Justin Adkins
Eastern Kentucky University (EKU)


REU program-Summer 2008
Univ. of Wisconsin - Madison
Madison, WI 53706

Research projects of other REU students
Useful links
My conclusions
About Me

Observational Modeling of Stellar Populations


Isochrones are sets of single-age models of single star stellar evolution shown for a range of stellar masses. Since isochrones mimick simple stellar popultions, they are commonly used to derive the basic parameters (age, metallicity, distance, etc.) for star clusters. An isochrone is plotted on a Color Magnitude Diagram (CMD) as a trendline where the actual cluster data should lie. Isochrones are calibrated based on data taken generally only in the optical spectrum (e.g., UBVRI filters). A Color Magnitude Diagram (CMD) is just a plot of color(~temperature) vs brightness(luminosity), or an "observation" version of the Hertzsprung-Russell (H-R) diagram. New advances in technology and understanding have led to interest in further observation toward the infrared (IR) spectrum (e.g., 2MASS: JHK and Spitzer/IRAC: [3.6][4.5][5.8][8.0]). The study of galaxies (ours and others) makes observing in the IR necesaary, to see through the dust clouds that obscure the vision of the visible light filters. The GLIMPSE Survey at the University of Wisconsin is a great example of this. However, most isochrones have not been thouroughly tested in the IR and need to be checked for accuracy within these IR bands. If they don't match the data, then clearly some rethinking must be done for the isochrones to accurately model all data, as a good model should.

So why do it?

  • Provides a new analysis technique to improve studies of star clusters.
  • Test different isochone "families" to quantitatively show which are most likely correct.
  • Help standardize isochrone modeling for future use.

What I am doing right now

Fig. 1 - CMD showing one "best fit" isochrone over actual data points for NGC 188.

This summer I have been working on running an IDL-based CMD fitting program (made by Trevor Irwin) to see which of the many isochrones fit the shape of the data CMD best. This is revolutionary because often scientists "fit" isochonres "by eye", picking the first isochrone that works, and do not fully explore the parameter space. We are hoping to make the isochrone fitting more quantitative and uniform by implementing this fitting program.

Above is one isochrone plotted over the real NGC 188 CMD. This was done for each combination of filters in UBVRI[3.6][4.5], with R vs. (R-[3.6]) being pictured. Further data will be incorporated from 2MASS and deeper JHK bands, as well as data from SDSS (Sloan Digital Sky Survey; ugriz), at a later date.


After a long part of the summer dedicated to verifying the accuracy of the fitting program itself, initial tests of the Padova isochrones are currently underway, with initial results for 42 color-magnitude combinations using UBVRI[3.6][4.5] bands. The "best fit" parameters given by the fitting program were recorded and plotted in a histogram (Fig. 2) to see how well the filters matched the same parameters (or to find trends in their differences).

Fig. 2 - Histograms showing the "best-fit" for the 42 color-magnitude combinations using UBVRI[3.6][4.5]
as a function of Log(Age), Magshift (extiction corrected distance modulus), and reddening (E(B-V).
The Blue line denotes the fit from von Hippel & Sarajedini 1998, Sarajedini et al. 1999

More data and refining of the fitting program will be necessary to fully quantify any errors in the isochrones, but some trends are becoming visible already. One can visibly see the isochrones fitting on the CMD's look like they fit well. More importantly, the parameter distributions for age, distance modulus, and reddening all contained the accepted value well within one standard deviation. These are very good early indications that the isochrones will accurately predict populations based on infrared data. Using a cleaner dataset, more filters, and evaluating more isochrone families (Yale-Yonsei, Dartmouth, etc.) will give us a greater understanding about the accuracy of these models, as well as a better understanding of the underlying physics required to make correct models.

About Me

I am currently a senior at Eastern Kentucky University. I will graduate in May with a Bachelor's Degree in General Physics with a minor in Mathematics. I assist teaching in some astronomy lab courses and I also tutor for the Physics Department. I am required to do a Senior Thesis project for the Honors Program at EKU, and I am working with Dr. Jessica Lair on light-curve analysis for Type Ia Supernovae.

Useful links

The following links are very useful for looking up info on UNIX, web page making, and astrophysical data and journals.

SIMBAD (Stellar database)

SIMBAD (Stellar database)

NED (Extragalactic database)

UNIX tutorial

Web page basics

NASA Astrophysics Data Service