Документ взят из кэша поисковой машины. Адрес оригинального документа : http://star.arm.ac.uk/~csj/nuffield/reports/anne_oleary_report.doc Дата изменения: Thu Aug 14 18:36:42 2008 Дата индексирования: Tue Oct 2 13:19:07 2012 Кодировка: Поисковые слова: binary star

Observation of The Binary Star NN Ser

Using the Faulkes Telescope





ABSTRACT


My project is based on the research carried out during a five week
placement at Armagh Observatory. I used the Faulkes Telescope in Hawaii to
observe several eclipses of the binary star NN Ser. This was recorded by
CCD camera, and the images were then used in photometry, to find the change
in magnitude of NN Ser over time. From the values gained through
photometry, I plotted a light curve using Excel. From this light curve I
was able to find the length of an eclipse. The Faulkes Telescope was used
to capture images during the eclipse of NN Ser in order to find out how
accurately the telescope could time the event.



[pic]
The Faulkes Telescope North at Sunset
Image Courtesy of the Faulkes Telescope Project


INTRODUCTION




The star NN Serpentis (NN Ser) is, in fact not one star but two. Known as a
binary star, it consists of two stars in the later stages of their life-
span. One star is a white dwarf: a very small, dense body of carbon. As it
is not undergoing nuclear fusion, it is the star's residual heat (cooling
from a temperature of 100 million degrees) that causes it to emit light.
The star is so dense that, although it has a mass equal to half of that of
the sun, it is only twice the size of the earth. The second star is a red
dwarf, a larger (one and a half times the size of Jupiter) but much less
dense star in which hydrogen undergoes fusion at a very slow rate. Thus,
although it is an active star, it emits much less light than the white
dwarf and is often described as a cool companion star. However, as the
stars are so far away, they remain unresolved (i.e. they appear as a single
point) and any interaction between the stars is observed as a change in the
light emitted.
[pic]

NN Ser is shown circled


NN Ser can be described as a variable star. This means that the light
emitted by the star does not remain constant. This is due to the fact that
the stars orbit one another and regularly eclipse one another relative to
the earth. The more noticeable eclipse occurs when the dimmer star passes
in front of the brighter one, blocking it from view and significantly
reducing the amount of light reaching the Earth. In the case of NN Ser, the
eclipse is total as one star is completely blocked by another. The light is
reduced to such a degree that the star appears to all but the most
sensitive telescopes to have vanished altogether. The period of the orbit
of the stars is 187 minutes, and the eclipse lasts for approximately
11minutes.

The timing of an eclipse provides a useful method of determining an
instrument's abilities. As the length of the eclipse of NN Ser is known, it
can be used to gauge the capabilities of a new telescope. In this case, the
telescope used was the Faulkes Telescope North which is located on the
mountain of Haleaka, on the Hawaiian island of Maui. There is also Faulkes
Telescope South, which is in Australia. The Faulkes Telescopes are research
quality reflecting telescopes which were constructed to make research grade
resources available to schools and students. The telescopes can be
controlled live via the internet.



[pic]
The Opening of the Telescope Enclosure at Dusk

Image courtesy of the Faulkes Telescope Project







METHODS AND MATERIALS


The Faulkes Telescope is equipped with a CCD (charge coupled device)
camera, which produces digital images that can be used for photometry. This
contains a piece of silicon which detects light through the photoelectric
effect, which causes it to emit electrons when light falls on it. These
electrons build up in individual photosites, and during processing, the
number of electrons is counted and is related to a certain intensity or
brightness of light.




Prior to my observations, I used the programmes Nightfall and Starlight pro
to model the light curve for NN Ser, based on the stars' size, temperature
and other characteristics of the binary system. This allowed me to consider
how I would handle my data after my observations were made.


Before observing an eclipse, I had the opportunity to collect some test
frames of NN Ser. A range of different filters and different exposure times
are used in order to find out the best exposure time for my observations.
The shorter the time of exposure, the more images can be taken during the
eclipse. This means there are more points on the light curve which makes it
more accurate. As the length of the eclipse is found from the light curve,
a more accurate curve means a more accurate value for the length of the
eclipse is found. However a shorter exposure has a worse signal to noise
ratio.

I chose the r' filter as it seemed to be the most sensitive for observing
my star, i.e. an exposure with this filter produced better images than
exposures of the same length with other filters. A ten second exposure was
chosen as a compromise between exposure time and a good signal to noise
ratio.

