Skip to content
Permalink
Branch: master
Commits on Apr 5, 2019
  1. Units of bpch ALD2_OCEAN_SOURCE diagnostic should be kgC/m2/s

    yantosca committed Apr 5, 2019
    The ALD2_OCEAN_SOURCE manual diagnostic (for bpch output) is defined
    with units of kg/m2/s but should be kgC/m2/s.  This is necessary to
    be consistent with diag3.F.
    
    NOTE: We are phasing out bpch output soon.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Apr 4, 2019
  1. Now dimension State_Met%Area_M2 as (I,J) instead of (I,J,1)

    yantosca committed Apr 4, 2019
    The State_Met%Area_M2 field was allocated as (I,J,1), but this causes
    issues with GCHP.  When GCHP sees a 3rd dimension, it will try to
    save diagnostics with all levels.
    
    The quick fix is to allocate State_Met%Area_M2 as an (I,J) array
    instead of as an (I,J,1) array.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  2. Add redesigned noontime J-value diagnostics

    yantosca committed Apr 4, 2019
    We now have added the noontime J-value diagnostics as 2 separate arrays:
    
    (1) State_Diag%JNoon: This is the cumulative sum of noontime J-values
        for each photolyzed species.  (Noontime is defined as between 11:00
        and 13:00 GMT).
    
    (2) State_Diag%JNoonFrac: This is the cumulative sum of the times when
        it was local noon at each grid box.  When this is time-averaged,
        it will become the fraction of time that local noon occurred at
        each grid box.
    
    Therefore, to compute the local noon J-value diagnostics, one must
    divide State_Diag%JNoon by State_Diag%JNoonFrac in post-processing.
    
    Also added some minor cosmetic changes elsewhere in code.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Mar 28, 2019
  1. Comment out MINIT array from strat_chem_mod.F90

    yantosca committed Mar 28, 2019
    The MINIT array is allocated but not currently used.  Therefore, it is
    taking up extra memory.  This could potentially impede simulations,
    especially in cloud computing environments.
    
    For now, we have commented out all instances of MINIT, as well as
    subroutine SET_MINIT.  At some further point in time we will delete
    these commented lines.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  2. Bug fix: Make sure planeflight writes out on last timestep of day

    yantosca committed Mar 28, 2019
    Luke Schiferl wrote:
    
       The planeflight diagnostic in v12.2.0 fails to read out points in
       the final dynamic timestep of the day. It seems that this is due to
       a remnant from before the transition from minutes to seconds used
       to describe timesteps.
    
    This has now been corrected.
    
    Also -- changed the GC version number to 12.3.1 in the file
    GeosCore/gc_classic_version.H.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  3. Revert to getting volcanic emissions from base emissions for BPCH dia…

    msulprizio committed Mar 28, 2019
    …gnostics
    
    In the original 12.3.0 benchmark, we updated to OMI-based volcanic
    emissions that were obtained from the AeroCom_Volcano extension.
    Those emissions proved to be too low, so we have reverted to the
    old volcanic emissions and therefore need to get those emissions from
    the base emissions (ExtNr=0) in hcoi_gc_diagn_mod.F90.
    
    Signed-off-by: Melissa Sulprizio <mpayer@seas.harvard.edu>
  4. Update GC version to 12.3.0

    msulprizio committed Mar 28, 2019
    Signed-off-by: Melissa Sulprizio <mpayer@seas.harvard.edu>
Commits on Mar 27, 2019
  1. Merge branch 'TagCO_fix' into dev/12.3.0

    msulprizio committed Mar 27, 2019
  2. Fix CO units and write secondary production into a netcdf file

    Beata Bukosa authored and msulprizio committed Mar 26, 2019
Commits on Mar 21, 2019
  1. Merge branch 'HEMCO' into dev/12.3.0

    yantosca committed Mar 21, 2019
    Removed a leftover file (hcox_aerocom_mod.X), and also made minor
    comment changes in hcox_volcano_mod.F90.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  2. Remove temporary file hcox_aerocom_mod.X

    yantosca committed Mar 21, 2019
    This file was committed by mistake.  We are removing it now.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Mar 20, 2019
  1. Rename hcox_aerocom_mod.F90 to hcox_volcano_mod.F90 (and fix diagnost…

    yantosca committed Mar 20, 2019
    …ics)
    
    In the HEMCO/Extensions folder, we have renamed hcox_aerocom_mod.F90
    to hcox_volcano_mod.F90.  We now have the option of using either
    AeroCom or OMI emissions data, so we have changed the name of the module
    accordingly.  Hopefully this should prevent confusion.
    
    We have also renamed the flag ExtState%Aerocom to ExtState%Volcano,
    and module routines HCOX_AeroCom_* to HCOX_Volcano_*.
    
    We have also fixed a bug: We now remove the MinDiagLev=2 parameter
    from being passed to HCO_EmisAdd.  This was causing the degassing
    and eruptive emissions to both be zero (because it was ignoring the
    category values).
    
