De Lee, Nathan Michael (NKU)

Modeling Stellar Companions in SDSS

The Sloan Digital Sky Survey (SDSS) is an international collaboration of astronomers including faculty at University of Kentucky (e.g. Renbin Yan and Ron Wilhelm). In this project, we are exploring how well we can model and detect stellar companions using SDSS data. Stellar companions include stars, brown dwarfs, and planets. One important aspect of this modeling is to understand how well we can recover simulated companions that are designed to look like data taken from the first and second generation of the Apache Point Observatory Galactic Evolution Experiment (APOGEE and APOGEE-2) which are sub-surveys of SDSS. We are testing a variety of different algorithms to figure out which ones provide the best balance of speed versus accuracy. When exploring the chi-squared space of a fit to radial velocity orbital data, the orbital period is one of the most important orbital elements. As a result, we are using period finding algorithms including Lomb-Scargle, AoV, and Fast Chi-squared. We are also trying Markov-Chain Monte Carlo rejection sampling techniques which are slower, but tend to be more accurate.

Software:

Modeling these simulations is primarily done in Python 3. The major python modules that we need include astropy, emcee, numpy, pymc3, exoplanet, scipy, matplotlib, corner, thejoker, phoebe, and schwimmbad. These modules can be installed via anaconda or pip. We also use a suite of tools written in C and Fortran called vartools 1.38. All of these modules, algorithms, and software packages are freely available. We also use the IDL language, which is commonly used in astronomy. We will use the freely available EXOFASTv2 software package. Although if IDL is not available, we can work without it.

Personnel:

PI Dr. Nathan De Lee (NKU)

Students:

NKU Undergraduates:

Kendra Herweck

Tim Faller

There is no one at UKY that works on this particular project.

Grants:

Grant No. 1616684. NSF grant, A survey of the types of stars and massive planets that make up binary pairs in the Milky Way Galaxy, provides in $85,693 in support.

Detecting and Characterizing RR Lyrae and Other Variable Stars in KELT

The Kilodegree Extremely Little Telescope (KELT) survey is a transiting planet finding survey that can also be used to find and characterize RR Lyrae (RRL) stars in the disk and inner halo of the Milky Way galaxy. RRL stars are of particular interest because they are standard candles and can be used to map out structure in the galaxy. The KELT survey represents a new generation of surveys that has many epochs over a large portion of the sky. KELT samples over 60% of the sky in both northern and southern hemispheres, and has a long-time-baseline of 4-10 years with a very high cadence rate of less than 20 minutes. This translates into 4,000 to 10,000+ epochs per lightcurve with completeness out to 3 kpc from the Sun. The computational side of this project involves determining ways to separate the RRL stars from all the other variable objects. We are exploring a number of avenues to accomplish this including doing cuts in photometric color space, limiting period and amplitude space, and using machine learning techniques.

Software:

There are on order 4 million lightcurves available in the KELT database. This project has two goals: 1) Use a training set of ~1400 lightcurves to develop an effective detection algorithm 2) Apply this detection algorithm to the 4 million lightcurves. These algorithms will use Python 3 with the following modules: astropy, emcee, numpy, pymc3, scipy, matplotlib, corner, and schwimmbad. These modules can be installed via anaconda or pip. We also use a suite of tools written in C and Fortran called vartools 1.38. All of these modules, algorithms, and software packages are freely available.

Personnel:

PI Dr. Nathan De Lee (NKU)

Collaborators:

I have worked with Dr. Ron Wilhelm at UKY on other RR Lyrae studies that involved KELT lightcurves, so he may be interested in working with me on this.

Grants:

There is no grant support for this project at this time.