Yan, Renbin

Group Research Activities:

Our group works on several research projects about galaxy evolution. We study the stellar population,

interstellar medium, active galactic nuclei in galaxies using observational data obtained from spectroscopy

surveys of stars and galaxies. Relevant to this application is a project in which we are building a large

stellar spectral library for a wide range of applications in astronomy. Stellar spectral libraries underpin

spectroscopic analysis in stellar, galactic, and extragalactic astronomy; they are fundamental for modeling

stellar populations in galaxies. We have obtained a total of nearly 90,000 spectra, making it the largest

empirical stellar library collected to date. The spectra have high signal-to-noise ratios, cover a wide wavelength

range with well-calibrated spectrophotometry. The sample covers a wide range of stellar types. In order to

facilitate the construction of stellar population models, we need to determine accurately the stellar parameters

for each star we targeted. This requires the computation power provided by the HPC.

Fitting the Spectral Energy Distribution of Galaxies

We are trying to understand the star formation history of galaxies, by decomposing

the population of stars in a galaxy into multiple stellar populations with different ages.

In practice, we try to find the linear combination of a large number of templates to match

the observed photometry and spectra indices. We have 12 unknown parameters and we

run an MCMC search in this 12-dimensional space. Thus, we require the high-performance

computing resource to conduct the calculation efficiently.

Students:

Nikhil Ajgaonkar, Graduate RA, Added 03/29/2021

Computational method:

Markov-Chain Monte Carlo using the emcee package.

Software:

Python (3.7 preferred)

Required python packages: numpy, scipy, matplotlib, random, astropy, pdb, os, math, time, multiprocessing

External package: emcee

Stellar Parameter Determination for the MaStar Spectral Library

For each spectrum in our library, we need to provide accurate stellar parameters: temperature, surface gravity,

and elemental abundances. This is achieved by fitting each spectrum to a huge grid of theoretical templates. To

efficiently sample the parameter space and interpolate between model grid points, we have written a custom

code that will run a Markov-Chain Monte-Carlo algorithm to find the best fit. This takes about a minute for each

spectrum on a single CPU. With a large number of spectra, we would like to use the high performance cluster in

order to complete the calculation in a reasonable amount of time.

Personell:

Renbin Yan, faculty, (PI)

Daniel Lazarz, Graduate RA

Computational Methods:

We use a custom code we developed. The main algorithm is Markov-Chain Monte Carlo.

Software:

Python (2.7 preferred)

Required python packages: numpy, scipy, matplotlib, pandas, astropy, pdb, os, math, time, pdb, requests, json, tqdm

External package: dustmaps (https://dustmaps.readthedocs.io/en/latest/) (require both bayestar2017 and bayestar2019 versions).

Collaborators:

No collaborators non graduate students yet.

Grants:

Publications: