Difference between revisions of "GEOS-Chem v11-01 benchmark history"
(→v11-01b-Run0) |
(→v11-01b-Run0) |
||
Line 88: | Line 88: | ||
--[[User:Melissa Payer|Melissa Sulprizio]] ([[User talk:Melissa Payer|talk]]) 20:24, 11 August 2015 (UTC) | --[[User:Melissa Payer|Melissa Sulprizio]] ([[User talk:Melissa Payer|talk]]) 20:24, 11 August 2015 (UTC) | ||
+ | |||
+ | ==== Comments about the 1-year benchmark v11-01b-Run0 ==== | ||
+ | |||
+ | ===== Aerosol differences ===== | ||
+ | |||
+ | '''''Colette Heald wrote:''''' | ||
+ | |||
+ | :I took a look at the aerosols in the benchmark. There's one thing I don't quite understand: why are the differences between v11-01b and v10-01i so much larger than the differences between v11-01b and v10-01-public at the surface for NIT, NH4, OCPI (see for example the difference plots for Jan or Jul)? According to the table that Melissa linked to the only difference between the two v10-01 runs are bug fixes and the inclusion of UCX and SOA. The NIT/NH4 may be a knock-on effect from changes in SO4 & HNO3, but I'm at a loss for why OCPI changes so significantly. Perhaps I've misunderstood the changes included those simulations? | ||
+ | |||
+ | '''''Melissa Sulprizio wrote:''''' | ||
+ | |||
+ | :The large differences in OCPI are due to the fact that biogenic OC emissions are not calculated in the SOA chemistry mechanism. Before HEMCO, we had the following code in carbon_mod.F: | ||
+ | |||
+ | IF ( LSOA ) THEN | ||
+ | |||
+ | ! Total HYDROPHILIC OC source [kg] | ||
+ | ! (Don't use archived TERP_ORGC if LSOA=T) | ||
+ | OCSRC(I,J,1) = ANTH_ORGC(I,J,1) + | ||
+ | & BIOF_ORGC(I,J,1) + | ||
+ | & BIOB_ORGC(I,J,1) | ||
+ | |||
+ | ELSE | ||
+ | |||
+ | ! Total HYDROPHILIC OC source [kg] | ||
+ | ! (Use archived TERP_ORGC for if LSOA=F) | ||
+ | OCSRC(I,J,1) = ANTH_ORGC(I,J,1) + | ||
+ | & BIOF_ORGC(I,J,1) + | ||
+ | & BIOB_ORGC(I,J,1) + | ||
+ | & TERP_ORGC(I,J) | ||
+ | |||
+ | ENDIF | ||
+ | |||
+ | Christoph Keller has treated OCPI in the same manner in HEMCO. If SOA is turned in in GEOS-Chem v10-01 and later versions, HEMCO will not calculate biogenic OCPI emissions. | ||
+ | |||
+ | Many of the other differences between v11-01b and v10-01i can be attributed to last-minute fixes that went into v10-01 prior to the public release (see the complete [[GEOS-Chem_v10-01#Fixes_and_updates_added_during_the_public_comment_period|table of fixes and updates]]). In particular, the following fixes may have contributed to the differences we’re seeing: | ||
+ | |||
+ | #[[GEOS-Chem_v10-01#Last-minute_fixes_prior_to_the_GEOS-Chem_v10-01_provisional_release|Remove Russia from MIX Asia mask file]] | ||
+ | #[[GEOS-Chem_v10-01#Switching_to_the_officially_released_GFED4_data_files|Switching to the officially released GFED4 data files]] | ||
+ | #[[GEOS-Chem_v10-01#Bug_fix_in_I3_field_interpolation|Bug fix in I3 field interpolation]] (affected temperature-dependent emissions) | ||
+ | |||
+ | '''''Jeff Pierce wrote:''''' | ||
+ | |||
+ | :Thanks Melissa. Just to be clear, the SOA mass goes to to the OCPI tracer when the interactive SOA is turned off, but SOA mass goes to other, SOA-specific tracers when the interactive SOA is turned on. | ||
+ | |||
+ | '''''Colette Heald wrote:''''' | ||
+ | |||
+ | :Thanks for the clarifications Melissa. I believe then, that other than the [[#Dust differences|dust issue]] that Li has flagged up, the aerosols look fine for v11.1 | ||
+ | |||
+ | ===== Dust differences ===== | ||
+ | |||
+ | '''''Li Zhang wrote:''''' | ||
+ | |||
+ | :After checking the results, I found that there are some problems with the Dust3 (DST3) concentrations both at the surface and 500 hPa. It is not only different to the 1-month benchmarks that Lizzie sent me, but also different to the results of our simulations based on GEOS-Chem adjoint code. Both in Lizzie's and my results, that the DST3 concentrations are larger after using our dust scheme. | ||
+ | :The DST3 should be supposed to increase, especially over the dust source regions after applying our improved dust scheme. From the dust emission map of DST3, I did see this increase in this benchmarks. However I am surprised that it shows decrease in the concentrations both at the surface and 500 hPa. I am not sure it is due to the coding or only the plotting mistake. It would be great that someone can double check it. | ||
+ | |||
+ | '''''Melissa Sulprizio wrote:''''' | ||
+ | |||
+ | :The dust differences that we see in the 1-year benchmark plots for v11-01b are a combination of the following updates: | ||
+ | #[[Mineral_dust_aerosols#Improved_dust_size_distribution_scheme|Improved dust size distribution scheme]] | ||
+ | #[[Mineral_dust_aerosols#Now_treat_DST2-DST4_as_coarse_mode_in_wet_scavenging|Now treat DST2-DST4 as coarse mode in wet scavenging]] | ||
+ | :The second item was identified by Duncan Fairley when we implemented his acid uptake on dust scheme (default off) into the standard code. | ||
+ | :The improved dust size distribution scheme decreased DST1 and DST2, while it increased DST3 and DST4. The wet scavenging fix to dust decreased DST2 and DST3. | ||
+ | |||
+ | '''''Colette Heald wrote:''''' | ||
+ | |||
+ | :Thanks for the clarification. I can see that this makes sense given the changes. I discussed this with Dave, and I'll just note here (more for the benefit of Jeff and Peter) that the net effect of these changes does appear to be moving us in the wrong direction. The model already underestimates dust export/transport and these further reductions will exacerbate that. Something for us to keep an eye out in the future... | ||
+ | |||
+ | '''''David Ridley wrote:''''' | ||
+ | |||
+ | :We’ve mainly been focusing on removal in the outflow from Africa, but there seems to be too much removal in general based on the gradient in dust AOD away from source. However, I’m finding a general lack of dust from Asia in GC too, which is contrary to some previous work. Much of low bias we’re seeing (relative to satellite) comes from earlier and later in the year than the Wang et al. (2012) and Ku et al. (2011) studies that I know of – although the latter study shifts dust emissions around, rather than decreasing them substantially. We’ve been looking over a period of several years, so it is tricky to relate that to the above studies that focus on periods of days to weeks. | ||
+ | :It sounds like Duncan’s fix makes sense, but it does highlight potential issues with emission and removal of dust in the model. | ||
+ | |||
+ | --[[User:Melissa Payer|Melissa Sulprizio]] ([[User talk:Melissa Payer|talk]]) 13:46, 19 August 2015 (UTC) | ||
== 1-month benchmarks == | == 1-month benchmarks == |
Revision as of 13:46, 19 August 2015
On this page we have posted complete information about all benchmark simulations (both 1-month and 1-year) for GEOS-Chem v11-01.
Contents
1-year benchmarks
v11-01b-RnPbBe
A 1-year Rn-Pb-Be simulation was performed using GEOS-Chem v11-01b. 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:
You may view the benchmark plots for the simulation by pointing your browser to:
http://ftp.as.harvard.edu/gcgrid/geos-chem/1yr_benchmarks/v11-01/v11-01b/RnPbBe/output/
--Melissa Sulprizio (talk) 20:26, 20 July 2015 (UTC)
v11-01b-Run0
Three GEOS-Chem model versions were compared to each other:
Color | Quantity Plotted | Met Type | Year | Emissions | Chemistry mechanism | Wet deposition | Annual Mean OH [105 molec/cm3] |
---|---|---|---|---|---|---|---|
Red | v10-01i-Run0 | GEOS-FP, 72L, 4x5 |
2013 | HEMCO emissions component fixes:
|
UCX chemistry mechanism | Same as v9-01-03e-Run0 | 11.125 |
Green | v10-01-public-release-Run0 | GEOS-FP, 72L, 4x5 |
2013 | " " | Benchmark chemistry mechanism, includes | " " | 11.723 |
Blue | v11-01b-Run0 | GEOS-FP, 72L, 4x5 |
2013 | + Update DMS climatology to Lana + Improved dust size distribution scheme |
+ Density of OA update + Update of PMN + O3 reaction products in globchem.dat file + Bug fix for black carbon in ucx_mod.F |
+ Impaction scavenging for hydrophobic BC + Homogeneous IN removal + Now treat DST2-DST4 as coarse mode in wet scavenging |
12.000 |
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/v11-01/v11-01b/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/v11-01/v11-01b/Run0/output/
--Melissa Sulprizio (talk) 20:24, 11 August 2015 (UTC)
Comments about the 1-year benchmark v11-01b-Run0
Aerosol differences
Colette Heald wrote:
- I took a look at the aerosols in the benchmark. There's one thing I don't quite understand: why are the differences between v11-01b and v10-01i so much larger than the differences between v11-01b and v10-01-public at the surface for NIT, NH4, OCPI (see for example the difference plots for Jan or Jul)? According to the table that Melissa linked to the only difference between the two v10-01 runs are bug fixes and the inclusion of UCX and SOA. The NIT/NH4 may be a knock-on effect from changes in SO4 & HNO3, but I'm at a loss for why OCPI changes so significantly. Perhaps I've misunderstood the changes included those simulations?
Melissa Sulprizio wrote:
- The large differences in OCPI are due to the fact that biogenic OC emissions are not calculated in the SOA chemistry mechanism. Before HEMCO, we had the following code in carbon_mod.F:
IF ( LSOA ) THEN
! Total HYDROPHILIC OC source [kg] ! (Don't use archived TERP_ORGC if LSOA=T) OCSRC(I,J,1) = ANTH_ORGC(I,J,1) + & BIOF_ORGC(I,J,1) + & BIOB_ORGC(I,J,1)
ELSE
! Total HYDROPHILIC OC source [kg] ! (Use archived TERP_ORGC for if LSOA=F) OCSRC(I,J,1) = ANTH_ORGC(I,J,1) + & BIOF_ORGC(I,J,1) + & BIOB_ORGC(I,J,1) + & TERP_ORGC(I,J)
ENDIF
Christoph Keller has treated OCPI in the same manner in HEMCO. If SOA is turned in in GEOS-Chem v10-01 and later versions, HEMCO will not calculate biogenic OCPI emissions.
Many of the other differences between v11-01b and v10-01i can be attributed to last-minute fixes that went into v10-01 prior to the public release (see the complete table of fixes and updates). In particular, the following fixes may have contributed to the differences we’re seeing:
- Remove Russia from MIX Asia mask file
- Switching to the officially released GFED4 data files
- Bug fix in I3 field interpolation (affected temperature-dependent emissions)
Jeff Pierce wrote:
- Thanks Melissa. Just to be clear, the SOA mass goes to to the OCPI tracer when the interactive SOA is turned off, but SOA mass goes to other, SOA-specific tracers when the interactive SOA is turned on.
Colette Heald wrote:
- Thanks for the clarifications Melissa. I believe then, that other than the dust issue that Li has flagged up, the aerosols look fine for v11.1
Dust differences
Li Zhang wrote:
- After checking the results, I found that there are some problems with the Dust3 (DST3) concentrations both at the surface and 500 hPa. It is not only different to the 1-month benchmarks that Lizzie sent me, but also different to the results of our simulations based on GEOS-Chem adjoint code. Both in Lizzie's and my results, that the DST3 concentrations are larger after using our dust scheme.
- The DST3 should be supposed to increase, especially over the dust source regions after applying our improved dust scheme. From the dust emission map of DST3, I did see this increase in this benchmarks. However I am surprised that it shows decrease in the concentrations both at the surface and 500 hPa. I am not sure it is due to the coding or only the plotting mistake. It would be great that someone can double check it.
