search

joshspeagle / brutus

Modeling stellar photometry with "brute force" methods
MIT License
32 stars 5 forks source link

Reddening priors #17

Closed joshspeagle closed 5 years ago

joshspeagle commented 5 years ago

Systematics tests on clusters have shown a bias in both MIST and Bayestar models to systematically underestimate distances to stars in clusters at lower masses. This effect is driven by the prior: all else being equal, it is preferable to push the mass down, which pushes the distances closer and the rest of the parameters (metallicity, reddening, etc.) adjust to compensate. One approach to help resolve his issue is to impose a (distance-)reddening prior that can help prevent as much “sloshing” of the photometry, so I want to implement options for an Av prior similar to the Rv prior as well as utilizing some version of the 3-D Bayestar2017 dust map.

joshspeagle commented 5 years ago

This has been resolved as of f8b331ec0f231a20ad92f4a6d69755f65958611f, although it hasn't been fully incorporated into the LOS modules as of yet. I'll open a new issue to track progress on that front.