De-absorbed AGN spectra

[GernotMaier]

What should we do if there are only de-absorbed spectra given in a paper?
(example is a MAGIC/VERITAS Mrk 501 paper)

I can certainly add a comment (including the EBL model used), but I am worried that the comments are not always read.

Best would be to introduce a flag or a data field (although here the model name and the redshift must then be included).

Any suggestions?

[cdeil]

I think not adding or not properly representing such models in gamma-cat would be a valid approach. It's the same as for the energy-dependent morphology that you mentioned. Those are rare cases and it's a lot of work to add them in a good, useful way (with code that can read and process the model, and documentation of how exactly the model is defined).

If you want to work toward supporting such models in gamma-cat, the practical approach is probably just to add all relevant info using extra fields, or comments with "TODO" that information XML should be added at some point.

Concretely, I think a new spec type (e.g. "ebl-pl") might be a good idea:
https://github.com/gammapy/gamma-cat/blob/master/input/schemas/dataset_source_info.schema.yaml#L313
So that it's immediately clear that it's not just a simple measure PL that we currently support.

[GernotMaier]

There is no spectral model given in the paper - the spectral points itself are given only after de-absorption.

But I see your point, let's:

• add the data with a comment with all the information
• open an issue so that we don't forget it; maybe somebody has time in the future to implement it (I guess it is more important to focus on the data collection for now)

[cdeil]

If it's just the spectral points, yes, you can add any keys or comments you like in the header.

And we can just keep this existing issue open as a reminder that we might want to write down a spec at some point and implement a serialisation format for EBL-absorbed models at some point.

[GernotMaier]

Coming back to this issue after a long time: if we want to add now spectral models after ebl absorption applied, is it simply about adding your suggestion?

Concretely, I think a new spec type (e.g. "ebl-pl") might be a good idea:
https://github.com/gammapy/gamma-cat/blob/master/input/schemas/dataset_source_info.schema.yaml#L313

Would be nice if we could add this.