-Chem
Feature Submitted by
Features affecting the full-chemistry simulation:
Retire BRAVO emissions and use CEDS for Mexico instead Melissa Sulprizio (GCST)
Update mask file applied to NEI2011 emissions Melissa Sulprizio (GCST)
Features NOT affecting the full-chemistry simulation:
Transport Tracers simulation: a common set of tracers in GEOS-Chem and GEOS Transport WG
Melissa Sulprizio (GCST)
ObsPack diagnostic Andy Jacobson (NOAA/ESRL)
Bob Yantosca (GCST)
GFAS biomass burning emissions (as an option) Mat Evans (York)
Killian Murphy (York)
Fix bug in stratosperic aerosol surface area netcdf diagnostics Jonathan Moch (Harvard)
Christoph Keller (GMAO)
Update from HEMCO 2.1.010 to HEMCO 2.1.011: GCST
Features affecting GCHP only:
Update ESMF version from v5.2.0rp2 to v7.1.0r Lizzie Lundgren (GCST)
New fields in GCHP output restart file Lizzie Lundgren (GCST)
Bug fix: use checkpoint output file as subsequent run restart Lizzie Lundgren (GCST)
Version, resolution, met fields used: GEOS-FP (72L), 4x5, July 2016
1-month benchmark finished on: Sat Jan 26 05:39:11 EST 2019
Performance statistics:
  • Compiled with ifort 17.0.4
  • Ran on 24 CPUs of holyjacob01.rc.fas.harvard.edu (Intel(R) Xeon(R) CPU E5-2680 v3 "Haswell" @ 2.50 GHz)
  • Wall time: 09:06:46
  • CPU time / wall time: 20.2766
  • % of ideal performance: 84.49%
  • Memory: 13.1582 GB
Compared to previous benchmark: 12.1.0
This update will impact:
(select all that apply with boldface)
Advection, BL Mixing, Convection, Met Fields, Dry Dep, Wet Dep, Stratosphere, Anthro Emiss, Biogenic Emiss, Biomass Emiss, Photolysis, Chemistry, Other (please specify):
Unit test results may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.2.0/output/GC_12.2.0_Jan25.results.html
Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.2.0/output/
Metrics
Global mean OH from log file (% change): 12.5253573268589 x 105 molec/cm3 (0.016%)
Methyl chloroform lifetime (% change): 4.9508 years (-0.022%)
Methane lifetime (% change): 8.3429 years (-0.023%)
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, MEK, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4, NH3, NH4, NIT, BCPI, OCPI, BCPO, OCPO, BrO, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG1, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOA0, ISOA1, ISOA2, ISOA3, ASOA1, ASOA3, SOAP, SOAS, HPALD, DHDN, ETHLN, IEPOXA, IEPOXB, IEPOXD, ISN1, RIPA, RIPB, RIPD, IMAE, SOAIE, SOAME, LVOC, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, HOI, I2, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, pFe, OH
Comments on SURFACE differences: All differences can be attributed to updating the US mask file and retiring BRAVO emissions and using CEDS for Mexico instead. These changes impacted species concentrations at the surface over southern Canada and northern Mexico for several species with anthropogenic emissions. These changes propagated downwind and throughout the troposphere. See more details in the comments for emission differences below.
At 500 hPa, list all species that changed by 10% or more: ISOP, MVK, MACR, NPMN, PRPE, DMS, NH3, NIT, OCPI, OCPO, HBr, ISOPND, ISOPNB, MOBA, MACRN, XYLE, MTPA, LIMO, MTPO, TSOA2, TSOA3, ISOA3, SOAP, HPALD, ETHLN, IEPOXB, ISN1, RIPA, RIPB, RIPD, LVOC, ISN1OG, MONITU, IPMN, ClNO2, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, I2O4, pFe
Comments on 500 hPa differences:
  • See comments for surface differences
  • The differences in NH3 and NIT at 500 hPa can be attributed to numerical noise from ISORROPIA
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, NH3, NH4, NIT, BCPI, OCPI, BCPO, OCPO, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG1, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOA0, ISOA1, ISOA2, ISOA3, ASOA1, ASOA3, SOAP, SOAS, EOH, HPALD, DHDN, ETHLN, IEPOXA, IEPOXB, IEPOXD, ISN1, RIPA, RIPB, RIPD, IMAE, SOAIE, SOAME, LVOC, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, HC187, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, HOI, I2, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, pFe, OH
Comments on ZONAL MEAN differences:
  • See comments for surface differences
  • Most differences aloft are small and are located around 30°N and 60°N, corresponding to the changes in emissions over Mexico and Canada.
  • The >10% differences for many species listed above are primarily in regions were concentrations are very low (e.g. above 200 hPa).
In the EMISSION RATIO maps, list all species that changed by 10% or more: Anthropogenic emissions: ACET, ALD2, ALK4, BC, BENZ, C3H8, CH2O, CO, EOH, MEK, NH3, NO, OC, PRPE, SO2, SO4, TOLU, XYLE
Comments on EMISSION RATIO differences: Differences in anthropogenic emissions are observed near the US border in southern Canada and northern Mexico. These differences are caused by:
  1. Updating the US mask file applied to NEI2011 emissions. The NEI2011 emissions extend into Canada and Mexico, but the new mask file introduced in this version ensures NEI2011 emissions are only applied over the contiguous US and CEDS emissions are used for Canada and Mexico. APEI emissions are used for Canada for species included in that inventory (NO, CO, SO2, SO4, NH4, BC, and OC).
  2. Retiring BRAVO emissions and using CEDS for Mexico instead. In earlier GEOS-Chem versions, BRAVO emissions were only applied to one 4° x 5° grid box over Baja California, which was not covered by the older NEI2011 mask file. Retiring BRAVO impacted NO, CO, SO2, and SO4 emissions over that one grid box.
Additional or summary comments:

The wall time, scalability, and memory usage for this benchmark increased from 12.0.0. This is primarily because this benchmark saves out monthly netCDF diagnostics in HISTORY.rc. In addition, there are ongoing issues on the /n/regal/ scratch disk on Harvard's Odyssey cluster which have been found to negatively impact job performance.

An 1-month benchmark for 12.2.0 was also done with gfortran 8.2.0 for comparison with the official benchmark using ifort 17.0.4. The change in compiler produces essentially no change in output, except perhaps at the level of numerical noise.

Compiler # Cores Wall Time
[hh:mm:ss]
CPU time /
Wall time
% Ideal Memory
[GB]
Global Mean OH
[molec/cm3]
MCF lifetime
[a]
ifort 17.0.4 24 8:36:56 20.5517 85.63% 13.1714 12.5236147781285 4.9817
gfortran 8.2.0 24 8:29:21 18.2056 75.86% 11.8865 12.523614312375907 4.9817
  • In the table above, the run with gfortran 8.2.0 has comparable wall time to the run with ifort 17.0.4. CPU time / Wall time and & ideal were lower in the gfortran run, likely because gfortran cannot fully take advantage of in-chip optimizations available on Intel CPUs.

A 1-year Transport Tracers benchmark simulation was also performed for this version. Please see this wiki post for more details.

Approval
Requires further investigation: No
Approved by: Daniel Jacob
Date of approval: 14 Feb 2019

--Melissa Sulprizio (talk) 23:30, 28 January 2019 (UTC)

12.1.0

This section includes the assessment form for the 1-month benchmark simulation of 12.1.0.

Description
New features added into GEOS-Chem
Feature Submitted by
Features affecting the full-chemistry simulation:
Budget diagnostics Lizzie Lundgren (GCST)
Chris Holmes (Florida State)
Anthropogenic PM2.5 dust source Sajeev Philip (Dalhousie)
GEOS-Chem restart file updates Melissa Sulprizio (GCST)
Updates and fixes for the CEDS anthropogenic emissions inventory:
Pengfei Liu (Harvard)
Melissa Sulprizio (GCST)
Jiawei Zhuang (Harvard)
HEMCO fixes: GCST
Fix mass conservation in non-local PBL mixing Jintai Lin (PKU)
Allow for SZA to 98 degrees when getting photolysis rates Eric Fleming (GMAO)
Features NOT affecting the full-chemistry simulation:
FlexGrid Stage 1: Use HEMCO to read met fields Jiawei Zhuang (Harvard)
Jintai Lin (PKU)
Add fixes for seasonal scale factors in CH4 simulations Bram Maasakkers (SRON)
Melissa Sulprizio (GCST)
Convert input data for RRTMG from bpch to netCDF for input via HEMCO Bob Yantosca (GCST)
Added netCDF diagnostics in the following areas: Bob Yantosca (GCST)
Add support for continuous integration with TravisCI – automatically test each commit as soon as it is pushed to Github Jiawei Zhuang (Harvard)
Bob Yantosca (GCST)
Move Fast-JX .dat files from the geos-chem-unittest repository to gcgrid/ExtData Chris Holmes (Florida State)
GCST
Update HEMCO configuration file for the tagged CO simulation Bob Yantosca (GCST)
Structural updates to facilitate interfacing GEOS-Chem into WRF Haipeng Lin (Peking U)
GCST
Structural updates to facilitate interfacing GEOS-Chem into GEOS-5 Christoph Keller (GMAO)
Lizzie Lundgren (GCST)
Features affecting GCHP only:
Fix masking bug to correctly apply anthropogenic emissions Lizzie Lundgren (GCST)
Remove erroneous 180 degree longitude shift in APEI regridding Lizzie Lundgren (GCST)
Move GCHP run directory creation from GEOS-Chem Unit Tester to GCHP repository Lizzie Lundgren (GCST)
Version, resolution, met fields used: GEOS-FP (72L), 4x5, July 2016
1-month benchmark finished on: Wed Nov 21 16:48:07 EST 2018
Performance statistics:
  • Compiled with ifort 17.0.4 NOTE: Updated from Intel 11.1.069
  • Ran on 24 CPUs of holyjacob01.rc.fas.harvard.edu (Intel(R) Xeon(R) CPU E5-2680 v3 "Haswell" @ 2.50 GHz)
  • Wall time: 06:10:38
  • CPU time / wall time: 23.2185
  • % of ideal performance: 96.74
  • Memory: 8.7716 GB
Compared to previous benchmark: 12.0.0
This update will impact:
(select all that apply with boldface)
Advection, BL Mixing, Convection, Met Fields, Dry Dep, Wet Dep, Stratosphere, Anthro Emiss, Biogenic Emiss, Biomass Emiss, Photolysis, Chemistry, Other (please specify):
Unit test results may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.1.0/output/GC_12.1.0_Nov21.results.html
Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.1.0/output/
Metrics
Global mean OH from log file (% change): 12.5233901731440 x 105 molec/cm3 (-0.40%)
Methyl chloroform lifetime (% change): 4.9519 years (0.73%)
Methane lifetime (% change): 8.3448 years (0.75%)
At the SURFACE, list all species that changed by 10% or more: NO, O3, PAN, CO, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG1, TSOG2, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOA0, ISOG1, ISOG3, ISOA1, ISOA2, ISOA3, ASOG1, ASOG2, ASOG3, ASOAN, ASOA1, ASOA2, ASOA3, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOXA, IEPOXB, IEPOXD, ISN1, RIPA, RIPB, RIPD, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, LVOCOA, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, HC187, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, CH2ICl, CH2IBr, HOI, I2, IBr, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, ISALA, ISALC, AERI, pFe, OH, HO2
Comments on SURFACE differences: Below we summarize the notable changes caused by specific updates.

Anthropogenic PM2.5 dust source (aka AFCID)

  • This update introduced a large increase in DST1 concentrations globally.
  • The AFCID inventory adds ~1.1 Tg per month (= ~13 Tg/yr) to the DST1 species.

Updates and fixes for the CEDS anthropogenic emissions inventory

  • These updates introduced large changes for many species by introducing a missing anthropogenic emissions sector and adding ship emissions for many species that did not previously have ship emissions. The ship emissions update dominates and can be seen as large changes in species concentrations over the oceans. Adding the RCO sector greatly increases (>100% change) species concentrations over land for OCPO, OCPI, BENZ, TOLU, XYLE, TSOA*, ISOA*, ASOA*, SOAP, SOAS, EOH, GLYX, HCOOH, globally

HEMCO bug fix: prevent excessive attempts to read the HEMCO restart file when using GEOS-Chem "Classic"

  • This fix increased concentrations for many species over land by increasing biogenic emissions (see comments on emission ratios below).

Allow for SZA to 98 degrees when getting photolysis rates

  • This update caused changes in species concentrations near the poles for NO, NPMN, N2O5, HNO4, ISOPND, PROPNN, NO2, NO3, HNO2, ISN1, LVOC, ISN1OG, MONITS, MONITU, HONIT, OH, HO2, and many of the Br, Cl, and I species.

The following updates impacted species concentrations slightly, but the changes did not amount to >10%.

At 500 hPa, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4s, NH3, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG1, TSOG2, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOA0, ISOG1, ISOG3, ISOA1, ISOA2, ISOA3, ASOG1, ASOG2, ASOG3, ASOAN, ASOA1, ASOA2, ASOA3, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOXA, IEPOXB, IEPOXD, ISN1, RIPA, RIPB, RIPD, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, LVOCOA, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, HC187, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, CH2ICl, CH2IBr, I2, IBr, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, pFe, OH, HO2
Comments on 500 hPa differences: See comments for surface differences
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, O3, PAN, CO, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, CHBr3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG1, TSOG2, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOA0, ISOG1, ISOG2, ISOG3, ISOA1, ISOA2, ISOA3, ASOG1, ASOG2, ASOG3, ASOAN, ASOA1, ASOA2, ASOA3, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOXA, IEPOXB, IEPOXD, ISN1, RIPA, RIPB, RIPD, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, LVOCOA, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, HC187, OCS, BrCl, HCl, H2402, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, H2O, BrSALA, BrSALC, CH3I, CH2I2, CH2ICl, CH2IBr, HOI, I2, IBr, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, ISALA, ISALC, AERI, pFe, OH, HO2
Comments on ZONAL MEAN differences: See comments for surface differences
In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthropogenic + biofuel : ALD2, ALK4, BC, BENZ, C3H8, CH2O, CO, EOH, MEK, NH3, NO, OC, PRPE, SO2, SO4, TOLU, XYLE
  • Biogenic: ACET, ALD2, APIN, BCAR, BPIN, CARE, CO, EOH, FARN, ISOP, LIMO, MBOX, MONX, MYRC, OMON, OSQT, PRPE, SABI
  • Ship: CO, NO, SO2
  • DST1 emissions
  • Lightning NO
Comments on EMISSION RATIO differences:
Additional or summary comments: Important new features in this version that do not impact benchmark output include:

The switch from ifort 11.1.069 to ifort 17.0.4 produces essentially no change in output, except perhaps at the level of numerical noise. The data below were obtained from 1-month benchmark simulations with GEOS-Chem 12.0.0 submitted by Melissa Sulprizio:

Compiler # Cores Wall Time
[hh:mm:ss]
CPU time /
Wall time
% Ideal Memory
[GB]
Global Mean OH
[molec/cm3]
MCF lifetime
[a]
ifort 11.1.069 24 6:04:37 22.2942 98.29% 7.1892 12.4148399780817 4.9852
ifort 17.0.4 24 5:22:46 23.1919 96.63% 7.4660 12.4148067059835 4.9852
  • In the table above, the run with ifort 17.0.4 was approximately a half-hour faster than the run with ifort 11.1.069. This is because the ifort 11 compiler is at least 10 years old and cannot take advantage of in-chip optimizations available in more recent Intel CPUs such as Haswell, Broadwell, Knights Landing, etc.
Approval
Requires further investigation: No
Approved by: Daniel Jacob
Date of approval: 23 Nov 2018

--Melissa Sulprizio (talk) 22:37, 21 November 2018 (UTC)

12.0.0

This section includes the assessment form for the 1-month benchmark simulation of 12.0.0.

