Finding RR Lyrae in Globular Cluster - M 13
What is a Globular Cluster?
A globular cluster is a
gravitationally bound collection of stars. These generally consist of about
10,000 to 1,000,000 stars and span anywhere from several tens to 200 light
years in diameter.
What is M 13?
M 13 is a globular cluster in
Hercules, its location in terms of R.A. and Dec being 16 41 41.44 and +36 27
36.9 respectively. It was first
discovered by Edmond Halley in 1714, who on discovering this Globular Cluster
said, ¡°It shows itself to the naked eye when the sky is serene and the moon
absent.¡± M 13 is estimated to have anywhere from 100,000 to over a million
stars. The center of M13 has more than 500 times the stars in the solar
neighborhood. Nine of these stars have been identified as variable stars called
RR Lyrae. M13 spans about 150 light years across in diameter and is about
20,000 light years away from the Earth. It is over 12 billion years old, which
is not surprising as globular clusters are among the oldest objects in the
galaxy.
What are Variable Stars?
These are stars that undergo
significant variations in Luminosity. There are two main kinds of variable
stars ¨C Cepheid and RR Lyrae. Since we are concerned with RR Lyrae here, we
will be talking further about these.
What are RR Lyrae?
RR Lyrae stars are pulsating
Horizontal Branch stars, that is, they change their brightness with a regular
period and are usually found in globular clusters. They were named so after the
first star discovered of this kind.
The period of pulsation for
these stars is relatively small ¨C from 8 or 9 hours to 1.2 days. This makes it
possible for us to observe changes in magnitude over a period of few hours of
observing.
Why do RR Lyrae pulsate?
The pulsation is caused due to
the fact that the star enters a physically unstable state internally causing
its surface to move in and out causing it to change in terms of size and
brightness. The brightness of a star depends on its temperature and size. As
the size contracts, the gas is getting compressed into smaller volume leading
to increase in temperature and as the size increases, the pressure is released
and temperature falls. In the case of RR Lyrae the rate at which the size is
changing is not fast enough for the amount of change in brightness that is
observed. It is the effect of change in surface temperature along with changes
in size that affects the brightness of RR Lyrae stars. So, when the star is
smallest and hottest, it is also the brightest. Converse of this is also true.
So starting with a star that
is in compressed or contracted state, the He+ in the atmosphere of
the star ionizes into He2+. This leads to the atmosphere being less
transparent and more opaque. This traps the energy flux within the envelope and
as the gas gets heated the pressure builds up. When this pressure exceeds the
internal gravitational force of the star, the star starts expanding. As the
envelope expands, the atmosphere begins to cool down and the He2+
gains an electron to become He+ again, leading to increase in the
transparency of the atmosphere, thereby allowing for the pressure to decrease
and the star to shrink back.
Why are RR Lyrae useful?
We can study the distance to
the globular clusters by comparing the apparent magnitudes with the absolute
magnitudes of RR Lyrae stars residing in them. If we know the period of RR
Lyrae stars to go from brightening to dimming, then we can find the absolute
luminosity of the star, using the Period ¨C Luminosity relationship. This
information then can be put into the Distance Modulus to give us the distance
to the RR Lyrae, and hence the distance to M13. Since, the periods of these
stars are not very long, these are ideal for our project.
Project A
We started off by wanting to
pick one RR Lyrae star in M13 and making its light curve, to then calculate the
distance to this star and hence to M13. In order to do this, we needed to find
good finder charts, so that we could have good check stars and be able to
identify our RR Lyrae star, without any error. This took an unexpectedly long
amount of time and searching. Unfortunately, we were unable to find a good
finder chart for this Globular Cluster. Finally, we did find one. However, it
was not a very good one as even though it did show us M13 with RR Lyrae in it,
the distance scale was such that we were unable to understand and interpret the
chart. Wanting to take advantage of a clear night after a series of cloudy
nights, we decided to try and make do with our imperfect finder chart,
something we soon learnt one should never do. At about 7:30 PM, we started
taking twilight flats. After we were done, unfortunately someone accidentally
moved the camera, which rendered our twilight flats useless. We went ahead with
the observing anyways, deciding to use dome flats instead. Once we had got a
picture of M13 and wanted to identify our RR Lyrae, we realized how useless our
finder chart was. This and a series of bad luck with weather brought us to
reconsider the project. So we switched to Project B, which was to identify
known RR Lyrae in M13.
Project B
Now our project had turned
into essentially observing for a series of hours one night to try and get as
many pictures of M13 as possible at different times. This would enable us to
compare the magnitudes of different stars and determine the ones that showed
any changes in magnitude. These would be our RR Lyrae. We by this time had
found two great finder charts. So we knew the positions of the RR Lyrae stars
and just had to verify that they really were RR Lyrae, by checking for changes
in their magnitudes in our images.
One of the Finder Charts
Check stars in our image.
RR Lyraes in M 13 on our image. The brightest one at the top is RR Lyr 36, of
the couple the one to your left is RR Lyr 9 and the one to the right is RR Lyr
5. (Numbering according to our Finder Chart)
Observing
We were able
observe and get images of M13 on two nights. The first of these nights enabled
us to get an excellent image of M13, which can be seen above. Other than this,
the night was futile as we talked about earlier under Project A. The second
night we got to observe M13 from about 11:00 PM to 4:30 in the morning. Here
are some more details about our experience on these two nights:
On the first
night, we took 3 sets of twilight flats in the R, V, and B filters. These
however were useless as our camera was moved. We decided to go ahead with the
observing anyways with the intention of taking dome flats later. We then used
Arcturus to focus and sync the telescope. After focusing and synching we
pointed the telescope to M13. We took 1 set of 2-minute exposures in the R, B,
and B filters. We did not use track and accumulate. Also, during the course of
re-focusing and readjusting the telescope after taking the twilight flats, we
accidentally moved the camera, rendering our flat fields useless for our
images.