When the contents of the silicon wafer in the CCD camera are processed, the
images first become available as JPEG files. JPEG files are a form of
compressed data, which makes it possible to quickly upload them to the
internet so that they can be viewed by the student. However, during the
compression process, much information is lost, and the files therefore
cannot be used for photometry, as they do not contain the full range of
values for the light absorbed by the photosites. Therefore, in order to
analyse my data, I required a file format known as FITS files.
Once I had received my FITS files, I used AIP4WIN to measure the change in
magnitude of NN Ser during its eclipse. This is done by comparing the
brightness of NN Ser to other stars in the image. In order to minimise
errors I used a check and comparison star to ensure that any variation in
magnitude was due to variation in NN Ser and not the in star I was
comparing it to.

This is a screen shot of AIP4WIN being used for photometry.

I then transferred the results of photometry to Excel where I plotted a
light curve. In order to plot a curve containing data from several
different sessions on the telescope, it was necessary to plot change in
magnitude against phase rather than against time.









This is done using the following equations:
teclipse = T0 + P.n
Ь = (t - teclipse)/P

Where:
teclipse is the time in Julian days when the star eclipses.
T0 is the ephemeris, a known time in the Julian calendar when the star
eclipsed.
P is the period of the star.
n is the number of cycles that have occurred between the two times.
Ь is the phase of the eclipse.

In order to use these equations, the dates and times of the observations
had to be converted to Julian days. Then an approximate value for n was
found by rearranging the formula and using a time from my observations
which was estimated to be in the middle of an eclipse. This value for n was
then rounded to give the integer number of cycles that had occurred. From
this, an accurate value for teclipse is found. Using the second formula,
the times of my observations were converted to phases.


Because the earth is in constant orbit around the sun, the distance from
the NN Ser to the earth, and therefore the time taken for the light to
reach the earth constantly varies. In order to give an accurate prediction
of times, many prediction times and records of events are given as the time
they occur as viewed from the sun or barycentre. This creates a difference
between the given time of the event and the time the event is observed from
earth. This difference varies between my sets of data as the earth's
position relative to the sun has changed. The Barycentric corrections
involved taking away a certain number of seconds from my observed times in
order to convert them to times as seen from the sun.
Time difference = t1 - t2


t1 = d/c earth


X
Star t2 = d/c

Barycentre/sun

RESULTS


I have included a film strip of NN Ser during the eclipse; this is made
from images captured during my first session on the Faulkes Telescope.

The light curve plotted by Excel is shown overleaf. In some places, data
was missing. These points occurred during the eclipse and so I have
replaced them with a value of ten, the maximum value that could reasonably
be expected.

























CONCLUSIONS


In order to explain some of the problems with my results, I have included a
sample light curve for NN Ser created with Excel using the information I
obtained with Nightfall Pro. I am using this as an example of what an ideal
set of data would look like


When comparing the sample curve with the curve I obtained, several problems
are highlighted. First is that the curve is not flat during the eclipse.
This can be explained by looking at the process of photometry. The star's
drop in magnitude was found by comparing NN Ser to other, constant stars in
the image. A drop in magnitude or dimming is found when fewer photons are
recorded per pixel. However, with fewer photons, a drop in accuracy occurs.
In this case, the drop in magnitude was very large and so the number of
pixels recorded was very small, therefore the percentage inaccuracy was
very large. These inaccuracies probably lead to unevenness in the light
curve.

It is also seen that the light curve is not centred at phase equal to zero.
This is due to the fact that I have not included Barycentric corrections in
my results. I found that when I attempted to do this, the data correlated
less and that it did not become significantly centred. The problems with
the Barycentric corrections indicate that the timing mechanism of the
telescope may require further investigation.











EVALUATION


If I were to repeat this project, I would increase the length of the
exposure times during the eclipse. The exposure time at ten seconds during
the eclipse resulted in large errors in photometry. As there were fewer
photons recorded, the percentage error was very great. In some cases, I did
not receive a value, perhaps because the software did not recognise that
there was a star present at that point in the image. An exposure time of
thirty seconds would ensure that the photometry was much more accurate.


The Faulkes Telescope provides a hugely valuable resource to students who
wish to carry out research projects. The service provided to students could
be improved by altering the interface which allows students to control the
telescope. If two half hour sessions could be recognised as one hour-long
session, the student would benefit from a continuous session of
observation.



















ACKNOWLEDGEMENTS


I would like to thank all the staff at Armagh Observatory for their
patience and help.
I would especially like to thank: Dr Simon Jeffery, my supervisor, who did
everything possible to help me with my project; Amir Ahmad who helped in Dr
Jeffery's absence; and Apostolos Christou who also sacrificed his time in
aid of my project and my general education in astronomy.

I would also like to thank the staff at the control centre for the Faulkes
Telescope and the staff in Hawaii for their help in making the telescope
available to me.