vairable changes name 0.5.5 to 0.5.13 and some zeros?

[peterdudfield]

Detailed Description

The variables seems to hange name from 0.5.5 to 0.5.13

Context

• forecast goes wrong, and we think it was due to nwp change

Old (from 4th June)

• ukv: ['dlwrf', 'dswrf', 'hcc', 'lcc', 'mcc', 'prate', 'r', 'sde', 'si10',
  't', 'vis', 'wdir10']
  ecmwf: ['tcc', 'dlwrf', 'dswrf', 'duvrs', 'hcc', 'lcc', 'mcc', 'prate',
  'sde', 'sr', 't', 'u10', 'u100', 'u200', 'v10', 'v100', 'v200',
  'vis']

New

• UKV: ['r', 't', 'vis', 'wdir10', 'si10', 'hcc', 'lcc', 'mcc', 'dlwrf',
  'dswrf', 'prate', 'sde']
• ECMWF: ['sr', 'duvrs', 'sde', 'clt', 'u10', 'v10', 't', 'dswrf', 'dlwrf',
  'lcc', 'mcc', 'hcc', 'vis', 'u200', 'v200', 'u100', 'v100',
  'prate']

Possible Implementation

• work out why this is happened?
• either update NWP-consumer, or update pvnet_app to rename
• or update ocf_datapipes

[peterdudfield]

tcc seems to be missing from ECMWF in the new one, and possibly renamed as clt?

[peterdudfield]

Could be fixed once this is done - openclimatefix/ocf_datapipes#326

Also could be fixed in pvnet_app - https://github.com/openclimatefix/pvnet_app/blob/main/pvnet_app/data.py#L267

[peterdudfield]

Found it: In the live ECMWF data, can see that all values are 0 in the second step

[Sukh-P]

Mentioned this to @devsjc earlier today and I am not sure but these logs look suspicious as they mention renamed variables and says something like "filling missing variables with 0" and has an exception so could be related to that issue

[peterdudfield]

I think this is all done now