Currently unresolved issues in GEOS-Chem

From Geos-chem
Revision as of 20:43, 5 March 2021 by Bmy (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

We have migrated GEOS-Chem bug reports to the Github issue trackers. This page is no longer being updated.


On this page we provide information about known issues (bugs, problems with emissions inventories, etc.) that have not yet been resolved in GEOS-Chem. For a complete list of all bugs and fixes by GEOS-Chem version, please see our Bugs and fixes wiki page. Also see our list of Known issues caused by compiler bugs.

Unresolved issues from GEOS-Chem 12

These issues were first identified in the GEOS-Chem 12 series of versions.

12.1.0

Issue Reported by Affects Status
Error reading restart files when using a fixed emissions year in HEMCO Anthony Y.H. Wong (Boston U.) All simulations RESOLVED.
  • This issue was resolved in HEMCO version 2.1.012.
  • Validated: GEOS-Chem 12.3.0
  • Approved: 01 Apr 2019
Inconsistency in H2402 concentrations in GCHP vs GEOS-Chem Classic GCST GCHP and GEOS-Chem "Classic" full-chemistry simulations. RESOLVED.
  • It was discovered that an input file read by GEOS-Chem "Classic" contained surface volumne mixing ratio data that was interpolated incorrectly. The file also had corrupted timestamps. Re-creating the file resolved the issue.
  • Validated: GEOS-Chem 12.3.2
  • Approved: 02 May 2019

--Bob Yantosca (talk) 16:24, 30 May 2019 (UTC)

12.0.0

Issue Reported by Affects Status
TOMAS simulations cannot use netCDF diagnostics GCST All TOMAS simulations UNRESOLVED.
  • Will require work by the TOMAS team and GCST.
TOMAS40 simulations exit with a floating point exception GCST TOMAS simulations with 40 sized-resolved bins UNRESOLVED.
  • The issue might be caused by the several chemistry and emissions updates that have been recently added into GEOS-Chem.
  • The GCST has referred the error to the TOMAS development team, who will investigate further.
RRTMG does not compile with GNU Fortran GCST All simulations using RRTMG UNRESOLVED.
  • A fix will likely involve an RRTMG software update from AER, Inc. Fixing this is not a GCST priority at this time. Need an interested user from the GC community to take this on.

--Bob Yantosca (talk) 19:04, 3 August 2018 (UTC)

Unresolved issues from GEOS-Chem v11-02

These issues were first identified in GEOS-Chem v11-02{a-f}.

Issue Reported by Affects Status
2x2.5 tropchem simulations encounter KPP integrator errors after the addition of halogen chemistry Katie Travis (MIT) GEOS-Chem simulations at 2x2.5 resolution using the tropchem mechanism UNRESOLVED.
  • As an interim fix, the tolerance has been changed from 2e-1_dp to 1e-2_dp. This prevents a simulation from crashing, but GEOS-Chem does still get convergence errors printed to the log file.

--Melissa Sulprizio (talk) 16:16, 21 March 2018 (UTC)

Unresolved issues from GEOS-Chem v11-01

These issues were first identified in GEOS-Chem v11-01.

Issue Reported by Affects Status
Mass is not conserved when using PRECISION=4 option GCST GEOS-Chem simulations using PRECISION=4 UNRESOLVED.
  • This option may result in less optimal mass conservation than with PRECISION=8, so at this time we will leave it as an option rather than as a standard feature.
Inconsistent mercury simulation diagnostics Bob Yantosca (GCST) Mercury simulations UNRESOLVED.
  • An incomplete code merge has apparently caused some mercury quantities to be archived to the wrong slots of certain diagnostic arrays.
  • GCST will work together with the Hg and POPs Working Group to resolve these issues.
Negative tracer concentrations occur near the poles Lizzie Lundgren (GCST) All GEOS-Chem simulations PARTIALLY RESOLVED:
  • It was noticed that negative tracer concentrations can occur near the poles.
  • Lizzie Lundgren (GCST) implemented a fix that removes the negatives.
    • Validated: v11-01a
    • Approved: 07 Jul 2015
  • But the cause of the negative tracer concentrations is currently unknown. This may be an area of open research rather than an actual bug.
HEMCO errors apparently caused by bugs in the Intel Fortran Compiler GCST All GEOS-Chem simulations PARTIALLY RESOLVED:
  • A bug in the Intel Fortran Compiler optimizer sometimes causes a pointer in the MAP_A2A regridding routines to become unexpectedly deallocated. This bug occurs when compiling with both DEBUG=n BOUNDS=y.
  • The quick fix is to compile with DEBUG=y BOUNDS=y. This will turn off optimization, and thus avoid the error.
Mercury simulation inconsistently failing unit tests Lizzie Lundgren (GCST) Mercury simulations using GEOS-5 meteorology RESOLVED:
  • Bob Yantosca (GCST) found and fixed 2 parallelization errors that resolved this issue.
  • Minor numerical noise differences still occur in slot #4 of the PL-HG2-$ diagnostic for geos5_4x5_Hg and geos5_2x25_Hg unit tests.
    • We can consider these unit tests to have "passed with warnings", because only the diagnostic outputs differ.
  • Fixes to be added in v11-01g

