Difference between revisions of "GEOS-Chem v10-01 benchmark history"

From Geos-chem
Jump to: navigation, search
(1-month benchmark v10-01d with tropospheric chemistry)
(1-month benchmark v10-01c with UCX chemistry)
Line 2,367: Line 2,367:
 
*Ran on 8 CPUs of bench@titan-09.as.harvard.edu (2.659 GHz x 8 CPU)
 
*Ran on 8 CPUs of bench@titan-09.as.harvard.edu (2.659 GHz x 8 CPU)
 
*Wall time: 5:51
 
*Wall time: 5:51
*Scalability: 6.7089
+
*[[Scalability|CPU time / wall time]]: 6.7089
 +
*% of ideal performance: 83.86%
 
|-
 
|-
 
|Compared to previous benchmark:
 
|Compared to previous benchmark:

Revision as of 18:47, 21 December 2015

On this page we have posted complete information about all benchmark simulations (both 1-month and 1-year) for GEOS-Chem v10-01.

1-year benchmarks

v10-01-public-release Rn-Pb-Be

A 1-year Rn-Pb-Be simulation was performed using GEOS-Chem v10-01-public-release. The simulation utilized 4° x 5° GEOS-FP met fields for the year 2013, with a 4-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:

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01-public-release/RnPbBe/output/

--Melissa Sulprizio (talk) 19:02, 15 June 2015 (UTC)

v10-01-public-release-Run0

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


Color Quantity Plotted Met Type Year Emissions Chemistry mechanism Annual Mean OH
[105 molec/cm3]
Red v10-01h-Run0 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component updates:

UCX chemistry mechanism, includes: 11.092
Green v10-01i-Run0 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component fixes:

" " 11.125
Blue v10-01-public-release-Run0 GEOS-FP,
72L, 4x5
2013 " " Benchmark chemistry mechanism, includes 11.723
Black Observations          

The output plots for Run0 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01-public-release/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/v10-01/v10-01-public-release/Run0/output/

--Melissa Sulprizio (talk) 20:02, 3 August 2015 (UTC)

v10-01i-Run0

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

Color Quantity Plotted Met Type Year Emissions Chemistry mechanism Annual Mean OH
[105 molec/cm3]
Red v10-01e-Run1 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component, with the following update included: UCX chemistry mechanism, includes: 11.843
Green v10-01h-Run0 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component updates:

" " 11.092
Blue v10-01i-Run0 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component fixes:

" " 11.125
Black Observations          

The output plots for Run0 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01i/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/v10-01/v10-01i/Run0/output/

--Melissa Sulprizio 17:03, 28 April 2015 (EDT)

Comments about the 1-year benchmark v10-01i

Randall Martin wrote:

Could there be an error in SO2 emissions that goes beyond MIX? In this plot (top left) I see more than a factor of 2 reduction in anthropogenic SO2 emissions in many regions worldwide.
Something similar appears to happen for NO, but the global emission totals for NO increase.
    ANTHROPOGENIC + BIOFUEL
     Tracer   v10-01e-geosfp-Run1   v10-01h-geosfp-Run0   v10-01i-geosfp-Run0
    ============================================================================
          NO            27.761212             30.087867             30.639205    Tg N
         SO2            48.892838             38.464923             38.087385    Tg S


Melissa Sulprizio wrote:

HEMCO has the capability to calculate total emissions per inventory, so I used that feature to diagnose how the emission inventories are changing between v10-01f, v10-01h, and v10-01i to see if that can explain the differences showing up in the benchmarks. Please see the attached comparisons for the SO2 anthropogenic emission inventories and NO anthropogenic emission inventories. The year column in the tables represents the base year of the emission inventory, which has changed for many of the inventories.
Anthropogenic SO2 emissions decrease for most of the updated inventories (EDGAR, NEI2011, EMEP), which would help explain the global decrease in SO2 concentrations. MIX shows an increase in the SO2 emission totals, but that is likely dominated by the increase in Indian SO2 emissions and the inclusion of Russia. As we see from the emission difference maps, SO2 emissions generally decrease over China. We used the old mask file that included Russia for v10-01h and v10-01i, so Russian emissions are still included here. We have applied the new MIX mask file as a last-minute fix for the v10-01 release.
The increase in global NO emissions is primarily driven by EMEP and AEIC emissions. There may be a problem in the IDL routine for calculating the anthro+biofuel totals for NO (I’m looking into this). In the NOx sources table in that same file, we don’t see large differences in the individual NO anthropogenic sources:
    NOx SOURCES
    Tracer   v10-01e-geosfp-Run1   v10-01h-geosfp-Run0   v10-01i-geosfp-Run0
    ============================================================================
       NOac             0.814670              0.814732              0.819548    Tg N
    NOan+bf            30.075295             30.087867             30.918083    Tg N

--Melissa Sulprizio 12:55, 6 May 2015 (EDT)

v10-01h-Run0

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

Color Quantity Plotted Met Type Year Emissions Chemistry mechanism Annual Mean OH
[105 molec/cm3]
Red v10-01c-Run1 GEOS-FP,
72L, 4x5
2013 Same as v9-02r-geosfp-Run1 UCX chemistry mechanism, includes: 11.885
Green v10-01e-Run1 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component, with the following update included: " " 11.843
Blue v10-01h-Run0 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component updates:

" " 11.092
Black Observations          

The output plots for Run0 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01h/Run0/output/pdf
mget *

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

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01h/Run0/output/pdf/

--Melissa Sulprizio 11:36, 13 April 2015 (EDT)

Comments about the 1-year benchmark v10-01h

Jeff Pierce wrote:

There are huge changes in dust and some in sea-salt. I'm not sure why these changed at all. There are some non-trivial changes to other aerosol species, but these could be due to the emissions inventory changes.

Melissa Sulprizio wrote:

The changes in dust and sea salt may attributed to the meteorology fields. In the previous 1-year benchmarks we ran for 2013/01 – 2013/08 then 2012/08 – 2013/01. We did that because of met field and OTD-LIS factor availability for GEOS-FP at the time. We have since downloaded the GEOS-FP met fields for all of 2013 and in this version (v10-01h) we updated the OTD-LIS factors for GEOS-FP through October 2014. Therefore, for this 1-year benchmark simulation we were able to run for the entirety of 2013.

Christoph Keller wrote:

FYI, there has also been a change in dry deposition that may be responsible for some of the observed changes: in the new version (10-01h), dry deposition of dust particles is done within the PBL mixing scheme (as for all other tracers). The former GEOS-Chem versions performed dust dry deposition within the dust chemistry routine.

Dylan Millet wrote:

I had a look at the benchmark plots, focusing on the VOCs. There are a lot of emission updates in this version and the composition changes seem to generally be explainable by that. A few notes:
  • isoprene emissions drop 30%, which is greater than the ~18% I expected based on tests when I implemented this change originally (10% decrease due to MEGAN changes + 8% decrease due to canopy change). But it's not such a difference that something is clearly wrong; may just be GEOS-FP vs GEOS-5.
  • anthro + biofuel propane: > factor of 2 decrease in emissions, 12 to 5 TgC. This looks to be largely due to a massive emission decrease in Asia. Is this right? Just want to check that emissions there aren't actually zero or anything.
  • CO generally shows a modest improvement compared to observations
  • some VOCs change quite a bit in the inter-mountain west, I suppose this is a deliberate change in NEI-11?
  • we're now using MEGAN for ALD2 (didn't previously have any biogenic emissions in the model) and PRPE, so emissions of these changed quite a bit
  • there are some strange spatial patterns in the October ratio plots for some species (e.g., ALK4, MEK), oscillating between 2x increases and 2x decreases ... I'm guessing this is because we're ratioing October of 2013 with October of 2012

Prasad Kasibhatla wrote:

I checked the emission ratio plots for biomass burning CO emissions for Jan, Apr, Jul, and Oct, as well as the annual total CO emissions from biomass burning against plots/totals that I created offline. The benchmark plots/totals match the plots/totals for CO that I generate offline. So I am confident that GFED3/4 emissions are correctly implemented in HEMCO in GC v10-01h.
A couple of quick notes:
  • The benchmark comparisons for biomass burning is not an apples-to-apples GFED3 vs GFED4 comparison because the GFED3 runs are with emissions for 2011 (because GFED3 does not go out to 2013) and the GFED4 runs are with emissions for 2013.
  • The GFED4 emissions are preliminary. The final version of GFED4 will be released in a month or so. At that point, we simply have to swap out the input files read in by GC.
  • One future change to the code is to be expected. GFED4 emission categories are slightly different compared to GFED3 emission categories. In the current implementation, I have mapped GFED4 dry matter burned categories into GFED3 categories offline so as to keep the emission factors the same. In the near future I expect to implement the GFED4 emission categories with updated emission factors.

Aaron van Donkelaar wrote:

The EMEP-related emission-ratio plots look good as well. One minor point of note is that the native resolution of the updated EMEP emissions are 50 km x 50 km, not 0.1° x 0.1° grid as previously stated on the wiki. Regridding to the finer 0.1° x 0.1° resolution allowed these input files to maintain a reasonable representation of their original grid characteristics, thereby permitting HEMCO to more accurately regrid onto a variety of GEOS-Chem grids. I've updated the wiki accordingly.

Chris Holmes wrote:

After seeing [the figures comparing v10-01i and v10-01f], I am comfortable that the PARANOX/HEMCO updates are consistent with what I expected. There is a small increase of 1-2 ppb O3 over the N. Pacific ship tracks, which are what I found in my personal benchmarks with just PARANOX changes. The changes of 10-15 ppb over the N. Atlantic appear to originate over Europe and North America (emission changes?) and overwhelm any expected effect of the PARANOX update.

Randall Martin wrote:

Maps that Melissa just made for the 1 month benchmark are helpful to identify which inventories are responsible for which changes.
           http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01h/emission_maps/
           http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01h/emission_differences/
These maps seem to indicate that the large (25%) reduction in global SO2 emissions is being driven by decreases from China. Does anyone know if this is expected? It appears somewhat surprising to me.

Yuxuan Wang wrote:

I noticed the large reduction of SO2 over China and was surprised too. SO2 emissions from China were reported to have decreased since 2006 and I expect this decrease trend to be included in the MIX Asia inventory which is used in v10-01i. The v10-01i benchmark run is for year 2012/2013 and presumably uses yearly specific emissions for SO2. If v10-01f_UCX uses the older INTEX-B inventory for SO2 emissions which is for the year 2006 (Melissa: could you please confirm this?), then the plotted difference in SO2 emissions over China between the two versions can be explained. This is just my guess.

Melissa Sulprizio wrote:

That is correct. The benchmarks for v10-01e and prior versions used Asian emissions from the Streets 2006 (INTEX-B) inventory, while v10-01h uses the MIX inventory. The MIX Asian emissions are available for years 2008 and 2010, so we use 2010 emissions in our benchmark simulations for v10-01h.

Qiang Zhang wrote:

Attached is a comparison between Streets 2006 and MIX 2010 over SE Asia. SO2 emissions didn't change too much over SE Asia. Large decrease of propane emissions and increase of ethane emissions are due to update of NMVOC speciation scheme, as documented in Li et al. (2014). Anthropogenic emissions of aldehydes and ketones are also significantly increased because OVOCs are included in the new NMVOC speciation scheme.
Daniel noticed change in far east Russia emissions. This is because of the new MIX 2010 inventory covers Asian parts of Russia and it uses different methods than EDGAR. To obtain a consistent emission estimates for Russia, I would suggest to use EDGAR for Asian Russia. We prepared a new MIX region mask file which rejected Asian Russia region.

Meng Li wrote:

I made plots of the comparisons between MIX and INTEX-B emission inventory for SO2 for January, April, July and October (which are attached) at 4 x 5 degree, and the results are identical to those from v10-01h benchmark. So I think there are no errors in MIX emissions in simulations of benchmark v10-01h.

--Melissa Sulprizio 11:16, 6 May 2015 (EDT)

v10-01e-Run1

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

Color Quantity Plotted Met Type Year Emissions Chemistry mechanism Photolysis Annual Mean OH
[105 molec/cm3]
Red v10-01c-Run0 GEOS-FP,
47L, 4x5
2013 Same as v9-02r-geosfp-Run1 Same as v9-02r-geos5-Run0 FAST-JX v7.0 photolysis mechanism, with the following fixes included: 12.550
Green v10-01c-Run1 GEOS-FP,
72L, 4x5
2013 " " UCX chemistry mechanism, includes: " " 11.885
Blue v10-01e-Run1 GEOS-FP,
72L, 4x5
2013 HEMCO emissions component, with the following update included: " " " " 11.843
Black Observations            

The output plots for Run1 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01e/Run1/output/pdf
mget *

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

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01e/Run1/output/pdf/

--Melissa Sulprizio 11:27, 13 April 2015 (EDT)

v10-01e-Run0

This 1-year benchmark simulation was approved by HEMCO developer Christoph Keller and the GEOS-Chem Steering Committee on 01 Dec 2014.

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

Color Quantity Plotted Met Type Year Emissions Photolysis Linoz Annual Mean OH
[105 molec/cm3]
Red v9-02r-geosfp-Run1 GEOS-FP,
47L, 4x5
2013 Same as v9-02r-geos5-Run0
+Anthropogenic emissions scaled to 2010 (latest available)
+ 2011 GFED3 emissions (latest available)
+ Olson 2001 land map
Same as v9-02r-geos5-Run0 ON 12.689
Green v10-01c-Run0 GEOS-FP,
47L, 4x5
2013 " " FAST-JX v7.0 photolysis mechanism, with the following fixes included: ON 12.550
Blue v10-01e-Run0 GEOS-FP,
47L, 4x5
2013 HEMCO emissions component, with the following update included: " " ON 12.578
Black Observations            

The output plots for Run0 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01e/Run0/output/pdf
mget *

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

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01e/Run0/output/pdf/

--Melissa Sulprizio 11:07, 20 November 2014 (EST)

Comments about the 1-year benchmark v10-01e

Jeff Pierce wrote:

The carbonaceous aerosols have definitely changed since the last benchmark, but now OCPI over tropical forests is less than 10% of the OCPI pre-HEMCO (page 35 of the ratios). Can you confirm that the 0.1*monoterpenes is being added to OCPI as it was pre-HEMCO?

Melissa Sulprizio responded:

I found an error in the MEGAN emissions. In the HEMCO configuration file, we have MEGAN_SOA turned on by default because HEMCO knows to ignore the SOA species if we are not running a SOA simulation. However, with MEGAN_SOA turned on, we don’t execute the part of the code where we add 0.1*monoterpenes to OCPI. Therefore, as you noted, we see lower OCPI concentrations over tropical forests. Turning MEGAN_SOA off fixes this problem.

Shiliang Wu wrote:

The O3 profiles show some large increases (up to 30ppb or more) in Jan for certain locations, such as Uccle(51), Hohenpeissenberg(48), Kagoshima(32), Shanghai(31), etc. - these don't look to be due to emission change, so I wonder what's causing this?

Christoph Keller wrote:

There were two open questions regarding the 1-yr benchmark (at least to my knowledge):
1. The ozone ‘blob’ forming over Europe in January
2. The much lower concentration of NITs and SO4s in v10-01e compared to v10-01c.
I think I have an explanation for both points, and I would argue that the new code makes more sense physically:
1. The ozone blob is actually a good thing as it removes an issue that was introduced in v9-02r. As Melissa pointed out to me, ever since v9-02r we have been seeing a drastic decrease of ozone over Europe (and to a lesser extent East Asia) in January. This was because PARANOX could produce - under low O3 conditions - incredibly high O3 deposition values (division of a high number by a very, very small number). This would basically create a black hole for ozone at the surface that would then remove more and more ozone from the entire column. Melissa did a number of sensitivity studies where she ran v10-01c with and without PARANOX, and those plots show this effect very nicely. The new code does not allow such a behavior anymore, which explains the higher ozone concentrations in January compared to v10-01e.
2. NITs and SO4s are closely linked to sulfate chemistry and thus alkalinity. Alkalinity is directly obtained from the total mass of emitted sea salt aerosol, e.g. we approximate alkalinity = SSA. Since SSA is a 2D field but alkalinity is 3D, the sea salt aerosol need to be distributed vertically. In the old code, alkalinity was set to the total mass of emitted sea salt aerosol for every grid box that was below the PBL (zero otherwise). In other words, if the PBL was at level 11, the integrated alkalinity over this column was 10 times higher than the total mass of emitted sea salt aerosols in the same column. I found that not very realistic and therefore changed the alkalinity calculation so that it takes into account the fraction of the PBL. I also checked in with Fabien Paulot who has been working on this and he also thinks that this is the better way of doing it. After all, it ensures that the total alkalinity equals the total amount of emitted sea salt aerosols. This explains the much lower values for NITs and SO4s that we see in v10-01e compared to v10-01c.

--Melissa Sulprizio 11:33, 1 December 2014 (EST)

v10-01c-Run1

This 1-year benchmark simulation was approved by the GEOS-Chem Steering Committee on 26 Jun 2014.

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

Color Quantity Plotted Met Type Year Chemistry mechanism Photolysis Linoz Annual Mean OH
[105 molec/cm3]
Red v9-02r-geosfp-Run1 GEOS-FP,
47L, 4x5
2013 Same as v9-02r-geos5-Run0 Same as v9-02r-geos5-Run0 ON 12.689
Green v10-01c-Run0 GEOS-FP,
47L, 4x5
2013 " " FAST-JX v7.0 photolysis mechanism, with the following fixes included: ON 12.550
Blue v10-01c-Run1 GEOS-FP,
72L, 4x5
2013 UCX chemistry mechanism, includes: " " ON 11.885
Black Observations            

The output plots for Run1 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run1/output/ps
mget *
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run1/output/pdf
mget *

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

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run1/output/pdf/

--Melissa Sulprizio 16:54, 23 June 2014 (EDT)

v10-01c-Run0

This 1-year benchmark simulation was approved by the GEOS-Chem Steering Committee on 26 Jun 2014.

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

Color Quantity Plotted Met Type Year Emissions Photolysis Linoz Annual Mean OH
[105 molec/cm3]
Red v9-02r-geosfp-Run0 GEOS-FP,
47L, 4x5
2013 Same as v9-02r-geos5-Run0
(scaled all emissions to 2005)
Same as v9-02r-geos5-Run0 ON 12.389
Green v9-02r-geosfp-Run1 GEOS-FP,
47L, 4x5
2013 Anthropogenic emissions scaled to 2010 (latest available)
+ 2011 GFED3 emissions (latest available)
+ Olson 2001 land map
" " ON 12.689
Blue v10-01c-Run0 GEOS-FP,
47L, 4x5
2013 " " FAST-JX v7.0 photolysis mechanism, with the following fixes included: ON 12.550
Black Observations            

The output plots for Run0 (both PostScript and PDF format) may be downloaded from:

ftp ftp.as.harvard.edu
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run0/output/ps
mget *
cd gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run0/output/pdf
mget *

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

http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v10-01/v10-01c/Run0/output/pdf/

--Melissa Sulprizio 14:56, 10 June 2014 (EDT)

1-month benchmarks

v10-01-public-release

We performed two 1-month benchmark simulations for v10-01-public-release:

  1. v10-01-public-release-Run0: 1-month benchmark with SOA turned off
  2. v10-01-public-release-Run1: 1-month benchmark with SOA turned on

1-month benchmark v10-01-public-release with SOA on

Here is the assessment form for 1-month benchmark simulation v10-01-public-release with SOA turned on (aka v10-01-public-release-Run1).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • GFED4: Prasad Kasibhatla (Duke), Christoph Keller (Harvard)
  • RRTMG fixes: David Ridley (MIT), GEOS-Chem Support Team
  • org.dat fix: David Ridley (MIT)
  • CH4 bands: Katie Travis (Harvard)
  • PARANOX fix: GEOS-Chem Support Team
  • Wetdep fixes: Viral Shah (UW)
  • SOA benchmark and fix: GEOS-Chem Support Team
  • QFED2: Katie Travis (Harvard), Christoph Keller (Harvard)
  • gamap_mod.F fix: Sebastian Eastham (MIT)
  • PGI fixes: Gan Luo (SUNY/Albany), GEOS-Chem Support Team
  • Makefile updates: GEOS-Chem Support Team
  • BROMOCARB_SEASON fix: Luke Schiferl (MIT), GEOS-Chem Support Team
  • HEMCO updates: Christoph Keller (Harvard)
  • I3 field fix: GEOS-Chem Support Team
  • Update ICOADS_SHIP files: GEOS-Chem Support Team
  • 0.25 x 0.3125 nested EU: Anna Protonatariou (Athens)
  • Removal of obsolete modules: GEOS-Chem Support Team
  • Mercury fixes: GEOS-Chem Support Team
  • TOMAS fixes: Jack Kodros (CSU), Jeff Pierce (CSU), GEOS-Chem Support Team
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Fri Jun 19 01:01:48 EDT 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 7:47
  • CPU time / wall time: 6.8285
  • % of ideal performance: 85.36%
Compared to previous benchmark: v10-01-public-release-Run0
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/v10-01/v10-01-public-release-Run1.tar.gz
  • NOTE: Unit tests for tagged CO were not performed, since this simulation is not yet 100% compatible with HEMCO.
Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01-public-release-Run1.tar.gz
Metrics
Global mean OH (from log file): 12.5101002745403 x 105 molec/cm3
Methyl chloroform lifetime: 4.9906 years
Did either of these change by more than 5%? No. The mean OH differs by -0.68%, and the MCF lifetime differs by 0.80%.
At the SURFACE, list all species that changed by 10% or more: NO, ALK4, ISOP, MVK, MACR, PMN, N2O5, DMS, OCPI, DST4, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, HNO2, Cl ,ClNO2, Cl2, Cl2O2
Comments on SURFACE differences:
  • Most differences are small number differences over the oceans or high latitudes where concentrations are low.
  • Differences in the biogenic species (ISOP and its derivatives) are likely caused by either (1) small-number differences or (2) the SOA chemistry. Additional biogenic species (e.g. LIMO, FARN, BCAR, OQST) are now are now included in the chemistry mechanism, but were absent from the prior benchmark (v10-01-public-release-Run0).
  • Differences in the OCPI tracer are due to the fact that biogenic OC emissions are not calculated for the SOA chemistry mechanmism. Christoph Keller wrote about this:
    • Regarding the diagnostics, biogenic OC emissions are not calculated for the SOA simulation because they are not used by GEOS-Chem. Because of that, I also think that the OC-BIOG diagnostics are misleading for the SOA simulation because they contain values, but those are not used… What we could do is to make sure that the biogenic OC diagnostics is defined but zero for the SOA simulation.
At 500 hPa, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, PRPE, N2O5, DMS, NH3, NIT, OCPI, DST4, ISOPN, MOBA, RIP, IEPOX, Cl, ClNO2, Cl2O2
Comments on 500 hPa differences:
  • See comments for SURFACE DIFFERENCES above.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, ALK4, ISOP, MVK, MACR, PMN, PRPE, C3H8, N2O5, SO4s, NH3, NIT, NITs, OCPI, BCPO, OCPO, SALC, Br, BrNO3, ISOPN, MOBA, RIP, IEPOX, OCS, CCl4, CH3CCl3, CFC11, H1211, H1301, H2402, Cl, ClNO2, Cl2O2
Comments on ZONAL MEAN differences:
  • See comments for SURFACE DIFFERENCES above.
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
  • This benchmark now contains biogenic emissions for SOA species LIMO, FARN, BCAR, OSQT. The prior benchmark (v10-01-public-release-Run0) did not contain any SOA species, and therefore emissions for those species were all zero.
  • The emission totals for NO soil and fertilizer emissions decreased slightly (both by -0.000043 Tg N). Because the soil NOx scheme depends on the amount of N that is deposited to the surface, we believe that these differences may be attributed to the impact of the SOA chemistry. The amount of NO in the atmosphere with SOA chemistry activated is likely a little bit different than in the prior benchmark, thus accounting for a slightly different amount of N being deposited to the surface.
Additional or summary comments:
Approval
Requires further investigation: No
Approved by: GCST
Date of approval: 16 Jun 2015