    One final fix: We corrected a typo in hcox_paranox_mod.F90, where
    ExtState%AeroCom should have been ExtState%ParaNOx in that module.
    There should have been no references to Volcano from ParaNOx.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Mar 19, 2019
  1. Merge branch 'HEMCO' into dev/12.3.0

    yantosca committed Mar 19, 2019
  2. Bug fix: Use ExtNrMono in MEGAN extension for CO, MTPA, MTPO, LIMO, SESQ

    yantosca committed Mar 19, 2019
    Biogenic emissions for species CO, MTPA, MTPO, LIMO, and SESQ were
    not being computed properly.  We need to pass the MEGAN monoterpenes
    extension number (Inst%ExtNrMono) to HCO_EmisAdd when computing
    emissions for these.  In this way, the automatic diagnostics will
    also be computed properly.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Mar 10, 2019
  1. Merge branch 'HEMCO' into dev/12.3.0

    msulprizio committed Mar 10, 2019
  2. Fix error reading restart file when using a fixed emissions year in H…

    msulprizio committed Mar 8, 2019
    …EMCO
    
    A user brought to our attention an issue caused by setting the emissions
    year in HEMCO_Config.rc when using GEOS-Chem 12.1.0. Starting in that
    version, we read in met fields and the GEOS-Chem restart file via HEMCO.
    Setting the emissions year seems to also override the year for those
    files when the user probably still wants to use the simulation year.
    
    Christoph Keller wrote:
    
       The year used in HEMCO is handled by the HcoClock object. It carries
       three year values: Clock%ThisYear is the current 'emission year',
       Clock%PrevYear is the emission year of the previous time step, and
       Clock%SimYear is the current simulation year. Clock%SimYear and
       Clock%ThisYear are always identical except for cases where the
       emission year is fixed (via HEMCO_Config.rc). When reading a file,
       HEMCO always uses Clock%ThisYear as the reference year. But it sounds
       like for some files, we would want to use Clock%SimYear instead. This
       would have to happen in routine HCO_GetPrefTimeAttr (in
       hco_tidx_mod.F90) when the current year is obtained:
    
        ! Get current time
        CALL HcoClock_Get( am_I_Root, HcoState%Clock,                   &
                           cYYYY    = cYr, cMM = cMt, cDD = cDy,        &
                           cWEEKDAY = cWd, cH  = cHr, cM  = cMn, RC = RC )
        IF ( RC /= HCO_SUCCESS ) RETURN
    
    Change to:
        ! Get current time
        CALL HcoClock_Get( am_I_Root, HcoState%Clock,                   &
                           cYYYY    = cYr, cMM = cMt, cDD = cDy,        &
                           cWEEKDAY = cWd, cH  = cHr, cM  = cMn,        &
                           sYYYY    = sYr, RC = RC )
        IF ( RC /= HCO_SUCCESS ) RETURN
        IF ( SOMETHING ) cYr = sYr
    
      The question is what the determination for this switch should be (which
      I just called 'SOMETHING' above)? A more complicated, but more robust
      approach would be to specifically mark these files in HEMCO_Config.rc
      (e.g. by adding an attribute to the time flag) and then add a
      corresponding flag to the FileData derived type (let's say
      'AlwaysUseSimYear'), which already contains a bunch of meta-data that
      goes with each input file. This flag would then determine whether to
      use cYr or sYr:
    
        IF ( Lct%Dct%Dta%AlwaysUseSimYear) cYr = sYr
    
    Signed-off-by: Melissa Sulprizio <mpayer@seas.harvard.edu>
Commits on Mar 7, 2019
  1. Merge branch 'GEOS' into dev/12.3.0

    lizziel committed Mar 7, 2019
    # Conflicts:
    #	HEMCO/Core/hco_error_mod.F90
    #	HEMCO/Core/hco_readlist_mod.F90
    #	HEMCO/Core/hco_state_mod.F90
    #	HEMCO/Core/hco_types_mod.F90
    #	HEMCO/Interfaces/hcoi_esmf_mod.F90
    #	Headers/species_database_mod.F90
    
    Signed-off-by: Lizzie Lundgren <elundgren@seas.harvard.edu>
  2. Merge branch 'HEMCO' into dev/12.3.0

    yantosca committed Mar 7, 2019
  3. Fix minor typo in logical assignment

    lizziel committed Mar 7, 2019
    Signed-off-by: Lizzie Lundgren <elundgren@seas.harvard.edu>
  4. Merge branch 'HEMCO' of github.com:geoschem/geos-chem into HEMCO

    yantosca committed Mar 7, 2019
    Merge the fix for GetIndex2Interp into the updated HEMCO branch
    containing GEOS-5 updates from Christoph Keller
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  5. Fix minor typo in logical assignment

    lizziel committed Mar 7, 2019
    Signed-off-by: Lizzie Lundgren <elundgren@seas.harvard.edu>
  6. Updates from Christoph Keller retrieved from GEOS development branch

    lizziel committed Mar 7, 2019
    Changes in include:
    - Now use internal counter to determine first-time reading
    - Only shift hh/mm when applying time shift
    - Cap time shift to same day (defaults to no)
    - Replace underscores in long-name string with spaces
    
