[BUG/ISSUE] Incorrect source of Be7Strat and Be10Strat below tropopause level in RnPbBe extension

[lizziel]

In the HEMCO RnPbBe extension, emissions for species Be7Strat and Be10Strat are set to the emissions for Be7 and Be10 at and above the tropopause level each time the Emissions component is run. This is done in the following code:

          IF ( L > ExtState%TropLev%Arr%Val(I,J) ) THEN
             IF ( Inst%IDTBe7Strat > 0 ) THEN
                Inst%EmissBe7Strat (I,J,L) = Add_Be7
             ENDIF
             IF ( Inst%IDTBe10Strat > 0 ) THEN 
                Inst%EmissBe10Strat(I,J,L) = Add_Be10
             ENDIF
          ENDIF

Because the tropopause level can change per column per timestep, and because emissions are not reset to zero below the tropopause level, there can be accumulated sources directly below the tropopause for Be7Strat and Be10Strat over time. For example, if TropLev goes up one level for a given column then the previous emission value at the level below the new TropLev persists and is erroneously applied again. This continues until a new TropLev is at or below the original TropLev. The result is an over-estimate of Be7Strat and Be10Strat.

This bug also introduces small differences between model output when breaking up a single segment transport tracer simulation into multiple consecutive runs, such as is done for the GCHPctm transport tracer benchmark.

[lizziel]

This issue is fixed in HEMCO commit 4468b68. It will be included in the 3.0.0 release that accompanies the 13.0.0 release of GEOS-Chem.