GEOS-Chem v8-01-03

From Geos-chem
Jump to: navigation, search

Overview

BETA RELEASE December 2008

What's new in this version

Contains everything in v8-01-02 plus the following updates:

Previous issues now resolved in v8-01-03

TPCORE replaced with version based on GMI model

Please see the discussion on our TPCORE wiki page about the implementation of the GMI version of TPCORE into GEOS-Chem.

Also see this PDF document for more information about the new TPCORE version as installed in GEOS-Chem v8-01-03.

--Bob Y. 09:42, 17 February 2010 (EST)

Output of ND51 does not match output in the ctm.bpch file

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

Claire Carouge wrote:

I found the problem [that Prasad Kasibhatla had reported] between ctm.bpch and ts_satellite.bpch. In fact for ND51, a variable is declared with a wrong type. So in diag51_mod.f you need to replace these lines :
     ! Local variables 
     INTEGER :: I, LT
by these:
     ! Local variables 
     INTEGER :: I
     REAL*8  :: LT
Doing so, I only get errors smaller than 5e-5 % between ctm.bpch and ts_satellite.bpch.

--Bob Y. 14:01, 15 December 2008 (EST)

ND22 (J-value) diagnostic

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

A problem was discovered in the ND22 diagnostic. If you typed the following in the diagnostic section of the input.geos file:

ND22: J-Values          : 47   all             
      => JV time range  :      11 13

then this was fine. However, if you requested a subset of tracers, then you would not get the correct number of tracers printed out.

Claire Carouge and Bob Yantosca have implemented a fix into diag3.f such that there is now a test for the name of each tracer. As a side-effect of this fix, you may no longer type all into the input.geos file. You must explicitly list each tracer number. For example, you may type:

ND22: J-Values          : 47   1 7 8 20 99              
      => JV time range  :      11 13

where tracers #1, #7, #8, #20 will cause J(NO2), J(HNO3), J(H2O2) and J(CH2O) (respectively) to be archived to the diagnostics. Tracer #99 will cause both J(OH) and J(O3) (which is really J(O1D)) to be archived.

--Bob Y. 15:53, 15 December 2008 (EST)

ND34 (biofuel emissions) diagnostic

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

Prasad Kasibhatla pointed out an inconsistency in the biofuel emissions diagnostics ND29 (slot #3) and ND34. There was a problem in the writing of ND34 when asking for a subset of tracers only. The code doesn't write out the array parts corresponding to the tracers asked for.

To fix it, you need to change the line in diag3.f

           ARRAY(:,:,1) = AD34(:,:,N) / SCALESRCE

by:

           DO MM = 1, NBFTRACE
              IF ( BFTRACE(MM) == NN ) THEN
                 MMB = MM
                 EXIT
              ENDIF
           ENDDO

           ARRAY(:,:,1) = AD34(:,:,MMB) / SCALESRCE

Also as a side-effect of this fix, you must explicitly declare the tracer numbers for which you wish to archive biofuel emissions. For example, use this line in input.geos:

ND34: Biofuel emissions :  1   1 4 5 9 10 11 18 19 20 21

--Bob Y. 16:01, 15 December 2008 (EST)

Change of location of calls to diagnostics in main.f

DIAG_2PM

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

We have moved the call to subroutine DIAG_2PM to after the call to subroutine INTERP. This will ensure that the correct value of tropopause pressure (TROPP) will be obtained for GEOS-4 met fields.

NOTE: For GEOS-5, the TROPP field is a 3-hour average and not and instantaneous field, so this move should have no effect.

--Bob Y. 16:18, 15 December 2008 (EST)

DIAG1, DIAG41, DIAG42

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

In GEOS-Chem driver program main.f, we have moved the following calls to diagnostic subroutines:

        !==============================================================
        !       ***** A R C H I V E   D I A G N O S T I C S *****
        !==============================================================
        IF ( ITS_TIME_FOR_DYN() ) THEN

           ! Accumulate several diagnostic quantities
           CALL DIAG1

           ! ND41: save PBL height in 1200-1600 LT (amf)
           ! (for comparison w/ Holzworth, 1967)
           IF ( ND41 > 0 ) CALL DIAG41

           ! ND42: SOA concentrations [ug/m3]
           IF ( ND42 > 0 ) CALL DIAG42

           !### Debug
           IF ( LPRT ) CALL DEBUG_MSG( '### MAIN: a DIAGNOSTICS' )
        ENDIF

from after transport to after wet deposition. This was to remove some historical baggage, and to make the ctm.bpch diagnostics more consistent with the ND48, ND49, ND50, ND51 timeseries diagnostics.

The impact of the reordering of the diagnostics

NOTE: We are phasing out binary punch diagnostics output, in favor of netCDF output.

In GEOS-Chem v8-01-02 and earlier releases, the order of operations as called from the main.f driver program was as follows:

  1. Advection (i.e. TPCORE)
  2. PBL mixing
  3. Cloud convection
  4. MONTHLY or DAILY MEAN DIAGNOSTICS (i.e. DIAG1, DIAG41, DIAG42)
  5. Dry Deposition
  6. Emissions
  7. Chemistry
  8. Wet Deposition
  9. TIMESERIES DIAGNOSTICS (i.e. ND48, ND49, ND50, ND51)

Subroutine DIAG1 is the main accumulator for several of the monthly or daily mean diagnostics (e.g. tracer concentrations, surface pressure, etc.) in ctm.bpch. DIAG41 archives afternoon PBL heights and DIAG42 archives SOA concentrations (these were added later).

Due to historical reasons, the call to DIAG1 (as well as later-added routines DIAG41 and DIAG42) had been placed the cloud convection. However, as the code developed and timeseries diagnostics (ND48, ND49, ND50, ND51) were incorporated into GEOS-Chem, these timeseries diagnostics were called after chemistry and wet deposition.

A couple of users have pointed out that the MONTHLY or DAILY MEAN diagnostics were not consistent with the TIMESERIES DIAGNOSTICS because they are called from different points within the main driver program. Therefore, in order to improve consistency between these all diagnostics, we now call the MONTHLY or DAILY MEAN DIAGNOSTICS (e.g. DIAG1, DIAG41, DIAG2) after chemistry and wet deposition.

Therefore, in GEOS-Chem v8-01-03, the new order of operations is as follows:

  1. Advection (i.e. TPCORE)
  2. PBL mixing
  3. Cloud convection
  4. Dry Deposition
  5. Emissions
  6. Chemistry
  7. Wet Deposition
  8. MONTHLY or DAILY MEAN DIAGNOSTICS (i.e. DIAG1, DIAG41, DIAG42)
  9. TIMESERIES DIAGNOSTICS (i.e. ND48, ND49, ND50, ND51)

NOTE: The switch in diagnostic location may result in some changes in tracer concentration as compared to previous versions of GEOS-Chem. This may especially be true for the dust and seasalt tracers.

--Bob Y. 11:04, 6 May 2009 (EDT)