Melissa Sulprizio wrote:
- The dust differences that we see in the 1-year benchmark plots for v11-01b are a combination of the following updates:
- The second item was identified by Duncan Fairley when we implemented his acid uptake on dust scheme (default off) into the standard code.
- The improved dust size distribution scheme decreased DST1 and DST2, while it increased DST3 and DST4. The wet scavenging fix to dust decreased DST2 and DST3.
Colette Heald wrote:
- Thanks for the clarification. I can see that this makes sense given the changes. I discussed this with Dave, and I'll just note here (more for the benefit of Jeff and Peter) that the net effect of these changes does appear to be moving us in the wrong direction. The model already underestimates dust export/transport and these further reductions will exacerbate that. Something for us to keep an eye out in the future...
David Ridley wrote:
- We’ve mainly been focusing on removal in the outflow from Africa, but there seems to be too much removal in general based on the gradient in dust AOD away from source. However, I’m finding a general lack of dust from Asia in GC too, which is contrary to some previous work. Much of low bias we’re seeing (relative to satellite) comes from earlier and later in the year than the Wang et al. (2012) and Ku et al. (2011) studies that I know of – although the latter study shifts dust emissions around, rather than decreasing them substantially. We’ve been looking over a period of several years, so it is tricky to relate that to the above studies that focus on periods of days to weeks.
- It sounds like Duncan’s fix makes sense, but it does highlight potential issues with emission and removal of dust in the model.
--Melissa Sulprizio (talk) 13:46, 19 August 2015 (UTC)
1-month benchmarks
v11-01b
Here is the assessment form for 1-month benchmark simulation v11-01b.
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: | v11-01, GEOS-FP (72L), 4x5, July 2013 |
1-month benchmark finished on: | Mon Jul 27 18:27:44 2015 |
Performance statistics: |
|
Compared to previous benchmark: | v11-0a |
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/v11-01/v11-01b/v11-01b.results.html
|
Plots may be viewed at: | http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v11-01/v11-01b/ |
Metrics | |
Global mean OH (from log file): | 12.6447244306942 x 105 molec/cm3 |
Methyl chloroform lifetime: | 4.9562 years |
Did either of these change by more than 5%? | No. The mean OH differs by 0.57%, and the MCF lifetime differs by -0.73%. |
At the SURFACE, list all species that changed by 10% or more: | NO, PAN, ALK4, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, C3H8, CH2O, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALA, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO2, NO3, HNO2, BrCl, Cl, ClO, HOCl, ClNO3, ClOO, OClO, Cl2, Cl2O2, MTPA, LIMO, MTPO, TSOG1, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, ISOG2, ISOA1, ISOA2, ISOA3, BENZ, TOLU, XYLE, ASOAN, ASOA1, ASOA2, ASOA3, OH, HO2 |
Comments on SURFACE differences: |
|
At 500 hPa, list all species that changed by 10% or more: | NO, ALK4, ISOP, HNO3, H2O2, MEK, ALD2, RCHO, MVK, MACR, PMN, PPN, R4N2, PRPE, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALA, SALC, Br2, Br, BrO, HOBr, HBr, BrNO2, BrNO3, MPN, ISOPN, MOBA, HAC, GLYC, MMN, RIP, IEPOX, MAP, NO3, HNO2, BrCl, HOCl, ClNO2, OClO, Cl2, MTPA, LIMO, MTPO, TSOG1, TSOG0, TSOA1, TSOA2, TSOA3, ISOG1, ISOG2, ISOA1, ISOA2, ISOA3, TOLU, XYLE, ASOG1, ASOG2, ASOAN, ASOA1, ASOA2, ASOA3, OH, HO2 |
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: | ISOP, HNO3, H2O2, MVK, MACR, PMN, PRPE, N2O5, HNO4, MP, DMS, SO2, SO4, SO4s, MSA, NH3, NH4, NIT, NITs, BCPI, OCPI, BCPO, OCPO, DST1, DST2, DST3, DST4, SALA, SALC, Br2, HOBr, HBr, BrNO3, ISOPN, MOBA, MMN, RIP, NO3, HNO2, Cl, HOCl, ClOO, MTPA, LIMO, MTPO, TSOG1, TSOG2, TSOG3, TSOG0, TSOA1, TSOA2, TSOA3, TSOAO, ISOG1, ISOG2, ISOG3, ISOA1, ISOA2, ASOG1, ASOG2, ISOA3, ASOG3, ASOGAN, ASOA1, ASOA2, ASOA3, OH, HO2 |
Comments on ZONAL MEAN differences: |
|
In the EMISSION RATIO maps, list all species that changed by 10% or more: |
|
Comments on EMISSION RATIO differences: |
|
Additional or summary comments: |
|
Approval | |
Requires further investigation: | No |
Approved by: | GCST, Eloise Marais, Tom Breider, Daniel Jacob |
Date of approval: | 04 Aug 2015 |
--Lizzie Lundgren (talk) 19:16, 30 July 2015 (UTC)
--Melissa Sulprizio (talk) 22:01, 30 July 2015 (UTC)
v11-01a
Here is the assessment form for 1-month benchmark simulation v11-01a.
