Degeneracies

[phajy]

We would like to explore model degeneracies - i.e., what different parameter combinations give similar iron lines. To do that we need some way of determining when two line profiles are the "same".

[wiktoriatarnopolska]

https://ui.adsabs.harvard.edu/abs/2014PhRvD..90f4002J/abstract

[phajy]

A nice way to explore the degeneracies might be to create a "fake" dataset that is of comparable quality to, e.g., a deep XMM-Newton observation or a good Athena observation. We can then use an MCMC fitting method (or something else) to explore the parameter degeneracies. The "fake" data could be a simple power law plus iron line for simplicity.

[fjebaker]

I, err, was told there might be some new Athena response files issued soon (if they aren't already).

[phajy]

Would be really nice to simulate some Athena observations. I'll see if I can find the latest responses...

[wiktoriatarnopolska]

I posted the code to reproduce Johannsen (2014) .

The code I posted successfully reproduces figure 9 (middle tile), however, using the same recipe I couldn't reproduce figure 2 which was my actual aim.

the code utilises a super basic normalisation and I also added code for non-normalised plots just to show the difference.

[wiktoriatarnopolska]

Fig. 9 middle tile

[wiktoriatarnopolska]

Fig. 9 last tile

[wiktoriatarnopolska]

Fig. 9 1st tile

For Johannsen doesn't much perfectly -- for us it very much does.

[wiktoriatarnopolska]

Degeneracies and Johannsen2014

• verify fig. 1 results (?), fig. 2
• fig. 5
• fig. 7, 9