jaxoplanet icon indicating copy to clipboard operation
jaxoplanet copied to clipboard

power2 limb-darkening?

Open saigrain opened this issue 9 months ago • 5 comments

Has anyone tried to implement a power2 limb-darkening law (cf Maxted 2018, Maxted & Gill 2018) in jaxoplanet (or even exoplanet)? Or is there any plan to do so? My understanding is that it gives a much better match to observed limb-darkening profiles near the limb than polynomial laws.

saigrain avatar Mar 31 '25 20:03 saigrain

Hi @saigrain, thanks for asking. There is no plan to implement it although it could be very interesting to have it!

I see the implementation is fairly easy (code provided in https://arxiv.org/abs/1812.01606). Let me know if you are working on an implementation on your side or I am also happy to implement it and open a pull request!

lgrcia avatar Apr 01 '25 14:04 lgrcia

@lgrcia: I may be wrong here but since this is a non-polynomial limb-darkening parameterization, would you implement this by generating it in the power2 parameter space but then fitting it with a high-order polynomial so that it works within the arbitrary polynomial limb-darkening framework we have?

soichiro-hattori avatar Apr 01 '25 14:04 soichiro-hattori

@soichiro-hattori, you are totally right, I had the same thought. But we don't have to implement it using starry at all. Check out the appendix in this paper and you'll see that the code is fairly simple. The light curve has a closed form and can be implemented with jax.numpy functions.

I pushed some working code to the power2 branch but I'd be very happy for @saigrain to open a PR if she wants to give the implementation a try!

EDIT: sorry wrong paper, corrected the url!

lgrcia avatar Apr 01 '25 14:04 lgrcia

Amazing, that was quick! I hadn't started my own implementation, so I will try testing the branch @lgrcia started.

saigrain avatar Apr 01 '25 15:04 saigrain

Not sure of the implications but just wanted you to be aware that there is a minor error in the 2018 papers that was noted in 2019 in regard to the physicality of the coefficients (which the 2018 code may not have corrected): https://iopscience.iop.org/article/10.3847/2515-5172/ab3a3e/ampdf

TylerFair avatar Apr 03 '25 06:04 TylerFair