Description
New features added into GEOS-Chem
Feature Submitted by
Features affecting the full-chemistry simulation:
Turn off TRASH emissions by default to avoid double counting GCST
Eloise Marais (U. Birmingham)
Add entries to HEMCO_Config.rc for SOAP and pFe in CEDS inventory GCST
Bug fix for liquid water content in gckpp_HetRates.F90 Jiayue Huang (U. Washington)
Quanjie Chen (U. Washington)
Bug fix for CH4 boundary conditions Lee Murray (Rochester)
Fixed typo in call to GAMMA_HOBr_AER Xuan Wang
GCST
Correct issues when specifying DICE-Africa emissions: Eloise Marais (U. Birmingham)
GCST
Add missing Criegee intermediate reaction Xin Chen (UMN)
Bug fix for ARCTAS_SHIP in HEMCO_Config.rc files Barron Henderson (EPA)
Avoid double counting CEDS and GEIA NH3 biofuel emissions Christoph Keller (NASA/GMAO)
GCST
Bug fix in FlexChem: do not zero out ACTA, EOH, HCOOH concentrations Katie Travis (MIT)
Add GFED4.1 data for 2015 and 2016; Use GFED4.1 2016 in benchmarks Kate O'Dell (CSU)
GCST
Features NOT affecting the full-chemistry simulation:
Minor fixes for mercury chemistry updates implemented in v11-02f Colin Thackray (Harvard)
Restore default timestep for MERRA-2 nested grid simulations to 10/20 Ke Li (Harvard)
Remove duplicate C3HI spectrum in FJX_spec.dat and FJX_j2j.dat Tomas Sherwen (York)
Fixed typo that prevented PM2.5 from being archived to netCDF output Jonathan Moch (Harvard)
GCST
Bring two-way nested-grid code option up-to-date Yingying Yan (PKU)
Jintai Lin (PKU)
Do not force L=1 by default in planeflight_mod.F Katie Travis (MIT)
GCST
Add PH2O2 to the default list of prod/loss families in the KPP mechanisms GCST
Update to HEMCO v2.1.007, which includes minor updates:
  1. Stop with error if multiple containers have the same name
  2. Bug fix for distributing emissions in the vertical dimension
  3. Add extra error checks in HEMCO standalone module
  4. Remove null string character from netCDF unit string
  5. Bug fix for HEMCO soil NOx error with ifort 17
GCST
Christoph Keller
Jenny Fisher
Bug fix for reference time in netCDF diagnostic files GCST
Updates for QFED biomass burning emissions:
  1. Update to QFED v2.5r1
  2. Fix QFED v2.4r6 files for 2016/10/13
Christoph Keller (NASA GMAO)
Fix issues in complexSOA and complexSOA_SVPOA restart files Eloise Marais (U. Birmingham)
Bug fixes for v11-02 simulations using the RRTMG radiative transfer model Chris Holmes (U. Florida)
Now always declare the netCDF time variable with an unlimited dimension Chris Holmes (U. Florida)
Version, resolution, met fields used: GEOS-FP (72L), 4x5, July 2016
1-month benchmark finished on: Mon Jul 30 21:05:43 EDT 2018
Performance statistics:
  • Compiled with ifort 11.1.069
  • Ran on 24 CPUs of holyjacob01.rc.fas.harvard.edu (Intel(R) Xeon(R) CPU E5-2680 v3 "Haswell" @ 2.50 GHz)
  • Wall time: 06:38:04
  • CPU time / wall time: 21.9237
  • % of ideal performance: 91.35%
  • Memory: 7.3247 GB
Compared to previous benchmark: v11-02-rc
This update will impact:
(select all that apply with boldface)
Advection, BL Mixing, Convection, Met Fields, Dry Dep, Wet Dep, Stratosphere, Anthro Emiss, Biogenic Emiss, Biomass Emiss, Photolysis, Chemistry, Other (please specify):
Unit test results may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.0.0/output/GC_12.0.0.results.html
Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/GC_12/12.0.0/output/
Metrics
Global mean OH from log file (% change): 12.5734575378792 x 105 molec/cm3 (+1.80%)
Methyl chloroform lifetime (% change): 4.9162 years (-1.91%)
Methane lifetime (% change): 8.2830 years (-1.85%)
O3 STE flux (% change): 495.1870 Tg O3/year (-0.18%)
At the SURFACE, list all species that changed by 10% or more: NO, O3, PAN, CO, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, NH3, NH4, NIT, NITs, BCPI, BCPO, OCPI, OCPO, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG*, TSOA*, ISOG*, ISOA*, ASOG*, ASOA*, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOX*, ISN1, RIP*, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, LVOCOA, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, MONITA, INDIOL, IPMN, HC187, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, CH2ICl, CH2IBr, HOI, I2, IBr, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, ISALA, ISALC, AERI, pFe, OH, HO2
Comments on SURFACE differences: Below we summarize the notable changes caused by specific updates.


Turn off TRASH emissions by default to avoid double counting

  • This update resulted in slightly reduced concentrations over land for many species.

Add entries to HEMCO_Config.rc for SOAP and pFe in CEDS inventory

  • This fix caused increases in SOAP, SOAS, and pFe concentrations globally.

Bug fix for CH4 boundary conditions

  • This fix resulted in increased CH4 concentrations in the troposphere.

Bug fix for liquid water content in gckpp_HetRates.F90

  • This fix reduced many Br and Cl species concentrations globally while increasing HBr, BrSALA, and BrSALC.

Fix typo in call to GAMMA_HOBr_AER

  • This fix caused slight changes in Br and Cl species concentrations, mainly over the oceans

Fixes to DICE-Africa emissions

  • These updates primarily caused increases in species concentrations near the surface over and downwind of Africa.

Bug fix in FlexChem: do not zero out ACTA, EOH, HCOOH concentrations

  • This fix caused large increases in ACTA and HCOOH at all levels

Add GFED4.1 data for 2016 and 2016

  • Updating the GFED4 year from 2014 to 2016 caused >10% increases and decreases in species concentrations over land globally.

The following updates impacted species concentrations slightly, but the changes did not amount to >10%.

At 500 hPa, list all species that changed by 10% or more: NO, PAN, CO, ALK4, ISOP, HNO3, ACET, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, NH3, NH4, NIT, NITs, BCPI, BCPO, OCPI, OCPO, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG*, TSOA*, ISOG*, ISOA*, ASOG*, ASOA*, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOX*, ISN1, RIP*, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, LVOCOA, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, MONITA, INDIOL, IPMN, HC187, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, CH2ICl, CH2IBr, HOI, I2, IBr, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, pFe, OH
Comments on 500 hPa differences: See comments for surface differences
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, PAN, CO, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, NPMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, NH3, NH4, NIT, NITs, BCPI, BCPO, OCPI, OCPO, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPND, ISOPNB, MOBA, PROPNN, HAC, GLYC, MVKN, MACRN, MAP, NO2, NO3, HNO2, BENZ, TOLU, XYLE, MTPA, LIMO, MTPO, TSOG*, TSOA*, ISOG*, ISOA*, ASOG*, ASOA*, SOAP, SOAS, EOH, MGLY, GLYX, ACTA, HPALD, DHDN, ETHLN, HCOOH, IEPOX*, ISN1, RIP*, IMAE, SOAIE, SOAME, SOAGX, SOAMG, LVOC, ISN1OG, ISN1OA, MONITS, MONITU, HONIT, IONITA, MONITA, INDIOL, IPMN, HC187, CH4, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, BrSALA, BrSALC, CH2I2, CH2IBr, HOI, I2, ICl, I, IO, HI, OIO, INO, IONO, IONO2, I2O2, I2O3, I2O4, pFe, OH, HO2
Comments on ZONAL MEAN differences: See comments for surface differences
In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthropogenic emissions: ACET, ALD2, ALK4, C3H8, CO, MEK, NH3, NO, OC, PRPE, SO2, SO4, XYLE
  • Biomass burning emissions: ACET, ALD2, ALK4, BC, BENZ, C2H6, C3H8, CH2O, CO, EOH, MEK, MTPA, NH3, NO, OC, PRPE, SO2, TOLU, XYLE
  • NO fertilizer & soil emissions
  • SO2 ship emissions
Comments on EMISSION RATIO differences:
Additional or summary comments:
Approval
Requires further investigation: No
Approved by: GCST
Date of approval: 02 Aug 2018

--Melissa Sulprizio (talk) 17:10, 31 July 2018 (UTC)

GCHP benchmarks

TBD