Conversation
kwargan
left a comment
There was a problem hiding this comment.
Looks good to me. Does the observation operator actually assimilate OMIEFF using apriori and averaging kernels like in the GSI?
|
I am sorry Ricardo but as as you might recall the |
No unfortunately I don't think |
There was a problem hiding this comment.
Given recent discussions with Ricardo, Kris and in a matter of documenting this when it is time to revisit:
We agreed that the AtmVertInterpLay should be maintained for the current T2O effort on FP. ColumnRetrieval will be used in the CF configuration using the same dataset possibly and then transitioned to FP over the years.
Keep in mind that AtmVertInterpLay, despite similarity in coding style, is not equivalent to what is in GSI currently especially regarding the efficiency/averaging kernels factors and apriori... so we shouldn't expect a match between GSI and JEDI using this operator. NOAA/EMC also reported TL/AD issues but this has to be proven than just said... the possible degradation in O3 analysis needs to be carefully adressed then. I am happy to help if needed.
Vertical column integration routines below
GSI intoz:
https://github.com/GEOS-ESM/GEOSana_GridComp/blob/develop/GEOSaana_GridComp/GSI_GridComp/intoz.f90
JEDI AtmVertInterpLay:
https://github.com/JCSDA/ufo/blob/develop/src/ufo/operators/atmvertinterplay/vert_interp_lay_mod.F90
JEDI ColumnRetrieval:
https://github.com/JCSDA/ufo/blob/develop/src/ufo/operators/columnretrieval/ufo_satcolumn_mod.F90
Description
This is done for consistency w/ more recent versions of GEOS which from a certain data onward replace OMI_AURA w/ OMIEFF_AURA - the two datasets are similar but different versions with the more recent version being given a GMAO-centric name.
Dependencies
Impact
None