--Bob Y. (talk) 14:13, 16 June 2015 (UTC)
--Melissa Sulprizio (talk) 20:54, 22 June 2015 (UTC)

1-month benchmark v10-01-public-release with SOA off

Here is the assessment form for 1-month benchmark simulation v10-01-public-release with SOA turned off (aka v10-01-public-release-Run0).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • GFED4: Prasad Kasibhatla (Duke), Christoph Keller (Harvard)
  • RRTMG fixes: David Ridley (MIT), GEOS-Chem Support Team
  • org.dat fix: David Ridley (MIT)
  • CH4 bands: Katie Travis (Harvard)
  • PARANOX fix: GEOS-Chem Support Team
  • Wetdep fixes: Viral Shah (UW)
  • SOA benchmark and fix: GEOS-Chem Support Team
  • QFED2: Katie Travis (Harvard), Christoph Keller (Harvard)
  • gamap_mod.F fix: Sebastian Eastham (MIT)
  • PGI fixes: Gan Luo (SUNY/Albany), GEOS-Chem Support Team
  • Makefile updates: GEOS-Chem Support Team
  • BROMOCARB_SEASON fix: Luke Schiferl (MIT), GEOS-Chem Support Team
  • HEMCO updates: Christoph Keller (Harvard)
  • I3 field fix: GEOS-Chem Support Team
  • Update ICOADS_SHIP files: GEOS-Chem Support Team
  • 0.25 x 0.3125 nested EU: Anna Protonatariou (Athens)
  • Removal of obsolete modules: GEOS-Chem Support Team
  • Mercury fixes: GEOS-Chem Support Team
  • TOMAS fixes: Jack Kodros (CSU), Jeff Pierce (CSU), GEOS-Chem Support Team
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Fri Jun 19 07:45:48 EDT 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 6:33
  • CPU time / wall time: 7.0051
  • % of ideal performance: 87.56%
Compared to previous benchmark: v10-01i
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/v10-01/v10-01-public-release-Run0.tar.gz
  • NOTE: Unit tests for tagged CO were not performed, since this simulation is not yet 100% compatible with HEMCO.
Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01-public-release-Run0.tar.gz
Metrics
Global mean OH (from log file): 12.5954701437773 x 105 molec/cm3
Methyl chloroform lifetime: 4.9818 years
Did either of these change by more than 5%? No. The mean OH differs by 0.23%, and the MCF lifetime differs by 0.25%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, CO, ALK4, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST4, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on SURFACE differences:
  • The differences in NOx-related species over the oceans are attributed to the update to PARANOX that fixed an issue in the PARANOX SUNCOS fields.
  • The mask file for MIX Asian emissions now excludes Russia and instead we use EDGAR v4.2 emissions for that country. This change has led to some differences over Russia.
  • The fix to the photolysis code in SET_PROF has caused differences in some species (e.g. H2O2, N2O5, HNO3, Br2) over the dust laden regions of northern Africa and the Middle East. This corrects an issue that was introduced in v10-01i.
  • In the prior benchmark (v10-01i), we were using the latitudinal CH4 distribution (from NOAA flask data) for the year 2007, which was latest available year at the time. During the period of public comment, Katie Travis provided additional years of latitudinal CH4 data. This benchmark used Katie's updated latitudinal CH4 data for the year 2013 instead of 2007. This may result in some changes in the chemistry.
At 500 hPa, list all species that changed by 10% or more: NO, ALK4, ISOP, HNO3, MEK, RCHO, MVK, MACR, PMN, R4N2, PRPE, C3H8, N2O5, MP, DMS, SO2, SO4, NH3, NIT, NITs, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALC, Br2, Br, HOBr, HBr, BrNO2, BrNO3, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO3, HNO2, BrCl, Cl, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH
Comments on 500 hPa differences:

See comments for SURFACE DIFFERENCES above.

In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, ALK4, ISOP, HNO3, H2O2, MVK, MACR, PMN, PPN, PRPE, C3H8, N2O5, HNO4, MP, DMS, SO2, SO4s, NH3, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, OCS, BrCl, HCl, CCl4, CH3CCl3, CFC11, H1211, H1301, H2402, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, H2O, HO2
Comments on ZONAL MEAN differences:

See comments for SURFACE DIFFERENCES above.

In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthropogenic emissions: ACET, ALD2, ALK4, C3H8, CH2O, CO, MEK, NH3, NO, PRPE, SO2, SO4
  • Biomass burning emissions: ACET, ALD2, ALK4, BC, C2H6, C3H8, CH2O, CO, MEK, NH3, NO, OC, PRPE, SO2
  • Biogenic emissions: ISOP
  • NO fertilizer emissions
Comments on EMISSION RATIO differences:
  • Differences in anthropogenic emissions over Russia was caused by an updated MIX Asia mask. The mask file for MIX emissions now excludes Russia and instead we use EDGAR v4.2 emissions for that country.
  • Anthropogenic NO changed over the oceans. This is caused by the update to PARANOX that fixed an issue in the PARANOX SUNCOS fields.
  • Differences in biomass burning emissions are caused by the switch to officially released GFED4 files.
  • The Bug fix in I3 field interpolation affects the following temperature-dependent emissions:
    • Biogenic and natural emissions computed via MEGAN (e.g. ISOP, ACET, PRPE and other hydrocarbon species)
    • NO emissions for these sectors: lightning, fertilizer, soils
    • Dust emissions (computed by the DEAD dust algorithm)
  • Biogenic emissions of MONX are now restored. These appear to have been suppressed in the prior benchmark.
  • GFED emissions from deforestation now take the weighted sum of deforestation and woodland emission factors, based on the humid tropical forest mask. This update causes the biomass burning emissions totals to be slightly different w/r/t the prior benchmark v10-01h.
Additional or summary comments:
  • Comparing this benchmark to the prior benchmark (v10-01i) will allow us to validate the updates that were added to GEOS-Chem during the period of public comment (May 1 to June 15, 2015).
  • The technical issues with RRTMG (parallelization errors and inefficient subroutine calls) have now been corrected.
  • Flexible precision has been implemented into the v10-01-public-release code.
  • The HEMCO standalone executable (hemco_standalone.x) is now built each time you compile GEOS-Chem. GEOS-Chem users can use the HEMCO standalone program to test if their emission inputs are giving the expected results before they try to run those emissions in a full GEOS-Chem simulation.
  • The v10-01-public-release version contains the latest updates for the TOMAS aerosol microphysics simulation. The TOMAS team is still validating the results (this is ongoing).
  • We have removed much legacy code that HEMCO has now made obsolete.
  • Dave Ridley wrote:
    • I’ve just eyeballed the difference in the J-Values that we saw between v10-01i and 10-01h versus the 10-01i and 10-01i_fix and it looks like a near perfect reversal… so the J-values should now be similar to the pre-RRTMG 10-01h values.
Approval
Requires further investigation: No
Approved by: GCST
Date of approval: 16 Jun 2015

--Bob Y. (talk) 14:14, 16 June 2015 (UTC)
--Melissa Sulprizio (talk) 20:35, 22 June 2015 (UTC)

v10-01i

We performed two 1-month benchmark simulations for v10-01i:

  1. v10-01i: 1-month benchmark with RRTMG turned off
  2. v10-01i_RRTMG: 1-month benchmark with RRTMG turned on

1-month benchmark v10-01i with RRTMG off

Here is the assessment form for 1-month benchmark simulation v10-01i with RRTMG turned off (aka v10-01i).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • GCRT: David Ridley (MIT), Colette Heald (MIT)
  • EMEP: Aaron van Donkelaar (Dalhousie)
  • PARANOX fixes: Christoph Keller (Harvard)
  • AEIC: Christoph Keller (Harvard)
  • CO2 emission updates: Ray Nassar (Environment Canada)
  • Planeflight fix: Karen Yu (Harvard)
  • 0.25x0.3125 nested-grid updates: Lin Zhang (PKU)
  • Drydep update: Katie Travis (Harvard)
  • HEMCO updates: Christoph Keller (Harvard)
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Thu Apr 23 04:30:25 EDT 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 6:37
  • CPU time / wall time: 6.9602
  • % of ideal performance: 87.00%
Compared to previous benchmark: v10-01h
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/v10-01/v10-01i.tar.gz

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01i.tar.gz
Metrics
Global mean OH (from log file): 12.5663562879762 x 105 molec/cm3
Methyl chloroform lifetime: 4.9693 years
Did either of these change by more than 5%? No. The mean OH differs by -1.25%, and the MCF lifetime differs by -1.73%.
At the SURFACE, list all species that changed by 10% or more: NO, O3, PAN, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, HH4, NIT, NITs, BCPI, BCPO, OCPI, BCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH
Comments on SURFACE differences:
  • Large differences in NO, O3, HNO3, (and probably indirectly PAN, NO, NO2, NO3) over surface oceans can be attributed to fixing the bug in PARANOX that was present in v10-01h. In v10-01h ship emissions were inadvertently turned off, but these were restored in v10-01i. Concentrations of these species return to expected levels in v10-01i.
  • The reprocessed AEIC aircraft emissions may be responsible for some differences observed in NO, CO, SO2, SO4, BCPI, OCPI, ACET, ALD2, ALK4, C2H6, C3H8, CH2O, PRPE, MACR, RCHO. (You can see emissions along the flight corridors in the plots for SO2 and SO4.)
  • Many species show a pattern of decreased concentration over the Sahara and increased concentration in other locations in the Northern Hemisphere. We believe this is due to the updated optical properties in the photolysis that were added with the RRTMG update.
  • The updated EMEP emissions cause some minor differences in species concentration over Europe (NO, SO2, SO4, ALK4, ALD2, PRPE, MEK)
  • Differences in DST4 are probably caused by numerical roundoff. These occur over oceans where concentrations are already very low.
  • Changes in OH are likely driven by both the change in aerosol optical properties for the photolysis, as well as the bug fix in PARANOX which restored realistic concentrations of O3 and nitrogen species.
At 500 hPa, list all species that changed by 10% or more: NO, ALK4, HNO3, H2O2, MEK, MVK, MACR, PMN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2 (minor changes), ClOO, OClO, Cl2, Cl2O2, OH
Comments on 500 hPa differences: See comments for SURFACE DIFFERENCES above.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, O3, PAN, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, BCPO, OCPO, DST2, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, CH3CCl3, CFC11, H1211, H24O2, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2,
Comments on ZONAL MEAN differences:
  • See comments on SURFACE DIFFERENCES above.
  • DST2, DST4, SALC, ISOPN, MOBA, CH3CCl3, CFC11, H1211, H24O2, Cl, ClO, ClNO3, ClNO2, ClOO, OClO, Cl2, and Cl2O2 appear to be small-number differences (in some cases at the very top of the atmosphere).
In the EMISSION RATIO maps, list all species that changed by 10% or more: ACET, ALD2, ALK4, C2H6, C3H8, CO, ISOP, NH3, NO, SO2, SO4
Comments on EMISSION RATIO differences:
  • The reprocessed EMEP emissions are responsible for differences in these species over Europe:
    • ALD2 (anthro + biofuel)
    • ALK4 (antho + biofuel)
    • CO (anthro + biofuel)
    • NO (anthro + biofuel)
    • NH3 (anthro only)
    • SO2 (column anthro)
    • SO4 (anthro + biofuel)
    • MEK (anthro + biofuel)
    • PRPE (anthro + biofuel)
  • The reprocessed AEIC aircraft emissions are responsible differences in these species:
    • NO, CO, SO2, SO4, BC, OC, NH3
    • Hydrocarbons: ACET, ALD2, ALK4, C2H6, C3H8, CH2O, PRPE, MACR, RCHO, MEK
  • A new HEMCO restart file is responsible for differences in NO fertilizer and soil emissions. A new HEMCO restart file was spun up to correct PARANOX_SUNCOS values that were one hour off in the original file.
  • GFED emissions from deforestation now take the weighted sum of deforestation and woodland emission factors, based on the humid tropical forest mask. This update causes the biomass burning emissions totals to be slightly different w/r/t the prior benchmark v10-01h.
Additional or summary comments:
  • In v10-01h, there existed a bug which caused ship emissions to be inadvertently turned off. This bug was corrected in v10-01i. Fixing this bug returns the O3 to expected levels over the N. Hemisphere oceans. But when compared to v10-01h (where O3 over the oceans is too low), we see large differences in the plots.
  • We also ran a special 1-month benchmark to test GEOS-Chem v10-01i with the RRTMG radiative transfer model turned ON. For more information about this benchmark simulation, please follow this link.
  • Dave Ridley pointed out a reduction in the OC and BC AOD in v10-01i. He found that the aerosol optics file (org.dat) used in v10-01i didn't include Randall Martin's updates for OC growth. Dave writes:
The change in the org optics seems to tie in with the reduction in the organic AOD we’re seeing, so that makes sense. The BC optics are the same, so the differences seen there shouldn’t be coming from the optics. There may have been a change in the density because of a lot of BC tweaks around the same time.
Approval
Requires further investigation: No
Approved by: Daniel Jacob
Date of approval: 30 Apr 2015

--Bob Y. 16:37, 30 April 2015 (EDT)

1-month benchmark v10-01i with RRTMG on

Here is the assessment form for 1-month benchmark simulation v10-01i with RRTMG turned on (aka v10-01I_RRTMG).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • GCRT: David Ridley (MIT), Colette Heald (MIT)
  • EMEP: Aaron van Donkelaar (Dalhousie)
  • PARANOX fixes: Christoph Keller (Harvard)
  • AEIC: Christoph Keller (Harvard)
  • CO2 emission updates: Ray Nassar (Environment Canada)
  • Planeflight fix: Karen Yu (Harvard)
  • 0.25x0.3125 nested-grid updates: Lin Zhang (PKU)
  • Drydep update: Katie Travis (Harvard)
  • HEMCO updates: Christoph Keller (Harvard)
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Mon Apr 27 11:48:13 EDT 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 15:32
  • CPU time / wall time: 4.0263
  • % of ideal performance: 50.33%
Compared to previous benchmark: v10-01i
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/v10-01/v10-01i_RRTMG.tar.gz

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01i_RRTMG.tar.gz/
Metrics
Global mean OH (from log file): 12.5663562879762 x 105 molec/cm3
Methyl chloroform lifetime: 4.9693 years
Did either of these change by more than 5%? No. The mean OH and MCF lifetimes are unchanged from 1-month benchmark v10-01i.
At the SURFACE, list all species that changed by 10% or more: None
Comments on SURFACE differences:
At 500 hPa, list all species that changed by 10% or more: None
Comments on 500 hPa differences:
In the ZONAL MEAN differences, list all species that changed by 10% or more: None
Comments on ZONAL MEAN differences:
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
Additional or summary comments:

Bob Yantosca wrote:

There were a few technical issues in the code that you should be aware of:
  • We had to turn off all parallelization in the interface between GEOS-Chem and RRTMG (module GeosCore/rrtmg_rad_transfer_mod.F). There is a parallelization error somewhere that causes the ND72 diagnostic output (tracers 44, 55, 56, 66, 67, 77, 78, 88 in diagnostic category RADMAP-$) yield different results when parallelization is turned OFF vs. when it is turned ON. We did not have time before IGC7 to investigate this further, so for the present we are just turning off all parallelization as the simplest fix.
  • There are several array temporaries created in subroutine calls to MCICA_SUBCOL_SW, MCICA_SUBCOL_LW, RRTMG_SW, and RRTMG_LW. These are caused by passing non-contiguous sub-slices of arrays to the subroutine. These can be fixed by using pointers to access the sub-slices of arrays. We will work in fixing this at a later time.
  • In GEOS-Chem v10-01, all fields of the State_Met object use (I,J,L) array ordering, to facilitate connection to the NASA GEOS-5 GCM. But in the module GeosCore/rrtmg_rad_transfer_mod.F, leftover code from a previous version still referred to State_Met%CLDF(L,I,J) and State_Met%AIRDEN(L,I,J), which caused out-of-bounds errors. This has now been fixed.
--Bob Y. 13:57, 23 April 2015 (EDT)

Dave Ridley wrote:

I think everything is good to go – there is definitely an issue in that the photolysis code isn’t seeing the right aerosol, and this affects things like HNO3 and H2O2 in dust laden regions. But the changes aren’t too dramatic other than those species, so I think we can sign off on the version release and get that sorted during the provisional period.
--Bob Y. 15:57, 1 May 2015 (EDT)
Approval
Requires further investigation: Some minor issues will be sorted out during the public comment period.
Approved by: David Ridley, Colette Heald
Date of approval: 01 May 2015

--Melissa Sulprizio 10:53, 28 April 2015 (EDT)
--Bob Y. 15:56, 1 May 2015 (EDT)

v10-01h

Here is the assessment form for 1-month benchmark simulation v10-01h

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • MEGAN updates: Dylan Millet (U. Minnesota)
  • Ship NOx updates: Chris Holmes (UC Irvine); Geert Vinken (Eindhoven)
  • EDGAR v4.2: Qiang Zhang (Tsinghua); Meng Li (Tsinghua); Sajeev Philip (Dalhousie)
  • MIX: Qiang Zhang (Tsinghua); Meng Li (Tsinghua)
  • NEI2011: Katie Travis (Harvard)
  • EMEP updates: Aaron van Donkelaar (Dalhousie)
  • OTD-LIS update: Lee Murray (Columbia)
  • GFED4: Prasad Kasibhatla (Duke), Christoph Keller (Harvard)
  • HEMCO bug fix: Christoph Keller (Harvard)
  • Planeflight fix: GEOS-Chem Support Team
  • WETLOSS fix: Carey Friedman (MIT)
  • NetCDF linking: GEOS-Chem Support Team
  • FINN: Jenny Fisher (U. Wollongong); Jingqiu Mao (Princeton); Christine Wiedinmyer (NCAR)
  • MASAGE NH3: Fabien Paulot (NOAA, formerly Harvard)
  • HTAP: Qiang Zhang (Tsinghua U.)
  • New HEMCO features: Christoph Keller (Harvard)
  • Optimization error fix: GEOS-Chem Support Team
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Thu Mar 26 21:09:41 EDT 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 6:33
  • CPU time / wall time: 6.9042
  • % of ideal performance: 86.30%
Compared to previous benchmark: v10-01f_UCX
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/v10-01/v10-01h.tar.gz

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01.tar.gz
Metrics
Global mean OH (from log file): 12.4116120372531 x 105 molec/cm3
Methyl chloroform lifetime: 5.0567 years
Did either of these change by more than 5%? No. The mean OH differs by -1.04%, and the MCF lifetime differs by 2.01%.
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, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, HCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on SURFACE differences:
  • Most of these differences are probably caused by the introduction of several new emissions inventories (see COMMENTS ON EMISSION RATIO DIFFERENCES below).
  • Also see the ADDITIONAL OR SUMMARY COMMENTS below.
At 500 hPa, list all species that changed by 10% or more: NO, O3, PAN, CO, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on 500 hPa differences:
  • Most of these differences are probably due to the introduction of several new emissions inventories (see COMMENTS ON EMISSION RATIO DIFFERENCES below).
  • Also see the ADDITIONAL OR SUMMARY COMMENTS below.
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, PMN, 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, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, OCS, BrCl, HCl, Cl4, CH3CCl3, H1211, H1301, H2402, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on ZONAL MEAN differences:
  • Most of these differences are probably due to the introduction of several new emissions inventories (see COMMENTS ON EMISSION RATIO DIFFERENCES below).
  • Also see the ADDITIONAL OR SUMMARY COMMENTS below.
In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthro+biofuel emissions: ACET, ALD2, ALK4, BC, C2H6, C3H8, CH2O, CO, MEK, NO, OC, PRPE, SO2, SO4
  • Biomass emissions: ACET, ALD2, ALK4, BC, C2H6, C3H8, CH2O, CO, MEK, NO, OC, PRPE, SO2
  • Biogenic emissions: ACET, ISOP, OC, PRPE
  • Other:
    • ACET from MONOT emissions, ACET from MBO emissions, ACET direct emissions
    • CO from MONOT emissions
    • NO fertilizer emissions, NO soil emissions, NO column lightning
    • SO2 volcanic emissions
Comments on EMISSION RATIO differences:
Additional or summary comments: General comments:
  • An apparent optimization error was causing data to be regridded incorrectly. This error has now been fixed in the code used to generate these benchmark output plots.
  • Small differences in the J-value ratios are attributed to the fact that HEMCO now reads the Hermann & Celarier (1997) UV Albedo data at 2° x 2.5° resolution, and now regrids to 4° x 5° resolution. The prior benchmarks read this data directly at 4° x 5° resolution.
  • The increase in wallclock time from 6:04 to 6:33 can be attributed to the extra I/O needed to read and regrid the hourly NEI2011 emissions.
  • In v10-01h, the application of dry deposition for aerosol species is now done in a centralized location before the KPP solver is called. In prior versions, this was done in many separate routines after the KPP solver was called.
  • In v10-01h, HEMCO now reads international time zone boundaries from an input file. In previous versions, each time zone was considered to be 1/24th of the world.

Also note the following discussions pertaining to various emissions now handled by HEMCO:

Approval
Requires further investigation:
Approved by:
Date of approval:

--Bob Y. 12:39, 27 March 2015 (EDT)

v10-01g

We did not run a 1-month benchmark simulation for v10-01g. This version consisted entirely of bug fixes and structural updates. We will validate these updates in the 1-month benchmark v10-01h, along with several HEMCO emissions updates.

