Open clusters offer valuable insight into many other areas of astronomy. Containing stars of the same age and composition, open clusters are ideal populations to study stellar evolution; they can be used as Galactic tracers - such as finding the rotational curve of the Milky Way - and are the basis for many other fields in astrophysics. Difficulties arise in using open clusters as tracers in that all information must be derived from other works which typically only consist of a few clusters for any given study. The result of this being that there is no uniformity as each study uses different photometric systems and different models. To rectify this issue, we are applying one set of models to the color-magnitude diagram (CMD) of one photometric system to determine age, metallicity, reddening and distance.
To do this, we chose the Girardi et al. (2002)
isochrones as our models and the 2MASS near-infrared survey as our
photometric system. As the isochrones did not have evenly spaced data
points, we modified the models by interpolating the main sequence to
fill in the gaps so the strongest concentration of the cluster would
align with the most populated part of the isochrone.
The age is determined by the shape of the isochrones. As the cluster
ages, the main sequence turn-off changes and moves further into the
red; there are 74 ages ranging from 6.6 to 10.25 log years in 0.05
increments. The animation on the right is showing an isochrone through
five log ages: 6.6, 7.6, 8.6, 9.6 and 10.25.
The metallicity of the cluster also changes the shape of the
isochrones. The animation on the left illustrates the subtle effect of
a change in metallicity for four ages, 7.8, 8.8, 9.8 and 10.25 log
years, through three different metallicities; sub-solar
(metal poor), solar metallicity and super-solar (metal rich).
The reddening
and distance of an open cluster is determined by the horizontal and
vertical shifts. The reddening is determined by both a vertical and
horizontal shift to fit the isochrone to the CMD. An example of this
shift for the open cluster NGC 188 is in the image below. The image on
the right is the isochrone placed with no reddening; the image on the
left is the isochrone with a reddening of E(B-V) = 0.082, a very low
value. The reddening is caused by interstellar extinction which
describes the absorption and scattering of light by the interstellar
medium (ISM).
For each cluster, the fitter will fit it to each age isochrone for all seven metallicities. The fitter will determine the best fit metallicity for each age; the final output is the best age and metallicity. The importance of membership probability, both 3-D proper motion members and radial velocities can be controlled with the fitter.
The image on the right shows the fitter results for four clusters - IC
4651, NGC 2682, Kharchenko 1 and NGC 1662 (clockwise from the upper
left image). The image can be enlarged by clicking on it; the larger
image opens in a new window. The grey dots are the stars; the blue triangles are
radial velocity members and the black dots are proper motion members.
On the images, the red fit is the best-fit isochrone with membership consideration, the blue isochrone the best-fit without membership consideration and the green isochrone is the previous catalog fit given by Dias et al. (2002).
Kharchenko 1 shows the most dramatic difference from the catalog fit.
For NGC 1662, the fit without a membership condition fails as it matches the main sequence to the area with the highest concentration of stars, but misses all the proper motion members.
For both IC 4651 and NGC2682, the main sequence of all three fits aligns very closely. At the MS turn-off the catalog fit and the new fits begin to diverge. The improvement of the catalog fit is much more apparent for IC 4651 where the turn-off differs about half a magnitude. The improvement of the fit for NGC 2682 is more subtle.
The image to the left is a magnification of the turn-off; a larger image is linked which opens in a new window. In the magnified image the new fit is clearly an improvement on the catalog fit as it fits the sub-giant branch much better.
After some final tests and checks to maximize the fitter results, we will be running clusters through the fitter.The percentage of binary stars in open clusters is a largely unknown number as only a small number can be determined by the current established methods of identification. The binaries that can identified are visual binaries, where the separation between the two stars is observable through a telescope, eclipsing binaries, which have an orbit in the same plane as the observer's line of sight so the stars eclipse each other, and spectroscopic binaries, in which the identification is biased towards close binaries with short periods as the changing of the radial velocity can be ascertained through multiple observations. We are planning on using the Color Tracks theory of Hurley and Tout (1998) to photometrically identify binary stars within open clusters. The Spectral Energy Distribution (SED) Fitter, which was written by Tom Robitaille of St. Andrews University in Scotland, can be used to identify single stars by the fluxes through different filters. Using the Color Tracks theory, we are compiling a new model grid to identify binary stars that have previously been unresolved.
