search

cmip6dr / Request

Request specification (which variables for which experiments, etc)
2 stars 0 forks source link

fx group partially missing in the online version of the DR (by MIP) ? #37

Closed jypeter closed 7 years ago

jypeter commented 7 years ago

When I open the online version of the PMIP request (but I get the same with the other projects), http://clipc-services.ceda.ac.uk/dreq/u/PMIP.html,

from the http://clipc-services.ceda.ac.uk/dreq/index/mip.html page,

I don't find the fx table in the 3.1 panel (Request variable group), but I can find it in the 3.3 panel (Request links)

Shouldn't fx be also present in 3.1 ?

I admit I have a hard time understanding the difference between 3.1 and 3.3, so maybe this is a feature and not a bug. Is there a document I can read to better understand the underlying links concept ?

martinjuckes commented 7 years ago

The documentation is here: http://proj.badc.rl.ac.uk/svn/exarch/CMIP6dreq/tags/latest/dreqPy/docs/dreqML.pdf . It probably needs some improvement.

The variable groups listed under http://clipc-services.ceda.ac.uk/dreq/u/PMIP.html are those defined by PMIP, under the links tab, you find everything requested .. which generally includes references to groups defined by PMIP and groups defined by other MIPs.

jypeter commented 7 years ago

I have found and read The Request XML document and Schema yesterday. The details and especially the table on page 8 help a lot, but I confess still have some trouble figuring out the difference between 3.1 and 3.3

If check fx in 3.3 for PMIP, I get PMIP:fx: Link to group: CMIP6 CMOR Table: fx [14] and if I check it for PMIP I get CMIP:fx: Link to group: CMIP6 CMOR Table: fx [14] so I understand that both PMIP and CMIP are requesting the same thing. But why is this not in 3.1?

And the fact that this mentions one CMOR table, but we have two tabs (fx and Ofx) in CMIP6_MIP_tables.xlsx is a bit confusing (#38)

martinjuckes commented 7 years ago

Some variable groups (e.g. http://clipc-services.ceda.ac.uk/dreq/u/0bb7eb9e-7f05-11e6-b027-ac72891c3257.html ) are requested by many MIPs, each for a different combination of experiments, so each MIP has a different requestLink associated with this one variable group. This makes it easier to be clear that they are all asking for exactly the same set of variables.