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: | v11-01, GEOS-FP (72L), 4x5, July 2013 |
1-month benchmark finished on: | Thurs Jul 02 00:00:12 EDT 2015 |
Performance statistics: |
|
Compared to previous benchmark: | v10-01-public-release with SOA on |
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): Tracer Unit Conversions, Air Quantities (e.g. box height and air mass), RRTMG |
Unit test results may be viewed at: | http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v11-01/v11-01a/v11-01a.results.html
|
Plots may be viewed at: | http://ftp.as.harvard.edu/gcgrid/geos-chem/1mo_benchmarks/v11-01/v11-01a/ |
Metrics | |
Global mean OH (from log file): | 12.5726451325898 x 105 molec/cm3 |
Methyl chloroform lifetime: | 4.9926 years |
Did either of these change by more than 5%? | No. The mean OH differs by 0.50%, and the MCF lifetime differs by -0.58%. |
At the SURFACE, list all species that changed by 10% or more: | NO, ISOP, HNO3, MVK, MACR, PMN, N2O5, NH4, NIT, NITs, OCPO, DST1, DST2, DST3, DST4, Br, BrO, HOBr, ISOPN, MOBA, GLYC, MMN, RIP, IEPOX, BrCl, Cl, ClO, ClNO2, ClOO, OClO, Cl2, Cl2O2, MTPA, LIMO, MTPO, ASOA2, ASOA3 |
Comments on SURFACE differences: |
|
At 500 hPa, list all species that changed by 10% or more: | NO, ISOP, MVK, MACR, PMN, PRPE, NH3, NIT, OCPO, DST3, DST4, Br, ISOPN, MOBA, GLYC, RIP, IEPOX, BrCl, Cl, ClO, ClNO2, ClOO, OClO, Cl2, Cl2O2, MTPA, LIMO, MTPO |
Comments on 500 hPa differences: |
|
In the ZONAL MEAN differences, list all species that changed by 10% or more: | NO, ALK4, ISOP, HNO3, MVK, MACR, PMN, PRPE, C3H8, N2O5, SO4s, NH3, NIT, NITs, BCPO, OCPO, DST1, DST2, DST3, DST4, SALC, Br, HBr, BrNO3, ISOPN, MOBA, GLYC, MMN, RIP, IEPOX, OCS, BrCl, CCl4, CH3CCl3, CFC11, H1211, H1301, H24O2, Cl, ClO, ClNO2, ClOO, OClO, Cl2, Cl2O2, MTPA, LIMO, MTPO, ISOA2, TOLU, |
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: |
|
Approval | |
Requires further investigation: | July 30, 2015 update: A moisture signature is observed in CO2 concentrations during carbon simulations with zero emissions and constant 370 ppm initial conditions. The moisture signature is expected for wet mixing ratio (v/v total air) but not dry mixing ratio (v/v dry air). This issue is currently under investigation. |
Approved by: | GCST |
Date of approval: | July 7, 2015 |
--Lizzie Lundgren (talk) 17:05, 2 July 2015 (UTC)