    Signed-off-by: Lizzie Lundgren <elundgren@seas.harvard.edu>
  7. Updates from Christoph Keller retrieved from GEOS development branch

    lizziel committed Mar 7, 2019
    Changes in include:
    - Now use internal counter to determine first-time reading
    - Only shift hh/mm when applying time shift
    - Cap time shift to same day (defaults to no)
    - Replace underscores in long-name string with spaces
    
    Signed-off-by: Lizzie Lundgren <elundgren@seas.harvard.edu>
  8. Merge branch 'HEMCO' into bugfix/Restart

    yantosca committed Mar 7, 2019
  9. HEMCO bug fix: Avoid out-of-bounds error in routine GetIndex2Interp

    yantosca committed Mar 7, 2019
    In routine GetIndex2Interp (in HEMCO/Core/hcoio_read_std_mod.F90),
    we need to make sure that tidx2 does not exceed nTime, which is the
    number of time slices in the file.  This can cause an out-of-bounds
    error.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Mar 5, 2019
  1. Further bug fix for HISTORY intervals that span leap years

    yantosca committed Mar 5, 2019
    (1) History/histcontainer_mod.F90: We no longer subtract the heartbeat
        timestep from the initial UpdateAlarm value.  This is because we
        now advance the clock in main.F before calling History_Update
        (which updates all diagnostic quantities in a collection).
    
    (2) History/history_mod.F90: Call HistContainer_UpdateIvalSec to
        recompute the update alarm interval (for monthly or hourly intervals)
        within History_Update.  This will make sure we adjust the update
        interval for leap year days.
    
    (3) GeosCore/main.F: We now call History_SetTime before History_Update.
        We need to advance the clock before History_Update so that we can
        properly adjust the update interval for leap year days.
    
    Also corrected an incorrect comment in Headers/state_met_mod.F90.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  2. Add fixes in hcoi_gc_diagn_mod.F90 to get SO2 volcano emissions from …

    msulprizio committed Mar 4, 2019
    …AeroCom_Volcano extension
    
    The volcano emissions previously came into HEMCO via the base emissions, but
    are now processed through the AeroCom_Volcano extension. This needed to be
    updated in hcoi_gc_diagn_mod.F90 in order to save out these emissions to BPCH.
    
    In hcox_aerocom_mod.F90, we also no longer need to hardcode year to
    the 2005-2012 time range because the OMI-based emissions are available
    for 1979-present.
    
    Signed-off-by: Melissa Sulprizio <mpayer@seas.harvard.edu>
Commits on Feb 27, 2019
  1. Merge branch 'bugfix/Restart' into dev/12.3.0

    yantosca committed Feb 27, 2019
Commits on Feb 26, 2019
  1. Fixed bug in AlarmIncrementMonths: increment month at end of loop

    yantosca committed Feb 26, 2019
    Fixed an issue in routine AlarmIncrementMonths (in module
    History/histcontainer_mod.F90).  We have to increment the month at the
    end of the DO loop after updating the nDays variable.  Otherwise this
    causes an incorrect count of days.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
  2. Add an improved test for setting interval alarms

    yantosca committed Feb 26, 2019
    Rewrote the algorithm to compute the Update Interval, File Write
    Interval, and File Close Interval.  Intervals of longer than 1 month
    or 1 year are now allowed.  Also, the interval computation should be
    much more robust when straddling leap year days.
    
    This update fixes an issue with the Restart collection (when used with
    the "End" frequency) not properly writing out files for intervals that
    straddle leap year days.
    
    Also note: the computation for deltaYMD and deltaHMS in history_mod.F90
    (which is used for the "End" frequency) has also been updated to be
    more robust.  The new algorithm should work for most common model
    simualtion intervals (i.e. integral numbers of minutes or hours or days
    or months).  There might be a few edge cases that would cause it to
    fail, but the new algorithm is definitely an improvement over the
    prior situation.
    
    Signed-off-by: Bob Yantosca <yantosca@seas.harvard.edu>
Commits on Feb 25, 2019
  1. Merge branch 'HEMCO' into dev/12.3.0

    yantosca committed Feb 25, 2019
  2. Bug fix for interpolation between files: make sure it works both for …

    Christoph Keller authored and yantosca committed Feb 25, 2019
    …interpolation into the future and into the past.
  3. Bug fix for calculation of interpolation weights: now use julian date…

    Christoph Keller authored and yantosca committed Feb 24, 2019
    …s instead of YMDhm.
Commits on Feb 22, 2019
  1. Merge branch 'dev/12.2.1' into dev/12.3.0

    yantosca committed Feb 22, 2019
Older
You can’t perform that action at this time.