On the second
night, between 7:00pm and 11:00pm, the sky was cloudy. We were unable to take
twilight flat fields. At the end of our observation run, we took dome flats
instead in each of the R, V, and B filters. After about 11:00pm, the sky began
to clear. We pointed the telescope to Arcturus for focusing and synching. After
focusing and synching, we pointed the telescope to M13. We then took several
1-minute exposures of M13 to make sure that the telescope is focus. We re-focus
the telescope. We started with 5-minutes exposures in the R, V, and B filters,
and then took three sets 10-minutes in the R, V, and B filters. Between the
last set of 10-minutes, we waited for about 45-minutes to take a set of
6-minute exposure images using track and accumulate.
Flat fields are use to reduce
unwanted signal effects caused by sensitivities across the CCD surface and by
dust on the lens. A speck of dust will show up on a flat field image as a doughnut
shaped smudge. By dividing the final astronomical image by the flat field we
are able to remove the dust. The ideal flat fields are twilight flats. The
ideal time to take twilight flats is during dawn and dusk as the sun is rising
or setting. Twilight flats are taken by pointing the telescope at the sky away
from the sun as it rises or sets. If twilight flats cannot be taken, dome flats
can be use instead. Dome flats are taken by illuminating a screen evenly with
light and pointing the telescope to the screen.
Photometry
Once we had all our images of M 13
along with flats, our first step was to divide the images with the flats taken
in their respective filter. Once, we had the flat fielded images in all the
filters, we proceeded to identify our check stars and match their magnitudes
with the known magnitudes in each filter. Once, we had the images all
standardized, we tried to identify the RR Lyrae stars using a finder chart.
This turned out to be a lot more difficult as our finder chart was not drawn at
a scale similar to the image we had. Neither was it drawn to scale as far as star sizes were
concerned. So, we spent a lot of hours trying to find another finder chart,
with no result. In the end we stuck with the chart we had and tried to
approximate the position of RR Lyrae the best we could, with the hope of
narrowing down to a few suspect RR Lyrae and then checking for magnitude
changes in the different images taken over a period of 6 hours.
Check
Stars:
|
Star |
B |
V |
S/N
(in V) |
RF (V) |
|
Check
A |
14.07 |
12.64 |
95.84 |
372.51 |
|
Check
B |
14.66 |
13.42 |
88.00 |
372.51 |
|
Check
C |
14.38 |
13.22 |
87.27 |
372.51 |
|
Check
D |
13.91 |
12.31 |
87.23 |
372.51 |
|
Check
E |
15.52 |
14.30 |
80.72 |
372.51 |
|
Check
F |
14.30 |
12.94 |
88.28 |
372.51 |
RR
Lyrae Stars:
|
Star |
Change
in V (Observed) |
RF |
|
RR
Lyr 5 |
~
0.36 |
372.51 |
|
RR
Lyr 9 |
~
0.37 |
372.51 |
|
RR
Lyr 36 |
~
0.06 |
372.51 |
Note: Our Finder Chart for the RR Lyrae
can be seen at: http://arxiv.org/PS_cache/astro-ph/pdf/0211/0211042.pdf
Another method we used to further verify
and check our identifications of these variable stars was to take the different
images and subtract them from each other, sticking to the same filter. This way
all the stars with approximately same magnitudes will be subtracted and what
will remain will be stars that had more of a magnitude change. These can be
then checked for changes in magnitude and be labeled RR Lyrae accordingly. Here
are some of our images and how we worked through them:
.
Error Analysis
1. Readout Noise and Quantization
Noise: These depend on the CCD and cannot be improved by the observer. For our
ST-8 CCD, the read noise is 15 e -- per pixel per read. The later kind of noise
has to do with Quantum Efficiency of the camera.
2. There were small differences in the
magnitudes of the stars as compared to the known values. So certain error is
introduced in the magnitudes of all our stars. This may be quantified as a
function of some property of stars or if some pattern can be discerned. An
interesting observation is that the magnitudes in the B filter are higher than
the magnitudes mentioned in the finder chart for B filter, while the magnitudes
in the V filter are lower than the values given in the finder chart. This is
true for each star.
3. We were unable to start observing
until 11 PM, which is quite late. It would have been better if we could have
started observing right after sunset, to get as complete a cycle as possible
for the pulsation period of the RR Lyrae in M 13.
4. We had to use dome flats. It would
have been better to have sky flats.
5. Selection of the object can be very
important. We should have put more effort into finding a good finder chart or
selected an object that we knew to have a great finder chart. This can be
really rewarding when it comes to verifying your results.
6. Quantitative Error: The idea is to
identify the RR Lyrae and then using some distance conversions, try to quantify
the displacement of our stars from the finder chart and Java Appalet, thereby
determining something like a translation error in our positions of RR Lyrae.
7. We were not very good at
refocusing a few times, so a few of our images are out of focus, which leads to
elongated stars with fuzzy edges making it difficult to choose the toggle box
size while doing photometry. This ends up affecting out magnitude
determinations.