--Melissa Sulprizio 18:04, 25 February 2015 (EST)

v10-01f

We performed two 1-month benchmark simulations for v10-01f:

  1. v10-01f_trop: 1-month benchmark with tropopsheric chemistry
  2. v10-01f_UCX: 1-month benchmark with UCX strat-trop chemistry (72 vertical levels)

1-month benchmark v10-01f with UCX chemistry

Here is the assessment form for 1-month benchmark simulation v10-01f with UCX chemistry (aka v10-01f_UCX).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • ALD2 photolysis update: Jingqiu Mao (Princeton), Sebastian Eastham (MIT)
  • Stratospheric Bry fixes: Johan Schmidt (Harvard)
  • UCX fix: Sebastian Eastham (MIT)
  • Two-way coupling: Jintai Lin (Peking U.), Yingying Yan (Peking U.)
  • Tagged CO fixes: Jenny Fisher (U. Wollongong)
  • Flexible precision: GEOS-Chem Support Team
  • Nested CH updates: Yuxuan Wang (Tsinghua/Galveston)
  • HEMCO updates: Christoph Keller (Harvard)
  • Scavenging by co-condensation fix: Duncan Fairlie (NASA/Langley)
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Thu Jan 8 20:26:53 EST 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu
  • Wall time: 6:04
  • CPU time / wall time: 6.8575
  • % of ideal performance: 85.72%
Compared to previous benchmark: v10-01e_UCX
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01f.tar.gz
(in directory v10-01f_UCX)

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01f.tar.gz<tt>
(in directory <tt>v10-01f_UCX
)
Metrics
Global mean OH (from log file): 12.5428873166735 x 105 molec/cm3
Methyl chloroform lifetime: 4.9566 years
Did either of these change by more than 5%? No. The mean OH differs by -3.40%, and the MCF lifetime differs by 3.58%.
At the SURFACE, list all species that changed by 10% or more: NO (S. pole), ALK4, ISOP (at poles and over oceans), HNO3 (paranox bug), MEK, MVK, MACR, PMN, PRPE, N2O5, HNO4, DMS, SO4S and NITS (HEMCO bug), Br2, BrO, BrNO3, MPN, ISOPN, MOBA,, GLYC, MMN, RIP, IEPOX, NO2, NO3, BrCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH
Comments on SURFACE differences:
  • NO: Difference over south pole appear may reflect small # differences where concentrations are very low.
  • ISOP and related species (ISOPN, MOBA, HAC, GLYC, MMN, RIP, MAP, IEPOX) display similar patterns of small # differences, where concentrations are very low.
  • Differences in Br/Cl species (Br2, BrO, BrNO3, BrCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2O2) are likely attributed to the updated stratospheric Bry files. Also, many of these species may also be exhibiting small # differences (esp. near the atmosphere top) where the concentrations are already low.
  • OCPI, OCPO: HEMCO now computes emission for hydrophilic and hydrophobic OC as separate species. In the v10-01e 1-month benchmark, HEMCO computed total BC and OC, then separated into hydrophilic and hydrophobic parts
  • NH3: Small diffs at the surface may be attributed to the bug fix in wet scavenging by co-condensation.
  • SO4s and NITs: This is a result of bug fix in how HEMCO computes the alkalinity. This fix is in v10-01f and the v10-01e 1-year benchmark, but not in the v10-01e 1-month benchmark.
  • HNO3: Differences over oceans were the result of a bug in the HEMCO PARANOX ship emissions module that was present in the v10-01e 1-month benchmark, but has since been fixed.
At 500 hPa, list all species that changed by 10% or more: NO, ALK4, ISOP, HNO3, MEK, MVK, MACR, PMN, PRPE, N2O5, HNO4, DMS, SO4s and NITs (HEMCO bug), NH3, NIT, OCPI, OCPO, DST4, Br2, Br, BrO, HBr, BRNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, NO2, NO3, HNO2, BrCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on 500 hPa differences:
  • See notes above
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, O3, PAN (near tropopause), PRPE, N2O5 (near trop), HNO4, MP, DMS, OCPI, CH4, H2O (in strat), OH, HO2, ISOP, ISOPN, MOBA, HAC, GLYC, MMN, RIP, MAP, IEPOX, ALK4, ACET (near top of atm), MEK, ALD2, RCHO, MVK, MACR, PMN, R4N2, PRPE, C3H8, CH2O, DMS, SO4s, NH3, NIT, NITs, BCPO, OCPO, DST2, DST3, DST4, SALC, MPN, NO2, NO3, HNO2, N2O, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, CHBr3, CH2Br2, CH3Br, BrCl, HCl, H1211, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2.
Comments on ZONAL MEAN differences:
  • Differences in NOx species (NO, NO2, NO3, HNO2, NHNO3, HNO4, N2O5, PAN) near the top of the atmosphere are likely attributed to the implementation of 2D boundary conditions for individual NOx species.
  • ISOP and related species (ISOPN, MOBA, HAC, GLYC, MMN, RIP, MAP, IEPOX) display similar patterns of small # differences, where concentrations are very low.
  • Differences in the following species are likely due to small # differences: ALK4, ACET (near top of atm), MEK, ALD2, RCHO, MVK, MACR, PMN, R4N2, PRPE, C3H8, CH2O, DMS, SO4s, NH3, NIT, NITs, BCPO, OCPO, DST2, DST3, DST4, SALC, MPN, NO2, NO3, HNO2, N2O:
  • Differences in stratospheric species (Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, CHBr3, CH2Br2, CH3Br, BrCl, HCl, H1211, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2) may be attributed to a combination of:
    1. The updated stratospheric Bry data files in HEMCO, and
    2. Small # differences where concentrations are already low
  • HNO3: Differences near the surface were the result of a bug in the HEMCO PARANOX ship emissions module that was present in the v10-01e 1-month benchmark, but has since been fixed. Differences at other altitudes may be caused by small number differences.
  • Differences in OCPO: HEMCO now computes emission for hydrophilic and hydrophobic OC as separate species. In the v10-01e 1-month benchmark, HEMCO computed total BC and OC, then separated into hydrophilic and hydrophobic parts.
  • SO4s and NITs: This is a result of bug fix in how HEMCO computes the alkalinity. This fix is in v10-01f and the v10-01e 1-year benchmark, but not in the v10-01e 1-month benchmark.
In the EMISSION RATIO maps, list all species that changed by 10% or more: NO (anthro+biofuel), SO2 (anthro+biofuel), SO4 (anthro+biofuel)
Comments on EMISSION RATIO differences:
  • The lower values of NO, SO2, SO4 anthro + biofuel w/r/t v10-01e appear to be caused by a diagnostic issue that has been solved in v10-01f.
  • Also, some slight differences in the emission totals for the following species are explained as follows:
    • OCbg (Biogenic OC): Due to an error in the v10-01e 1-month benchmark, this diagnostic had returned all zeroes. This issue was fixed for the v10-01e 1-year benchmark and in the v10-01f 1-month benchmark.
    • The slight numerical differences in NOfe and NOso (fertilizer and soil NOx emission) was caused by updates made to introduce the flexible precision into v10-01f. Certain constants in routine BIOFIT are now defined with full precision in v10-01f.
    • The differences in BChp and OChp (hydrophilic tracer from hydrophobic tracer) are due to the fact that in v10-01e, HEMCO carried total BC and OC, whereas in v10-01f, HEMCO now carries the hydrophilic and hydrophobic fractions as separate tracers: BCPI, BCPO, OCPI, OCPO. Also, in v10-01f, the Bond et al emissions files were reprocessed in order to split the hydrophilic and hydrophobic fractions.
Additional or summary comments:
Approval
Requires further investigation: No
Approved by: Bob Y.
Date of approval: 09 Jan 2015

--Melissa Sulprizio 17:38, 6 January 2015 (EST)

1-month benchmark v10-01f with tropospheric chemistry

Here is the assessment form for 1-month benchmark simulation v10-01f with tropospheric chemistry (aka v10-01f_trop).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • ALD2 photolysis update: Jingqiu Mao (Princeton), Sebastian Eastham (MIT)
  • Stratospheric Bry fixes: Johan Schmidt (Harvard)
  • UCX fix: Sebastian Eastham (MIT)
  • Two-way coupling: Jintai Lin (Peking U.), Yingying Yan (Peking U.)
  • Tagged CO fixes: Jenny Fisher (U. Wollongong)
  • Flexible precision: GEOS-Chem Support Team
  • Nested CH updates: Yuxuan Wang (Tsinghua/Galveston)
  • HEMCO updates: Christoph Keller (Harvard)
  • Scavenging by co-condensation fix: Duncan Fairlie (NASA/Langley)
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Wed Jan 7 14:59:29 EST 2015
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu
  • Wall time: 2:49
  • CPU time / wall time: 7.0954
  • % of ideal performance: 88.69%
Compared to previous benchmark: v10-01e_trop
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01f.tar.gz
(in directory v10-01f_trop)

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01f.tar.gz
(in directory v10-01f_trop)
Metrics
Global mean OH (from log file): 13.5245943486961 x 105 molec/cm3
Methyl chloroform lifetime: 4.7552 years
Did either of these change by more than 5%? No. The mean OH differs by -0.15%, and the MCF lifetime differs by -0.14%.
At the SURFACE, list all species that changed by 10% or more: NO, ISOP, HNO3, MVK, MACR, PMN, SO2, SO4s, NH4, NIT, OCPI, OCPO, DST4, ISOPN, RIP, NH3, HNO3
Comments on SURFACE differences:
  • NO: Difference over south pole appear may reflect small # differences where concentrations are very low.
  • ISOP, MVK, MACR, PMN, ISOPN, RIP, SO2, DST4: Appear to be small # differences over the ocean where concentrations are already low.
  • OCPI, OCPO: HEMCO now computes emission for hydrophilic and hydrophobic OC as separate species. In the v10-01e 1-month benchmark, HEMCO computed total BC and OC, then separated into hydrophilic and hydrophobic parts
  • NH3: Small diffs at the surface may be attributed to the bug fix in wet scavenging by co-condensation.
  • SO4s and NITs: This is a result of bug fix in how HEMCO computes the alkalinity. This fix is in v10-01f and the v10-01e 1-year benchmark, but not in the v10-01e 1-month benchmark.
  • HNO3: Differences over oceans were the result of a bug in the HEMCO PARANOX ship emissions module that was present in the v10-01e 1-month benchmark, but has since been fixed.
At 500 hPa, list all species that changed by 10% or more: NO, SO4s, NIT, NITs, OCPI, OCPO, DST4, ISOPN, RIP
Comments on 500 hPa differences:
  • NO: Differences at S. Pole may be caused by small # differences where concentrations are very low.
  • ISOPN, RIP: Appear to be small # differences, where concentrations are already low.
  • SO4s and NITs: This is a result of bug fix in how HEMCO computes the alkalinity. This fix is in v10-01f and the v10-01e 1-year benchmark, but not in the v10-01e 1-month benchmark.
  • NIT: Small differences can be attributed to numerical drift caused by ISORROPIA.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, C3H8, SO2, SO4s, NH3, NIT, NITs, OCPO, DST4, ISOPN, MAP, Br, Br2, BrNO3, NH3
Comments on ZONAL MEAN differences:
  • NO: Differences at S. Pole may be reflect small # differences where concentrations are very low.
  • Br, Br2, BrNO3: These differences are introduced by the new stratospheric Bry files via HEMCO. In the v10-01e code, we read these data at 4° x 5° resolution. In v10-01f, we read these at 2° x 2.5° and then regrid to 4° x 5°. Also, the new stratospheric Bry files correct the issue where the daytime & nighttime stratospheric Bry were the same.
  • ISOP, MVK, MACR, PMN, ISOPN, MAP, DST4, SO2, C3H8: These appear to be caused by small # differences where concentrations are already low.
  • SO4s and NITs: This is a result of bug fix in how HEMCO computes the alkalinity. This fix is in v10-01f and the v10-01e 1-year benchmark, but not in the v10-01e 1-month benchmark.
  • NH3: Small diffs at the surface may be attributed to the bug fix in wet scavenging by co-condensation.
  • NIT: Small differences can be attributed to numerical drift caused by ISORROPIA.
  • HNO3: Differences over oceans were the result of a bug in the HEMCO PARANOX ship emissions module that was present in the v10-01e 1-month benchmark, but has since been fixed.
In the EMISSION RATIO maps, list all species that changed by 10% or more: NO (anthro+biofuel), SO2 (anthro+biofuel), SO4 (anthro+biofuel)
Comments on EMISSION RATIO differences:
  • The lower values of NO, SO2, SO4 anthro + biofuel w/r/t v10-01e appear to be caused by a diagnostic issue that has been solved in v10-01f.
  • Also, some slight differences in the emission totals for the following species are explained as follows:
    • OCbg (Biogenic OC): Due to an error in the v10-01e 1-month benchmark, this diagnostic had returned all zeroes. This issue was fixed for the v10-01e 1-year benchmark and in the v10-01f 1-month benchmark.
    • The slight numerical differences in NOfe and NOso (fertilizer and soil NOx emission) was caused by updates made to introduce the flexible precision into v10-01f. Certain constants in routine BIOFIT are now defined with full precision in v10-01f.
    • The differences in BChp and OChp (hydrophilic tracer from hydrophobic tracer) are due to the fact that in v10-01e, HEMCO carried total BC and OC, whereas in v10-01f, HEMCO now carries the hydrophilic and hydrophobic fractions as separate tracers: BCPI, BCPO, OCPI, OCPO. Also, in v10-01f, the Bond et al emissions files were reprocessed in order to split the hydrophilic and hydrophobic fractions.
Additional or summary comments:
Approval
Requires further investigation: No
Approved by: Bob Y.
Date of approval: 08 Jan 2015

--Melissa Sulprizio 17:38, 6 January 2015 (EST) and
--Bob Y. 14:42, 8 January 2015 (EST)

v10-01e

We performed two 1-month benchmark simulations for v10-01e:

  1. v10-01e_trop: 1-month benchmark with tropopsheric chemistry
  2. v10-01e_UCX: 1-month benchmark with UCX strat-trop chemistry (72 vertical levels)

1-month benchmark v10-01e with UCX chemistry

Here is the assessment form for 1-month benchmark simulation v10-01e with UCX chemistry (aka v10-01e_UCX).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Thu Nov 6 00:26:55 EST 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 5:44
  • CPU time / wall time: 7.0316
  • % of ideal performance: 87.90%
Compared to previous benchmark: v10-01d_UCX
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01e.tar.gz
(in directory /v10-01e_UCX)

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01e.tar.gz
(in directory /v10-01e_UCX)
Metrics
Global mean OH (from log file): 12.9849685422903 x 105 molec/cm3
Methyl chloroform lifetime: 4.7803 years
Did either of these change by more than 5%? No. The mean OH differs by -0.74%, and the MCF lifetime differs by 0.49%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, Cl, ClO, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH, HO2
Comments on SURFACE differences:
  • HEMCO now reads all input data from netCDF files and automates the emission calculation process. Specifically, all input data (base emissions, scale factors, masks) are regridded onto the simulation grid before any data manipulation is performed. This fact alone can cause local emission differences on the order of 10%, e.g. for GFED-3 emissions over some areas in Africa and South America.
  • NO emissions over the US are approx. 25% higher compared to v10-01d. This is partly because of the regridding artifact mentioned above and partly because the NEI2005 emissions have now been lumped into the surface level. In v10-01d, NEI-2005 input data were distributed over 5 levels but only the lowest 2 were used for emission calculation.
  • The difference in ISOP and related biogenic species can be attributed to the fact that the MEGAN biogenic emissions in HEMCO were modified to only use the current day's leaf area index (LAI). In v10-01d and prior versions, the LAI was interpolated from the previous and current months. This change may be responsible for the large relative change of e.g. ISOP above 60 degrees N. Absolute concentrations are very small in the areas where relative changes are large.
  • Differences in DMS may be attributed to the fact that HEMCO uses a new updated air-sea exchange module. The air-sea exchange module centralizes this computation and uses the same parameterization scheme for all compounds. (In v10-01d and prior versions, this computation was repeated in several locations in the code.) The new air-sea exchange module seems to give lower DMS concentrations than the prior code.
  • The C2H6 differences are due to biofuel emissions over Europe, North America and Asia, which were not taken into account in v10-01d. This effect is most obvious in Asia, where biofuel emissions are highest.
  • SO2: HEMCO uses a different volcanic input file (still the same Diehl inventory) and a new vertical regridding scheme, which may be responsible for the SO2 differences. The same is also true for the AEIC emissions, which are now vertically regridded on-the-fly by the new vertical regridding scheme.
  • UCX stratospheric species are almost all similar to v10-01d.
At 500 hPa, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, R4N2, PRPE, C2H6, N2O5, HNO4, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, Cl, ClNO2, OClO, Cl2, Cl2O2
Comments on 500 hPa differences:
  • See comments on surface differences. Differences in SO2 and SO4 may be due to the new AEIC aircraft emission inventory as well as differences between the bpch and netCDF files of the volcanic emissions.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, PRPE, C3H8, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, BCPO, OCPO, DST1, DST2, DST3, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, OCS, BrCl, CCl4, CH3CCl3, CFC11, H1211, H1301, H2402, Cl, HOCl, ClNO3, ClNO2, ClOO, OClO, Cl2, Cl2O2, OH
Comments on ZONAL MEAN differences:
  • Most zonal diffs are small in absolute value but may be large in % value. This may reflect small-number statistics.
  • Differences in DMS and ACET may be attributed to the new air-sea exchange module used by HEMCO.
  • Differences of ISOP and other related biogenic species over the Northern boreal latitudes may be attributed to the change that was made in MEGAN. MEGAN now uses the current day's LAI instead of interpolating it from the previous & current months.
  • Changes in H2O2 may be related to wet deposition.
In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthropogenic emissions: ACET, ALD2, ALK4, C2H6, C3H8, CH2O, CO, MEK, NH3, NO, PRPE, SO2, SO4
  • Biomass emissions: ACET, ALD2, ALK4, C2H6, C3H8, CH2O, CO, MEK, OC, PRPE, SO2
  • Biogenic emissions: ACET, DMS, ISOP, PRPE
  • OTHER:
    • ACET from MONOT emissions, ACET from MBO emissions, ACET direct emissions, ACET from ocean source emissions
    • CO from MONOT emissions
    • NO aircraft emissions, NO fertilizer emissions, NO soil emissions, NO lightning emissions
    • SO2 aircraft emissions, SO2 ship emissions
Comments on EMISSION RATIO differences:
  • HEMCO does not separate anthropogenic and biofuel emissions. In some plots we are comparing the total of anthro+biofuel (v10-01e) to v10-01d.
  • HEMCO now passes ACET ocean sink to dry deposition, and doesn't explicitly write it out as a diagnostic
  • HEMCO now lumps eruptive and non-eruptive volcano emissions into one diagnostic (SO2-EV-$)
  • Differences in aircraft and volcano emissions w/r/t v10-01d are likely introduced by the MESSy vertical regridding routines.
  • Differences in ISOP and other key biogenic-emitting species are likely due to the change in how MEGAN uses leaf-area-index data w/r/t v10-01d.
  • ACET and DMS emissions now use a different air-sea exchange module (in HEMCO) w/r/t v10-01d.
Additional or summary comments:
  • GEOS-Chem v10-01d and prior versions use a klunky emissions interface. There was a significant amount of hardwiring, especially in the application of scale factors and masks. In v10-01e, HEMCO has completely replaced this legacy emissions interface. We have tried to replicate the emissions output r of v10-01d to the greatest extent possible. But in v10-01e+HEMCO, we have also brought in a couple of newer inventories (volcanoes, etc.) while also introducing a new vertical regridding scheme. So we will not be able to replicate the v10-01d emissions to 100% accuracy.
  • In v10-01d, the PARANOX ship plume model was called immediately before PBL mixing. In v10-01e, PARANOX is now called from within HEMCO at the same time that all the other emissions are computed.
  • As of this benchmark, the following simulations are not yet fully compatible with HEMCO.
  • The APM aerosol microphysics will have to be completely re-integrated once v10-01 ships.
Approval
Requires further investigation: No
Approved by: Bob Y. and Christoph Keller
Date of approval: 06 Nov 2014

--Melissa Sulprizio 06:29, 6 November 2014 (EST)
--Bob Y. 10:24, 6 November 2014 (EST)

1-month benchmark v10-01e with tropospheric chemistry

Here is the assessment form for 1-month benchmark simulation v10-01e with tropospheric chemistry (aka v10-01e_trop).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Wed Nov 5 18:31:03 EST 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-10.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 2:40
  • CPU time / wall time: 7.2007
  • % of ideal performance: 90.00%
Compared to previous benchmark: v10-01d_trop
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01e.tar.gz
(in directory v10-01e_trop)

NOTE: Unit tests for tagged CO and TOMAS were not performed, since these simulations are not yet 100% compatible with HEMCO.

Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01e.tar.gz
(in directory v10-01e_trop)
Metrics
Global mean OH (from log file): 13.5883620833673 x 105 molec/cm3
Methyl chloroform lifetime: 4.7619 years
Did either of these change by more than 5%? No. The mean OH differs by -0.62%, and the MCF lifetime differs by 0.22%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, H2O2, ACET, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, OH, HO2
Comments on SURFACE differences:
  • HEMCO now reads all input data from netCDF files and automates the emission calculation process. Specifically, all input data (base emissions, scale factors, masks) are regridded onto the simulation grid before any data manipulation is performed. This fact alone can cause local emission differences on the order of 10%, e.g. for GFED-3 emissions over some areas in Africa and South America.
  • NO emissions over the US are approx. 25% higher compared to v10-01d. This is partly because of the regridding artifact mentioned above and partly because the NEI2005 emissions have now been lumped into the surface level. In v10-01d, NEI-2005 input data were distributed over 5 levels but only the lowest 2 were used for emission calculation.
  • The difference in ISOP and related biogenic species can be attributed to the fact that the MEGAN biogenic emissions in HEMCO were modified to only use the current day's leaf area index (LAI). In v10-01d and prior versions, the LAI was interpolated from the previous and current months. Many of the differences are evident in the boreal forests, and in the Amazon, where a small change in LAI could result in a big increase in ISOP. ISOP increases by about ~5ppb in these regions.
  • Differences in ACET and DMS may be attributed to the fact that HEMCO uses a new updated air-sea exchange module. The air-sea exchange module centralizes this computation and uses the same parameterization scheme for all compounds. (In v10-01d and prior versions, this computation was repeated in several locations in the code.) The new air-sea exchange module seems to give lower concentrations than the prior code.
  • Differences in C2H6 may be attributed to biofuel emissions over Europe, North America and Asia, which were not taken into account in v10-01d. This effect is most obvious in Asia, where biofuel emissions are highest.
  • SO2: HEMCO uses a different volcanic input file (still the same Diehl inventory) and a new vertical regridding scheme, which may be responsible for the SO2 differences. The same is also true for the AEIC emissions, which are now vertically regridded on-the-fly by the new vertical regridding scheme.
At 500 hPa, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, R4N2, PRPE, C2H6, N2O5, HNO4, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, OCPO, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2
Comments on 500 hPa differences:
  • See comments on surface differences.
  • Differences in SO2 and SO4 may be due to the new AEIC aircraft emission inventory as well as differences between the bpch and netCDF files of the volcanic emissions.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, PRPE, C3H8, CH2O, C2H6, N2O5, HNO4, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, OCPI, BCPO, OCPO, DST3, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, OH
Comments on ZONAL MEAN differences:
  • Most zonal diffs are small in absolute value but may be large in % value. This may reflect small-number statistics.
  • Differences in DMS and ACET may be attributed to the new air-sea exchange module used by HEMCO.
  • Differences in C2H6 may be related to how aircraft emissions are vertically regridded by HEMCO.
  • Differences of ISOP and other related biogenic species over the Northern boreal latitudes may be attributed to the change that was made in MEGAN. MEGAN now uses the current day's LAI instead of interpolating it from the previous & current months.
  • Changes in H2O2 may be related to wet deposition.
In the EMISSION RATIO maps, list all species that changed by 10% or more:
  • Anthropogenic emissions: ACET, ALD2, ALK4, C2H6, C3H8, CH2O, CO, MEK, NH3, NO, PRPE, SO2, SO4
  • Biomass emissions: ACET, ALD2, ALK4, C2H6, C3H8, CH2O, CO, MEK, OC, PRPE, SO2
  • Biogenic emissions: ACET, DMS, ISOP, PRPE
  • OTHER:
    • ACET from MONOT emissions, ACET from MBO emissions, ACET direct emissions, ACET from ocean source emissions
    • CO from MONOT emissions
    • NO aircraft emissions, NO fertilizer emissions, NO soil emissions, NO lightning emissions
    • SO2 aircraft emissions, SO2 ship emissions
Comments on EMISSION RATIO differences:
  • HEMCO does not separate anthropogenic and biofuel emissions. In most plots we are comparing the total of anthro+biofuel (v10-01e) to v10-01d, so differences are apparent.
  • HEMCO now passes ACET ocean sink to dry deposition, and doesn't explicitly write it out as a diagnostic
  • HEMCO now lumps eruptive and non-eruptive volcano emissions into one diagnostic (SO2-EV-$)
  • Differences in aircraft and volcano emissions w/r/t v10-01d are likely introduced by the MESSy vertical regridding routines.
  • Differences in ISOP, ACET, DMS, and other key biogenic-emitting species are likely due to the change in how MEGAN uses leaf-area-index data w/r/t v10-01d.
  • ACET and DMS emissions now use a different air-sea exchange module (in HEMCO) w/r/t v10-01d.
  • The decrease in NH3 over SE Asia may be related to the Streets NH3 inventory.
Additional or summary comments:
  • GEOS-Chem v10-01d and prior versions use a klunky emissions interface. There was a significant amount of hardwiring, especially in the application of scale factors and masks. In v10-01e, HEMCO has completely replaced this legacy emissions interface. We have tried to replicate the emissions output r of v10-01d to the greatest extent possible. But in v10-01e+HEMCO, we have also brought in a couple of newer inventories (volcanoes, etc.) while also introducing a new vertical regridding scheme. So we will not be able to replicate the v10-01d emissions to 100% accuracy.
  • As of this benchmark, the following simulations are not yet fully compatible with HEMCO.
  • The APM aerosol microphysics will have to be completely re-integrated once v10-01 ships.
Approval
Requires further investigation: No
Approved by: Bob Y. and Christoph Keller
Date of approval: 06 Nov 2014

--Melissa Sulprizio 06:23, 6 November 2014 (EST)
--Bob Y. 10:13, 6 November 2014 (EST)

v10-01d

We performed two 1-month benchmark simulations for v10-01d:

  1. v10-01d_trop: 1-month benchmark with tropopsheric chemistry
  2. v10-01d_UCX: 1-month benchmark with UCX strat-trop chemistry (72 vertical levels)

1-month benchmark v10-01d with UCX chemistry

Here is the assessment form for 1-month benchmark simulation v10-01d with UCX chemistry (aka v10-01d_UCX).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • ISOPO2 isomerization reaction fix: Ploy Achakulwisut (Harvard)
  • FAST-JX v7.0 update: Sebastian Eastham (MIT), Jingqiu Mao (Princeton)
  • NetCDF define mode: GEOS-Chem Support Team
  • Nested grid parallelization fix: Jintai Lin (Peking U.)
  • ND44 fix for sea salt: Kateryna Lapina (CU Boulder)
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Mon Jun 2 20:00:41 EDT 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-08.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 5:52
  • CPU time/ wall time: 6.8095
  • % of ideal performance: 85.12%
Compared to previous benchmark: v10-01c with UCX chemistry
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01d.tar.gz
(in directory v10-01d_UCX)
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01d.tar.gz
(in directory v10-01d_UCX)
Metrics
Global mean OH (from log file): 13.0813437885412 x 105 molec/cm3
Methyl chloroform lifetime: 4.7621 years
Did either of these change by more than 5%? No. The mean OH differs by -0.25%, and the MCF lifetime differs by 0.34%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ISOP, H2O2, MVK, MACR, PMN, PRPE, CH2O, N2O5, HNO4, MP, DMS, NIT, DST4, Br, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, Cl, OClO, Cl2O2, HO2
Comments on SURFACE differences:
  • HAC shows large increases at the surface. We attribute this to the fix for J(HAC) added to this version.
  • The other species listed here (particularly ISOP and its oxidation products) show large percent differences. But these occur in places where the concentrations of these species are already very low. We may attribute these differences to small-number statistics.
At 500 hPa, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, PRPE, CH2O, N2O5, DMS, NH3, NIT, DST4, BrNO2, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, Cl, OClO, Cl2O2
Comments on 500 hPa differences:
  • HAC shows large increases at 500 hPa. We attribute this to the fix for J(HAC) added to this version.
  • The other species listed here (particularly ISOP and its oxidation products) show large percent differences. But these occur in places where the concentrations of these species are already very low. We may attribute these differences to small-number statistics.
In the ZONAL MEAN differences, list all species that changed by 10% or more: PAN, ALK4, ISOP, MVK, MACR, PMN, PRPE, C3H8, N2O5, SO4s, NH3, NIT, NITs, DST2, DST3, SALC, Br, BrNO3, ISOPN, MOBA, HAC, MMN, RIP, IEPOX, MAP, OCS, CCl4, CH3CCl3, CFC11, H1211, H1301, H2402, Cl, ClNO2, Cl2O2
Comments on ZONAL MEAN differences:
  • HAC shows large zonal mean differences. We attribute this to the fix for J(HAC) added to this version.
  • Several species (particularly ISOP and its oxidation products) show large percent differences but small (~0.1 ppbv) absolute differences.
  • The "numerical noise" patterns in the NH3 and NIT percent differences may be attributed to ISORROPIA II.
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
  • None needed.
Additional or summary comments:
  • This benchmark was done to test a few last minute-fixes into the code before integrating HEMCO into GEOS-Chem. HEMCO will be tested with benchmark simulation v10-01e.
  • The fix for the ND44 drydep diagnostic described above only affects simulations done where the non-local PBL mixing is turned OFF. In this benchmark (and all GEOS-Chem benchmarks), we use the non-local PBL mixing scheme.
Approval
Requires further investigation: NO
Approved by: Melissa Sulprizio and Bob Yantosca
Date of approval: 03 Jun 2014

--Melissa Sulprizio 14:47, 2 June 2014 (EDT)
--Bob Y. 11:39, 3 June 2014 (EDT)

1-month benchmark v10-01d with tropospheric chemistry

Here is the assessment form for 1-month benchmark simulation v10-01d with tropospheric chemistry (aka v10-01d_trop).

Description
New features added into GEOS-Chem:

Features affecting the full-chemistry simulation in this benchmark:

Features not affecting the full-chemistry simulation in this benchmark:

Developer name(s) and institution(s):
  • ISOPO2 isomerization reaction fix: Ploy Achakulwisut (Harvard)
  • FAST-JX v7.0 update: Sebastian Eastham (MIT), Jingqiu Mao (Princeton)
  • NetCDF define mode: GEOS-Chem Support Team
  • Nested grid parallelization fix: Jintai Lin (Peking U.)
  • ND44 fix for sea salt: Kateryna Lapina (CU Boulder)
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Mon Jun 16 13:24:57 EDT 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-08.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 2:47
  • CPU time/wall time: 6.8980
  • % of ideal performance: 86.23%
Compared to previous benchmark: v10-01c with tropospheric chemistry
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01d.tar.gz
(in directory v10-01d_trop)
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01d.tar.gz
(in directory v10-01d_trop)
Metrics
Global mean OH (from log file): 13.5883620833673 x 105 molec/cm3
Methyl chloroform lifetime: 4.7516 years
Did either of these change by more than 5%? No. The mean OH differs by -0.29%, and the MCF lifetime differs by 0.35%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ISOP, H2O2, MVK, MACR, PMN, PRPE, CH2O, N2O5, HNO4, MP, DMS, NIT, DST4, Br, BrO, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP
Comments on SURFACE differences:
  • HAC shows large increases at the surface. We attribute this to the fix for J(HAC) added to this version.
  • The other species listed here (particularly ISOP and its oxidation products) show large percent differences. But these occur in places where the concentrations of these species are already very low. We may attribute these differences to small-number statistics.
At 500 hPa, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, PRPE, CH2O, N2O5, DMS, NH3, NIT, DST4, BrNO2, BrNO3, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP
Comments on 500 hPa differences:
  • HAC shows large increases at 500 hPa. We attribute this to the fix for J(HAC) added to this version.
  • The other species listed here (particularly ISOP and its oxidation products) show large percent differences. But these occur in places where the concentrations of these species are already very low. We may attribute these differences to small-number statistics.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, ALK4, ISOP, MVK, MACR, PMN, PRPE, C3H8, N2O5, SO4s, NH3, NIT, NITs, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP
Comments on ZONAL MEAN differences:
  • HAC shows large zonal mean differences. We attribute this to the fix for J(HAC) added to this version.
  • Several species (particularly ISOP and its oxidation products) show large percent differences but small (~0.1 ppbv) absolute differences.
  • The "numerical noise" patterns in the NH3 and NIT percent differences may be attributed to ISORROPIA II.
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
  • None needed.
Additional or summary comments:
  • This benchmark was done to test a few last minute-fixes into the code before integrating HEMCO into GEOS-Chem. HEMCO will be tested with benchmark simulation v10-01e.
  • The fix for the ND44 drydep diagnostic described above only affects simulations done where the non-local PBL mixing is turned OFF. In this benchmark (and all GEOS-Chem benchmarks), we use the non-local PBL mixing scheme.
Approval
Requires further investigation: NO
Approved by: Melissa Sulprizio and Bob Yantosca
Date of approval: 03 Jun 2014

--Melissa Sulprizio 14:47, 2 June 2014 (EDT)
--Bob Y. 11:39, 3 June 2014 (EDT)

v10-01c

Because GEOS-Chem v10-01c was the version where we introduced the UCX chemistry mechanism, we performed two 1-month benchmark simulations:

  1. v10-01c_trop: 1-month benchmark with tropopsheric chemistry
  2. v10-01c_UCX: 1-month benchmark with UCX strat-trop chemistry (72 vertical levels)

--Bob Y. 13:15, 30 May 2014 (EDT)

Final recommendation for J-values

We were not able to include this final recommendation for J(HAC) and J(PAN) into the v10-01c_trop and v10-01c_UCX 1-month benchmarks. We will include these in the 1-year benchmarks for v10-01c.

--Bob Y. 17:16, 30 May 2014 (EDT)

1-month benchmark v10-01c with UCX chemistry

Here is the assessment form for 1-month benchmark simulation v10-01c with UCX chemistry (aka v10-01c_UCX).

Description
New features added into GEOS-Chem:

Features affecting the chemistry in this benchmark:

Features not affecting the chemistry in this benchmark:

Developer name(s) and institution(s):
Version, resolution, met fields used: v10-01, GEOS-FP (72L), 4x5, July 2013
1-month benchmark finished on: Thu May 29 00:01:33 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-09.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 5:51
  • CPU time / wall time: 6.7089
  • % of ideal performance: 83.86%
Compared to previous benchmark: v10-01c with tropospheric chemistry
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01c/v10-01c_UCX.tar.gz
(in directory v10-01c_UCX)
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01c.tar.gz
(in directory v10-01c_UCX)
Metrics
Global mean OH (from log file): 13.1136815995777 x 105 molec/cm3
Methyl chloroform lifetime: 4.7459 years
Did either of these change by more than 5%? No. The mean OH differs by -3.77%, and the MCF lifetimes differ by 0.23%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, HNO3, H2O2, MEK, RCHO, MVK, MACR, PMN, PPN, PRPE, N2O5, HNO4, MP, DMS, SO4s, NH3, NIT, NITs, DST4, Br2, Br, BrO, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, NO2, NO3, HNO2, OH, HO2
Comments on SURFACE differences:
  • Differences in the ratio plots of ISOP and its oxidation products over the oceans probably reflect small number differences.
At 500 hPa, list all species that changed by 10% or more: NO, ALK4, ISOP, HNO3, MEK, RCHO, MVK, MACR, PMN, PRPE, C3H8, N2O5, HNO4, MP, DMS, SO2, NH3, NIT, NITs, DST4, Br2, Br, BrO, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, NO2, NO3, HNO2, OH, HO2
Comments on 500 hPa differences:
  • The "noise" pattern in the NIT and NH3 plots reflect the numerical drift caused by ISORROPIA II
In the ZONAL MEAN differences, list all species that changed by 10% or more: All species, except for MSA and OCPI, changed by 10% or more.
Comments on ZONAL MEAN differences:
  • It is understandable that we would see differences of this magnitude in these species given that we have activated stratospheric chemistry. We are also using the full 72-layer grid as opposed to the 47-layer grid in the tropopause-only simulation.
  • The following species show small absolute differences but large percent differences:
    • ALK4, ISOP, MEK, ALD2, RCHO, MVK, MACR, PMN, R4N2, PRPE, C3H8, CH2O, C2H6, SO2, SO4s, MSA, NH3, NH4, NITs, BCPO, OCPO, DST1, DST2, DST3, DST4, SALA, SALC, BrNO2, CHBr3, CH2Br2, ISOPN, HAC, GLYC, MMN, HNO2
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
  • NOx sources (anthropogenic, aircraft, lightning, soils) changed very slightly in this run. This may be an artifact of comparing the data on the full 72-layer grid to the 47-layer grid of the previous benchmark.
  • The acetone ocean sink also decreased slightly (-0.000638 Tg C). This may also be an artifact of comparing the 72-level grid to the 47-level grid.
Additional or summary comments:
  • This simulation uses the KPP chemistry solver, while all past benchmarks have used SMVGEAR. UCX simulations are extremely slow when SMVGEAR is used, so KPP is the recommended solver.
  • Many of the structural updates (i.e. to speed up GEOS-Chem, to reduce the memory footprint, to fix issues in the specialty simulations) do not impact the full-chemistry simulation. This was validated with unit tests and difference tests.
  • IMPORTANT: We were unable to add the Final recommendation for J-values into this 1-month benchmark. We will add this into the 1-year benchmarks for v10-01c.
Approval
Requires further investigation: NO
Approved by: Daniel Jacob, Jingqiu Mao, Sebastian Eastham
Date of approval: 29 May 2014

--Bob Y. 13:11, 30 May 2014 (EDT)

1-month benchmark v10-01c with tropospheric chemistry

Here is the assessment form for 1-month benchmark simulation v10-01c with tropospheric chemistry (i.e. UCX off, aka v10-01c_trop).

Description
New features added into GEOS-Chem:

Features affecting the chemistry in this benchmark:

Features not affecting the chemistry in this benchmark:

Developer name(s) and institution(s):
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Wed May 28 17:57:45 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-09.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 2:49
  • Scalability: 6.7835
Compared to previous benchmark: v10-01b
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: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01c.tar.gz (in v10-01c_trop directory)
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01c.tar.gz (in v10-01c_trop directory)
Metrics
Global mean OH (from log file): 13.6273583437531 x 105 molec/cm3
Methyl chloroform lifetime: 4.7349 years
Did either of these change by more than 5%? No. The mean OH differs by -0.77%, and the MCF lifetimes differ by 1.37%.
At the SURFACE, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, H2O2, MEK, RCHO, MVK, MACR, PMN, PPN, PRPE, CH2O, N2O5, HNO4, MP, DMS, NIT, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO3, HNO2, OH, HO2
Comments on surface differences:
  • Most of the differences in ISOP and its products are over the oceans. This could be due to small number differences, since there is not a lot of ISOP in these locations.
  • We believe that the following differences are due to photolysis (FAST-JX in v10-01c_trop vs. FAST-J in v10-01b):
    • Differences in NO at the south pole
    • Differences in bromine species, particularly at the poles
    • Differences observed in DST4
At 500 hPa, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, H2O2, ACET, MEK, MVK, MACR, PMN, PRPE, CH2O, N2O5, HNO4, MP, DMS, SO2, NH3, NIT, DST4, Br2, Br, BrO, BrNO2, BrNO3, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2
Comments on 500 hPa differences:
  • Most of the differences in ISOP and its products are over the oceans. This could be due to small number differences, since there is not a lot of ISOP in these locations.
  • We believe that the following differences are due to photolysis (FAST-JX in v10-01c_trop vs. FAST-J in v10-01b):
    • Differences in NO at the south pole
    • Differences in bromine species, particularly at the poles
    • Differences observed in DST4
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, PAN, ALK4, ISOP, H2O2, ACET, MEK, MVK, MACR, PMN, PRPE, C3H8, CH2O, N2O5, MP, DMS, SO4s, NH3, NIT, NITs, Br2, Br, BrO, HOBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO3, HNO2, OH, HO2
Comments on ZONAL MEAN differences:
  • The following species show large percent differences but small absolute differences:
    • NO, PAN, ALK4, ISOP, MEK, MVK, MACR, PMN, R4N2, PRPE, CH2O, C3H8 (at top of atmosphere), N2O5, DMS, SO4s (at top of atmosphere), NITs (at top of atmosphere), DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, CHBr3, ISOPN, MOBA, NO2, NO3, HNO2,
  • Differences in ACET are probably caused by v10-01c having the correct pressure dependency for J(ACET) but v10-01b not having this fix applied.
  • Differences in NH3 and NIT are due to the numerical instability in ISORROPIA II.
  • Differences in HAC and GLYC are due to photolysis (FAST-JX vs. FAST-J).
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
  • The NO anthropogenic emissions only very slightly higher in v10-01c_trop (by 0.000343 Tg N). This is most likely due to the fix for FAST-JX wavelength bins propagating through to the PARANOX ship emissions plume model.
  • The ACET ocean sink is only slightly smaller (-0.106) in v10-01c_trop.
Additional or summary comments:
  • Many of the structural updates (i.e. to speed up GEOS-Chem, to reduce the memory footprint, to fix issues in the specialty simulations) do not impact the full-chemistry simulation. This was validated with unit tests and difference tests.
  • IMPORTANT: We were unable to add the Final recommendation for J-values into this 1-month benchmark. We will add this into the 1-year benchmarks for v10-01c.
Approval
Requires further investigation: NO
Approved by: Daniel Jacob, Jingqiu Mao, Sebastian Eastham
Date of approval: 29 May 2014

--Bob Y. 14:20, 30 May 2014 (EDT)

v10-01b

Here is the assessment form for 1-month benchmark simulation v10-01b.

Description
New features added into GEOS-Chem:
Developer name(s) and institution(s):
  • PROPNN molecular weight: Jenny Fisher (U. Wollongong)
  • TOMS O3 regridding: Jintai Lin (Peking U.)
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Wed Mar 5 15:06:00 EST 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-11.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 2:55
  • Scalability: 6.7949
Compared to previous benchmark: v10-01a
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):
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01b.tar.gz
Metrics
Global mean OH (from log file): 13.7339945052522 x 105 molec/cm3
Methyl chloroform lifetime: 4.6706 years
Did either of these change by more than 5%? No. Both the mean OH and MCF lifetimes differ by < 0.01%.
At the SURFACE, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, NIT, DST4, ISOPN, MOBA, RIP
Comments on SURFACE differences:
  • Differences in most species are due to small number differences over the oceans or poles, where concentrations are small.
  • Differences in NIT can be attributed to numerical noise from ISORROPIA.
At 500 hPa, list all species that changed by 10% or more: NO, ISOP, NH3, NIT, DST4, ISOPN, MOBA, RIP
Comments on 500 hPa differences:
  • Differences in most species are due to small number differences over the oceans or poles, where concentrations are small.
  • Differences in NIT can be attributed to numerical noise from ISORROPIA.
In the ZONAL MEAN differences, list all species that changed by 10% or more: ALK4, ISOP, C3H8, SO4s, NH3, NIT, GLYC
Comments on ZONAL MEAN differences:
  • Differences in most species are due to small number differences above ~200 hPa, where concentrations are small.
  • Differences in NIT can be attributed to numerical noise from ISORROPIA.
In the EMISSION RATIO maps, list all species that changed by 10% or more: None
Comments on EMISSION RATIO differences:
Additional or summary comments:
  • Differences caused by the TOMS O3 update are not seen in the benchmark output. The benchmark simulation uses O3 columns from the GEOS-FP met fields, so TOMS O3 columns are not used.
Approval
Requires further investigation: No
Approved by: Jenny Fisher, Fabien Paulot
Date of approval: 06 Mar 2014

--Melissa Sulprizio 11:34, 6 March 2014 (EST)

v10-01a

Here is the assessment form for 1-month benchmark simulation v10-01a.

Description
New features added into GEOS-Chem:
Developer name(s) and institution(s):
  • Patrick Kim (Harvard)
Version, resolution, met fields used: v10-01, GEOS-FP (47L), 4x5, July 2013
1-month benchmark finished on: Mon Feb 3 15:05:32 EST 2014
Performance statistics:
  • Ran on 8 CPUs of bench@titan-09.as.harvard.edu (2.659 GHz x 8 CPU)
  • Wall time: 2:48
  • Scalability: 6.8703
Compared to previous benchmark: v9-02r w/ GEOS-FP meteorology
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):
Plots may be viewed at: ftp://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v10-01/v10-01a.tar.gz
Metrics
Global mean OH (from log file): 13.7338316128900 x 105 molec/cm3
Methyl chloroform lifetime: 4.6707 years
Did either of these change by more than 5%? No. The mean OH differs by -0.02%, and the MCF lifetimes differ by 0.07%.
At the SURFACE, list all species that changed by 10% or more: NO, O3, PAN, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, N2O5, DMS, NH3, NH4, NIT, OCPI, DST4, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, PROPNN, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO3, OH
Comments on SURFACE differences:
  • Differences in isoprene related tracers can be attributed to changing biogenic emissions because we now force the MODIS LAI year to 2008.
  • Differences in NO, DST4, and the bromine species are due to small number differences where concentrations are low.
  • Differences in NH4 and NIT can be attributed to numerical noise from ISORROPIA.
At 500 hPa, list all species that changed by 10% or more: NO, ISOP, MVK, MACR, PMN, PRPE, N2O5, DMS, NH3, NIT, DST4, Br2, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, NO3
Comments on 500 hPa differences:
  • Differences in NO, DST4, and Br2 are due to small number differences where concentrations are low.
  • Differences in NH3 and NIT can be attributed to numerical noise from ISORROPIA.
In the ZONAL MEAN differences, list all species that changed by 10% or more: NO, ALK4, ISOP, MVK, MACR, PMN, PRPE, C3H8, N2O5, NH3, NIT, NITs, DST4, Br2, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX
Comments on ZONAL MEAN differences:
  • Differences in isoprene related tracers can be attributed to changing biogenic emissions because we now force the MODIS LAI year to 2008.
  • Differences in NO, ALK4, C3H8, N2O5, NITs, DST4, Br2, are due to small number differences where concentrations are low.
  • Differences in NH3 and NIT can be attributed to numerical noise from ISORROPIA.
In the EMISSION RATIO maps, list all species that changed by 10% or more: ISOP, ACET, OC, PRPE, CO from monoterpenes
Comments on EMISSION RATIO differences:
  • Biogenic emissions differ because we now force the use of MODIS LAI for 2008 if the simulation year is beyond 2008. MODIS LAI for 2009 was used in the benchmark v9-02r w/ GEOS-FP met, but there are large differences in the 2009 file that still need to be investigated.
Additional or summary comments: We still need to determine what caused the changes in the MODIS LAI input files. As per our earlier discussion, we have put in an error trap in drydep_mod to make sure that the changes are only made for simulations using GEOS-FP and the Olson 2001 land map.
Approval
Requires further investigation: No
Approved by: Patrick Kim
Date of approval: 5 Feb 2014

--Melissa Sulprizio 12:16, 3 February 2014 (EST)