available simulations for CORDEX-CMIP5

[JavierDiezSierra]

@larsbuntemeyer @jesusff

The script for downloading CMIP5-CORDEX-EUR-11 data for the evaluation experiment, including the variables ['tas', 'tasmin', 'tasmax', 'pr', 'psl', 'orog', 'sftlf'], is ready.

Before starting the download, I have checked the availability of both RCMs and variables.

Here is the list of available RCMs:

'CLMcom-ETH.COSMO-crCLIM-v1-1.v1'
'CLMcom.CCLM4-8-17.v1'
'CNRM.ALADIN53.v1'
'CNRM.ALADIN63.v1'
'DHMZ.RegCM4-2.v1'
'DMI.HIRHAM5.v1'
'GERICS.REMO2015.v1'
'ICTP.RegCM4-6.v1'
'IPSL.WRF381P.v1'
'KNMI.RACMO22E.v1'
'MOHC.HadREM3-GA7-05.v1'
'MPI-CSC.REMO2009.v1'
'RMIB-UGent.ALARO-0.v1'
'SMHI.RCA4.v1'

Minor issues encountered:

• IPSL.WRF381P.v1: Does not have monthly data, only daily.

• Variables not found:
  DHMZ.RegCM4-2.v1.psl
  IPSL.WRF381P.v1.orog
  IPSL.WRF381P.v1.sftlf
  RMIB-UGent.ALARO-0.v1.tasmin
  RMIB-UGent.ALARO-0.v1.tasmax
  RMIB-UGent.ALARO-0.v1.psl

[JavierDiezSierra]

I have just started to download cmip5-cordex data to: /mnt/CORDEX_CMIP6_tmp/aux_data/cmip5-cordex

@larsbuntemeyer

[larsbuntemeyer]

@JavierDiezSierra The catalog.csv now holds only the CMIP5 data. Should we make two catalogs for CORDEX-CMIP5 and CORDEX-CMIP6? I see that you translated the attributes to CMIP6, that's good, maybe we should add a mip_era column to the catalog, so we can distinguish CMIP5 and CMIP6 data., then we can have both in one catalog.. Let me check how to handle a master catalog...

[JavierDiezSierra]

@larsbuntemeyer Yes, I find both alternatives acceptable, a master catalog or two individual catalogs. Let me know if you need any help.

[jesusff]

yes, the mip_era should be the way to go to distinguish both data streams.

@JavierDiezSierra I renamed on disk the main directory from cmip5-cordex to cordex-cmip5, to keep consistency with CORDEX-CMIP6. At what stage is the old DRS mapped to the CORDEX-CMIP6 one? The data on disk should (and they do now) follow their old, native DRS

[larsbuntemeyer]

Agreed, i'll add a mip_era column!

[JavierDiezSierra]

@larsbuntemeyer Many thanks! Let me know once the new catalog is created (or I can create it if you prefer) with both CMIP5 and CMIP6 CORDEX simulations to include the CMIP5 analysis in the eobs.ipynb notebook. Currently, it is not working...

[JavierDiezSierra]

@larsbuntemeyer I have just updated the catalog to include both CMIP5 and CMIP6. I plan to commit the changes this morning.

:)

[JavierDiezSierra]

Some CMIP5 simulations were removed for several issues:

• Corrupted files: '/mnt/CORDEX_CMIP6_tmp/aux_data/cordex-cmip5/CORDEX/output/EUR-11/ICTP/ECMWF-ERAINT/evaluation/r1i1p1/RegCM4-6/v1/mon/tas/v20190502'

• Error when combining by coords with xarray: '/mnt/CORDEX_CMIP6_tmp/aux_data/cordex-cmip5/CORDEX/output/EUR-11/DMI/ECMWF-ERAINT/evaluation/r1i1p1/HIRHAM5/v1/mon/tas'