The Color Tracks theory of Hurley and Tout (1998) considers the color
shift of binary systems of two main sequence stars of different masses.
The image on the right illustrates this phenomenon, where each black
dot is a binary of different mass ratio between the primary and secondary stars with the mass of the primary
star as labeled for each group of six binaries. The main sequence single stars
lie on the bottom black line and the binaries made up of two equal mass
MS stars lie along the top black line, these binaries being 0.75 magnitude brighter than the MS
single stars. The MS and equal mass binaries create the "tracks"
of the theory and we're interested in finding the stars within these tracks. The image on the right (click to enlarge) shows that the variance of the secondary star as compared to the primary from negligible to equal mass, the color of the star becomes redder in addition to becoming brighter. This is true because lower mass stars appear redder than higher mass stars.
To identify binary stars photometrically, we are using the SED Fitter created by Tom Robitaille of St. Andrews University.
An SED
is the plot of flux intensity or brightness versus the wavelength of
light; the intensities in different wavelengths vary by temperature,
metallicity and surface gravity (size). As we are looking at stars on
the main sequence, the temperatures of stars are directly related to
the masses. Below are two hypothetical SEDs;
the left is a hot star and the right, a cool star.
As can be seen in the image, cooler, less massive stars have lower intensities in the visual
spectrum, but higher intensities in longer wavelengths, making the fluxes in the mid-infrared - the Spitzer Space
Telescope IRAC bands - very useful for identifying these stars.
The SED
Fitter identifies the type of star by comparing the fluxes through
certain wavelengths to a grid of models. We created a binary grid by
combining two MS stars ranging from a zero mass secondary star - a single star - to an equal mass secondary star.
To the right is a crude illustration of a binary made up of two MS stars, a massive, hot star and a cooler, less massive star. The SEDs of the two MS stars stars are shown in gray and the composite binary SED in black. As can be seen in the image, the binary has a very small difference in the visual bands from that of the primary hot star SED, but the binary has greater intensity in the infrared bands.
We are using NGC 188 as our prototypical cluster. It is a very old, well studied cluster so it allows us to check that we are properly identifying binary stars where there are overlaps in the photometric and spectroscopic methods. We combined the photometric data from multiple surveys: the optical (UBVRI) data set of Stetson et al. (2004), the near-infrared data (JHKs) from 2MASS and the mid-infrared (3.6, 4.5, 5,8, 8.0 micron) data come from the Spitzer IRAC camera.
After a few more investigations of the binary models, we will be able to start running NGC 188 through the fitter and investigating the identified binaries.In January 2008, I presented a poster of the preliminary results produced by the fitter in at the 211th meeting of the American Astronomical Society (AAS) in Austin, TX. View a pdf version of the poster.
I am finishing up my undergraduate career at Colby College where I am a Physics major and Italian Studies minor. After taking
Introduction to Astrophysics in Spring '06, I became interested in
learning more about astronomy which led me to an REU at UW - Madison.
When not busy with physics and math courses, I enjoy my involvement with the Colby Music Department taking lessons and playing piccolo and flute for the Colby Symphony Orchestra. I also enjoy learning interesting new trivia - especially related to maps and geography. When not taking physics and math, I like to apply my map-related trivia by traveling. This year I am looking forward to living in Bologna, Italy and studying at the Università di Bologna in the fall of 2007.
I was very fortunate to be advised by and study with Dr. Peter Frinchaboy and have the support and guidance from the WOCS group - Dr. Robert Mathieu, Aaron Geller and Ella Braden - this summer; I am very grateful for all the assistance that they gave me.
I would especially like to thank Ed Miekiewicz for being the best REU Director that one could have.
I would be amiss to forget to mention my partner in crime and clusters this summer, Natalie Gosnell.