Here at the MO we are at the stage where our first piControl run has started and its performance is too poor for us to maintain the full list of MIP requested variables that our model is capable of. As such we are looking to prune the diagnostic list by removing requests from MIPs that are not interested in the data from this particular configuration (e.g. RFMIP are not interested in data from our medium resolution HadGEM3, nor our UKESM model).
For a given (CMORvar, experiment) combination I need to determine which MIPs want the CMORvar and at what priority.
It is easier, I think, to explore this through a specific example; I’m looking for information on which MIPs want the variable “tsl” in mip table “Lmon” (referred to as Lmon/tsl hereafter) in the piControl experiment.
If I look at the objects within the data request (dreqPy v01.00.15) that join up the Lmon/tsl CMORvar and the piControl experiment I think the following connections are made;
For MIPs PMIP, C4MIP and CFMIP I can trace the following route through the data request objects;
And from the requestVar attributes I can see that these MIPs want this variable at priority 1, 2 and 2 for PMIP, C4MIP and CFMIP respectively.
There is also a single requestVar & requestVarGroup that connect Lmon/tsl and piControl via individual requestLink and requestItem objects for the following MIPs;
AerChemMIP, CMIP, FAFMip, GeoMIP, RFMIP, VolMIP and HighResMIP
The priority associated with this requestVar is 2 (i.e. I assume that all these MIPs want this variable at priority 2).
Combining the above lists of MIPs gives
AerChemMIP, C4MIP, CFMIP, CMIP, FAFMIP, GeoMIP, HighResMIP, PMIP, RFMIP, and VolMIP
Looking at the piControl spreadsheet available via [1] the “MIPs (requesting)” column for Lmon/tsl contains
AerChemMIP, C4MIP, CFMIP, CMIP, FAFMIP, GMMIP, GeoMIP, HighResMIP, LS3MIP, LUMIP, PMIP, RFMIP, VIACSAB, VolMIP
I.e. the spreadsheet appears to suggest that the following additional MIPs are interested in our variable; GMMIP, LS3MIP, LUMIP and VIACSAB
From the spreadsheet describing the data requested by GMMIP from CMIP experiments [2] it looks like GMMIP are only interested in the amip experiment. Looking at the equivalent pages for LS3MIP, LUMIP and VIACSAB suggests that they are not interested in the piControl experiment.
Could you confirm whether I have correctly interpreted the links between CMORvar, experiment and MIP objects within the data request?
TIA,
Matthew
Here at the MO we are at the stage where our first piControl run has started and its performance is too poor for us to maintain the full list of MIP requested variables that our model is capable of. As such we are looking to prune the diagnostic list by removing requests from MIPs that are not interested in the data from this particular configuration (e.g. RFMIP are not interested in data from our medium resolution HadGEM3, nor our UKESM model). For a given (CMORvar, experiment) combination I need to determine which MIPs want the CMORvar and at what priority.
It is easier, I think, to explore this through a specific example; I’m looking for information on which MIPs want the variable “tsl” in mip table “Lmon” (referred to as Lmon/tsl hereafter) in the piControl experiment.
If I look at the objects within the data request (dreqPy v01.00.15) that join up the Lmon/tsl CMORvar and the piControl experiment I think the following connections are made; For MIPs PMIP, C4MIP and CFMIP I can trace the following route through the data request objects;
And from the requestVar attributes I can see that these MIPs want this variable at priority 1, 2 and 2 for PMIP, C4MIP and CFMIP respectively.
There is also a single requestVar & requestVarGroup that connect Lmon/tsl and piControl via individual requestLink and requestItem objects for the following MIPs; AerChemMIP, CMIP, FAFMip, GeoMIP, RFMIP, VolMIP and HighResMIP
The priority associated with this requestVar is 2 (i.e. I assume that all these MIPs want this variable at priority 2).
Combining the above lists of MIPs gives AerChemMIP, C4MIP, CFMIP, CMIP, FAFMIP, GeoMIP, HighResMIP, PMIP, RFMIP, and VolMIP
Looking at the piControl spreadsheet available via [1] the “MIPs (requesting)” column for Lmon/tsl contains AerChemMIP, C4MIP, CFMIP, CMIP, FAFMIP, GMMIP, GeoMIP, HighResMIP, LS3MIP, LUMIP, PMIP, RFMIP, VIACSAB, VolMIP I.e. the spreadsheet appears to suggest that the following additional MIPs are interested in our variable; GMMIP, LS3MIP, LUMIP and VIACSAB
From the spreadsheet describing the data requested by GMMIP from CMIP experiments [2] it looks like GMMIP are only interested in the amip experiment. Looking at the equivalent pages for LS3MIP, LUMIP and VIACSAB suggests that they are not interested in the piControl experiment.
Could you confirm whether I have correctly interpreted the links between CMORvar, experiment and MIP objects within the data request? TIA, Matthew
[1] http://clipc-services.ceda.ac.uk/dreq/data/tabs02/cmvme_TOTAL_piControl_3_3.xlsx [2] http://clipc-services.ceda.ac.uk/dreq/tabs03/expt_GMMIP_CMIP_3_3.html