--Bob Yantosca (talk) 18:28, 22 April 2016 (UTC)

Unresolved issues from GEOS-Chem v10-01

These issues were first identified in GEOS-Chem v10-01.

Issue Reported by Affects Status
Cannot compile GEOS-Chem v10-01 with Intel Fortran Compiler v17 GCST All GEOS-Chem v10-01 simulations RESOLVED:
  • The solution involved removing OPTIONAL arguments from subroutines that are contained in INTERFACE blocks.
  • We recommend those who have Intel Fortran 17 to update to GEOS-Chem v11-01, in which this issue is fixed.
  • Validated: v11-01j
  • Approved: 03 Dec 2016
European grid definition bug fix in Tagged O3 simulations Anna Protonatariou (Athens) Tagged O3 simulations RESOLVED.
  • Validated: v11-01d
  • Approved: pending approval
Bug fix for PBLTOP and MTTOP values in tagged ozone simulation Anna Protonatariou (Athens) Tagged O3 simulations RESOLVED.
  • Validated: v11-01d
  • Approved: pending approval
Floating point overflow in MERRA/GEOS-FP convection Viral Shah (U. Washington) Simulations using convection RESOLVED.
  • Validated: v11-01d
  • Approved: pending approval
Fix for reading hourly NEI2011 emissions Viral Shah (U. Washington) Simulations using NEI2011 RESOLVED.
  • Validated: v11-01d
  • Approved: pending approval
Bug fix for offline dust aerosols when UCX is on Seb Eastham (Harvard) Simulations using UCX RESOLVED.
  • Validated: v11-01d
  • Approved: pending approval
Bug fix for monoterpenes in ND46 diagnostic Jared Brewer Full-chem simulations RESOLVED.
  • Validated: v11-01c
  • Approved: 14 Sep 2015
Incorrect computation of moisture in air quantities Lizzie Lundgren (GCST) All GEOS-Chem simulations RESOLVED:
  • Validated: v11-01a
  • Approved: 07 Jul 2015
Convection mass conservation issue for GEOS-FP Lizzie Lundgren (GCST) All simulations done with GEOS-FP RESOLVED:
  • Validated: v11-01a
  • Approved: 07 Jul 2015
Fix bug in RRTMG that prevents storage of O3 and CH4 radiative forcings Lizzie Lundgren (GCST) All simulations done with the RRTMG radiative transfer model RESOLVED:
  • Validated: v11-01a
  • Approved: 07 Jul 2015
Incorrect FAST-JX photolysis rates for species: Jingqiu Mao (Princeton)
Chris Chan Miller (Harvard) 		Full-chemistry, SOA, and aerosol microphysics simulations PARTIALLY RESOLVED.
  • Corrected acetaldehyde photolysis cross sections have now been added.
    • Validated: v10-01f
    • Approved: 13 Jan 2015
  • The carbonyl nitrate photolysis rate remains an active area of open research.

--Bob Yantosca (talk) 21:08, 20 January 2017 (UTC)

Unresolved issues from GEOS-Chem v9-02

These issues were first identified in GEOS-Chem v9-02.

Issue Reported by Affects Status
Potential issue with soil NOx in GCAP simulation GEOS-Chem Support Team Full-chemistry simulations using GCAP meteorology and ship emissions OPEN. We are continuing to look at this issue and hope to have a fix for GEOS-Chem v10-01.
Potential parallelization problems in TOMAS "ND60" diagnostic Sal Farina
GEOS-Chem Support Team 		Simulations using TOMAS aerosol microphysics OPEN. This issue only affects the ND60 diagnostic output itself, but not the tracer concentrations computed by TOMAS. For this reason, we are moving forward with the TOMAS benchmarks for GEOS-Chem v9-02.
Isoprene emissions are lower when using GEOS-FP meteorology Dylan Millet (U. Minnesota) Simulations using biogenic emissions OPEN. This does not necessarily indicate a flaw in GEOS-Chem, but is a matter of active research. It is known that the surface temperatures in GEOS-FP are biased low compared to GEOS-5, which is likely causing this issue.
Soil NOx emissions are lower when using GEOS-FP meteorology Prasad Kasibhatla (Duke) Simulations soil NOx emissions OPEN. This does not necessarily indicate a flaw in GEOS-Chem, but is a matter of active research. It is known that the surface temperatures in GEOS-FP are biased low compared to GEOS-5, which is likely causing this issue.
Need to adjust pulsing threshold when using GEOS-FP meteorology Patrick Kim (Harvard)
Karen Yu (Harvard) 		Simulations soil NOx emissions OPEN. This does not necessarily indicate a flaw in GEOS-Chem, but is a matter of active research. The threshold for pulsing simulations will have to be adjusted so that GEOS-FP nested-grid simulations at 0.25° x 0.3125° yield emissions consistent with the global 4° x 5° and 2° x 2.5° simulations.
O3 completely removed at Shanghai grid box with GEOS-FP 4x5 simulation Patrick Kim (Harvard)
GEOS-Chem Support Team 		Full-chemistry with ship emissions using GEOS-FP 4° x 5° met RESOLVED.
  • Validated: v10-01i
  • Approved: Pending approval
  • Chris Holmes provided several updates for ship NOx chemistry, which include a fix for this issue
  • PARANOX is now implemented as a HEMCO extension. The original PARANOX code has been removed from GEOS-Chem.
Regional biofuel emissions may be neglected when using RETRO or RCP emissions Chris Holmes (UC Irvine) Simulations using global and regional biofuel emissions RESOLVED.
  • Validated: v10-01i
  • Approved: Pending approval
  • All emissions are now done via HEMCO.
  • The code that had been causing this error was removed from GEOS-Chem.
Bug in tagged Hg emissions Carey Friedman (MIT) Tagged Hg simulations RESOLVED.
  • Validated: v10-01i
  • Approved: Pending approval
  • Hg emissions are now done via HEMCO.
  • The code that had been causing this error was removed from GEOS-Chem.
Bug in ship CO emissions Chris Holmes (UC Irvine) Simulations using EMEP ship emissions RESOLVED.
  • Validated: v9-02j
  • Approved: 28 May 2013
Bug fixes in biofuel_mod.F and emfossil.F Christoph Keller (Harvard) Simulations using biofuel and anthropogenic emissions RESOLVED.
  • Validated: v9-02j
  • Approved: 28 May 2013
Ozone over the eastern US is too high when using the Olson 2001 land map Patrick Kim (Harvard)
Natasha Goss (Harvard) 		Full-chemistry simulations RESOLVED.
  • Validated: v10-01a
  • Approved: 05 Feb 2014
Incorrect value of molecular weight for the PROPNN tracer in input.geos Jenny Fisher (Wollongong) Full-chemistry, SOA, and aerosol microphysics simulations. RESOLVED.
  • Validated: v10-01b
  • Approved: 06 Mar 2014
  • For users of GEOS-Chem v9-02, we recommend users update their input.geos files as described in the link at left.
Bug in determining when to use TOMS O3 columns GEOS-Chem Support Team Full-chemistry, SOA, and aerosol microphysics simulations. RESOLVED.
  • Validated: v10-01c
  • Approved: 29 May 2014
Timeseries diagnostics need to be updated to allow for more transported tracers Prasad Kasibhatla (Duke) All simulations, especially full-chemistry RESOLVED.
  • Validated: v10-01c
  • Approved: 29 May 2014
Bug in FAST-J acetone photolysis pressure dependency Jingqiu Mao (Princeton)
GEOS-Chem Support Team 		All full-chemistry simulations (including SOA, dicarbonyl, and aerosol microphysics options) RESOLVED.
Missing error checks for RCP in input_mod.F Jenny Fisher (U. Wollongong) Simulations using the RCP future scenarios RESOLVED.
Bug fixes and updates for the tagged CO simulation Jenny Fisher (U. Wollongong) Tagged CO simulations RESOLVED.
  • Validated: v10-01f
  • Approved: 13 Jan 2015
Avoid seg faults when starting simulations at times other than 00 UTC Chris Holmes (UC Irvine) Simulations using the BRAVO anthro emissions and ARCTAS SO2 ship emissions RESOLVED.
Compilation error in ND63 diagnostic Cenlin He (UCLA) Simulations using the ND63 PARANOX diagnostic, using certain IFORT compiler versions. RESOLVED.

--Bob Y. 18:35, 13 January 2015 (EST)

Unresolved issues from GEOS-Chem v9-01-03

These issues were first identified in GEOS-Chem v9-01-03.

