Wrong date being used for EUI observer coordinate

[dstansby]

Currently in sunpy 3.0.x, the date/time of m.observer_coordinate is taken from m.date, which is taken from the DATE-OBS FITS keyword. For Solar Orbiter (and possibly other missions), the coordinate metadata is valid at the time given by DATE-AVG however, which is different to DATE-OBS (see 3.1.1.10 of the Metadata Definition for Solar Orbiter Science Data).

This will be fixed in 3.1 by https://github.com/sunpy/sunpy/pull/5449/files, but we should issue a fix for the 3.0.x branch in the EUI map source.

[wtbarnes]

Should .date on the EUI source just default to 'DATE-AVG'? That seems like the easiest fix here.

[dstansby]

Yep, I think that's the best thing to do.