search

tomasstolker / species

Toolkit for atmospheric characterization of directly imaged exoplanets
https://species.readthedocs.io
MIT License
22 stars 9 forks source link

Make the reported luminosity be the luminosity only of the atmosphere #113

Closed gabrielastro closed 2 weeks ago

gabrielastro commented 2 months ago

Currently, the - luminosity = reported under Calculate multi-photometry and plotted in the corner plot (as log(L/LSun)) is the (reddened) photospheric luminosity PLUS the integrated luminosity of the additional blackbody.

I would like to suggest that the much more useful quantity is the atmospheric luminosity only; the (additional) blackbody is usually not an earnest fit that is reliable enough to be integrated, especially if no far-infrared data or so are available.

species already conveniently shows L_disc/L_atm in the corner plot to let the user check the importance of the blackbody, which is very nice. However, the main output of interest is the luminosity of the atmosphere, I would say, especially in the light of comparing the results from different atmospheric models.

The less obvious question is whether to report the de-reddened luminosity and/or the "empirical" one that comes from fitting a reddened atmospheric model. Both are useful: L_bol,empirical for comparing between observations and/or models, and L_bol,deredded for use with evolutionary models (assuming that the extinction is exterior enough to the planet for it not to affect its cooling, which is ok to first order or much better, e.g. for extinction really from the ISM).

Thanks to your help, I do have a script to compute the intrinsic and the de-reddened luminosities :nerd_face: but maybe it would be useful in general :slightly_smiling_face: and if the calculation were done automatically!

gabrielastro commented 2 months ago

(Thanks for adding the label.) This would affect also in a positive way the luminosity that can easily be added to the legend of plot_spectrum through leg_param for an easy overview. I guess luminosity should be renamed to be explicit.

And on a related note, it would be nice to be able to just add e.g. mass or, especially, luminosity to the legend entry, without however having to set all other parameters, especially since they are different for most model families. The current default settings are great and it would be nice to just have the option to add luminosity/mass/Bayesian evidence/… Thanks if this is not too much trouble to implement!

tomasstolker commented 1 month ago

Thanks for the suggestion! I have changed luminosity to log_lum, log_lum_atm, and log_lum_disk. You can use get_parameters() from ReadModel to select the model parameters and then add e.g. mass to create a leg_param list.

gabrielastro commented 1 month ago

Thanks a lot! These quantities are de-reddened (i.e., the intrinsic total and partial luminosities), yes? L_atm,dereddened = 4 π R² σ Teff⁴ and similarly for the disc luminosity. There could still be some use, especially in the corner plot, for the reddened (extincted) atmospheric luminosity, but thanks for the addition!

tomasstolker commented 1 month ago

These are all intrinsic luminosities!

tomasstolker commented 2 weeks ago

Closing this issue since it has been resolved