Issue Reported by Affects Status
Issues with Intel Fortran Compiler (IFORT) version 12 GEOS-Chem Support Team Some users have reported issues when migrating from IFORT 11 to IFORT 12. OPEN. Switching compiler versions seems to help.
Negative ship emissions caused by PARANOX plume model Geert Vinken (Eindhoven) Full-chemistry simulations using ship emisisons RESOLVED. Negative O3 values from ship emissions are now treated as a loss via dry deposition, rather than being set to zero.
  • Validated: v9-02l
  • Approved: 26 Jun 2013
Bug fix in hdf_mod.F Sajeev Philip (Dalhousie) Timeseries diagnostics saved out to HDF5 format RESOLVED.
  • Validated: v9-02b
  • Approved: 29 Oct 2012
Bug fixes for tagged CO simulation Jenny Fisher (Wollongong) Tagged CO simulations RESOLVED.
  • Validated: v9-02b
  • Approved: 29 Oct 2012
Bug fix in nei2005_anthro_mod.F Sajeev Philip (Dalhousie) Nested-grid simulations using NEI 2005 emissions RESOLVED.
  • Validated: v9-02b
  • Approved: 29 Oct 2012

--Bob Y. 12:38, 25 April 2014 (EDT)

Unresolved issues from GEOS-Chem v9-01-02

These issues were first identified in GEOS-Chem v9-01-02.

Issue Reported by Affects Status
Numerical noise in ISORROPIA II GEOS-Chem Support Team
Shannon Capps (Georgia Tech) 		Full-chemistry and offline aerosol chemistry simulations OPEN. We are contemplating replacing ISORROPIA II by another package in a future version of GEOS-Chem.
RETRO anthropogenic ethane emissions are too low Bess Corbitt (Harvard) Full chemistry simulations RESOLVED. We have replaced the RETRO ethane emissions with the ethane emissions from Yaping Xiao's offline simulation.
High bias in nighttime ground-level aerosol Sajeev Philip (Dalhousie) All full-chemistry and aerosol simulations using GEOS-5? RESOLVED. Sajeev Philip implemented this fix.
  • Validated: v9-02a
  • Approved: 16 Oct 2012
Mis-indexing bug in emep_mod.F GEOS-Chem Support Team Offline simulations only RESOLVED.
Bug in nei2005_anthro_mod.F GEOS-Chem Support Team Offline simulations only RESOLVED.
Parallelization problem in LINOZ Prasad Kasibhatla (Duke) Full-chemistry and tagged Ox simulations RESOLVED.
  • Validated: v9-02r
  • Approved: 14 Dec 2013
Dependency between dry deposition and soil NOx emissions GEOS-Chem Support Team
Rynda Hudman 		Full-chemistry simulations RESOLVED. This is historical baggage dating back to the earliest versions of GEOS-Chem. The implementation of the new updated soil NOx emissions algorithm has corrected this issue.
  • Validated: v9-02b.
  • Approved: 29 Oct 2012

--Bob Y. 12:36, 25 April 2014 (EDT)

Unresolved issues from GEOS-Chem v9-01-01

These issues were first identified in GEOS-Chem v9-01-01.

Issue Reported by Affects Status
KPP may cause run to exit with error GEOS-Chem Support Team Full-chemistry simulations OPEN. This will likely be resolved when FLEXCHEM is implemented into GEOS-Chem.
Error in partition.f when compiling with IFORT 10 Prasad Kasibhatla (Duke) Full-chemistry simulations OPEN. One solution is to upgrade the compiler version.
Rn-Pb-Be simulation is currently not compatible with non-local PBL mixing scheme Helen Amos (Harvard) Rn-Pb-Be simulations RESOLVED.
  • Validated: v9-02n
  • Accepted: 03 Sep 2013
Wrong tracer names in Rn-Pb-Be simulation can cause drydep diagnostics to be missing GEOS-Chem Support Team Rn-Pb-Be simulations RESOLVED. We added extra tracer names (e.g. Rn222, Pb210, Be7 to the SELECT CASE statement in input_mod.F.

--Bob Y. 12:40, 25 April 2014 (EDT)

Unresolved issues from GEOS-Chem v8-03-02

These issues were first identified in GEOS-Chem v8-03-02.

Issue Reported by Affects Status
Surface ozone over the US is 5-10 ppb too high. Daniel Jacob (Harvard) Full-chemistry simulation Improved treatment of isoprene nitrates chemistry in GEOS-Chem v8-03-02 has resulted in positive bias in the ozone simulation over the US. This is an area of active research in the Oxidants and Chemistry Working Group.

--Bob Y. 16:34, 21 February 2014 (EST)

Retrieved from "https://wiki.seas.harvard.edu/geos-chem/index.php?title=Currently_unresolved_issues_in_GEOS-Chem&oldid=49295"