Reprojection of magnetogram observations may be misleading

[samaloney]

Provide a general description of the issue or problem.

Consider a case where B == B_los, reproject the B_los observations to a new observer at+/-90 deg, B_los should be zero but if you reproject it won't be zero?

[ayshih]

Any reprojection can be misleading if either the measured quantity is not all at exactly rsun (e.g., coronal structures) or the measured quantity depends on observed direction (e.g., anisotropic emission, but also vector components).

For this specific issue, I don't know how we can do anything about it in a non-source-specific way because I don't know how we'd ever be able to programmatically distinguish between a map of a magnetic-field component (e.g., B_los) and a map of the total magnetic field (which doesn't depend on observed direction).

[samaloney]

But we know the specific sources this is an issue for e.g. MDI and HIM right now so we could add a specific warning for these?

[ayshih]

Yes, that's why I used the adjective "non-source-specific". It's a shame that the suggestion to implement measurement-type map classes never caught on...

My hesitation is the slippery slope if we start adding warnings like this:

• Do we add a warning for any image source that shows coronal structures (because they will be reprojected as if they are on the solar surface)?
• Do we add a warning for any image source that shows faculae (where the brightness depends on viewing angle)?