Bayesian MCMC galaxy SED modeling

In these days, galaxy SED modeling becomes popular since the current and forthcoming large scale surveys will produce multi-bands photometric data points sufficient to sample the galaxy integrated spectral energy distribution, which allows us to infer many interesting aspects of galaxy properties including stellar mass, star formation history, dust contents etc.

Currently there are several versions of widely used software package to model the galaxy spectra (i.e. stellar population thesis code). However the model parameter space is large and not well constrained, and thus in most cases, people use the population synthesis code with very strong assumptions about IMF and star formation history, which significantly affect our inference of galaxy stellar mass.

To improve this situation, Bayesian MCMC approach to model the galaxy integrated SED is very powerful solution not only for estimating the parameter and its uncertainty but also for assessing the statistical power of different models based on different assumptions made by user (different IMF, star formation history and multiple population etc.), using the Bayes factor model selection.

The successful demonstration of this approach to the galaxy morphology analysis (GALPHAT) convinces me that it will be straight forward to apply BIE to galaxy SED modeling. This is certainly my next project in the near future.

Link to the publically avaliable stellar population synthesis codes

Bruzual & Charlot (GALAXEV)

CIGALE : SED fitting code



GALEV : useful site


~ by ilsangyoon on April 3, 2011.

2 Responses to “Bayesian MCMC galaxy SED modeling”

  1. Hi, glad to hear about this idea and I am very very interested about it !

    I used to work on stellar population with early type galaxies and tried to compare the two method of direct fitting and bayesian, but the extremely large parameter space is hard to handle since I’ve no experience of MCMC, and I can only see it works on SED and spectral index, but not the spectrum itself. But I guess this should be no problem for your method, right?

    I guess you must saw this one: , so what kind of improvement do you expect in your work? more than two population model or more realistic SFH?

    Hope to hear more about this in the future !

    BTW: The MILES is new but seems like pretty promising library:

    • Thanks for the comment. People start to be getting convinced that SED modeling can be improved by MCMC approach. I posted some more thoughts about this new direction using full Bayesian approach. The potential impact of this new approach looks very strong as we get more data from large scale survey especially at high redshift.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: