Astro 122 - Nightlab Correction Project
Observations of Long Period Variable Stars
The objects of the exercise are to:
1. appreciate that the sky is not unchanging;
2. examine directly part of the way in which most stars die;
3. explore how scientific data are acquired and used.
these ends, you will look at the variations of a "long-period
variable" star, otherwise known
as a "Mira" variable (named after the first such star discovered).
These are giant stars that undergo huge
variations in brightness over a period of a year or so (can range
from about 100 to 1000 days).
Mira variables are dying stars with carbon-oxygen cores that have
lost their equilibria and are pulsating and changing their
brightnesses. They are in this state for 100,000 years or so. At
the same time they are losing matter through massive winds, and
will eventually lose their entire outer envelopes, nearly exposing
their old nuclear-burning cores. At the end, they will produce
planetary nebulae that are seen as shells of gas surrounding hot
stars, and then will die as white dwarfs.
Your job is to use Stardial to observe the Mira star RR Aquarii
(in the constellation
All the images Stardial has collected since 1996 are available
on-line: you can browse them by date, or by location on the sky.
For this experiment, you will need to examine images near the
location Right Ascension = 21h 15m, Declination = 0.
(To better understand celestial coordinates see page 27, Box 2-1, in your text.)
Here's how to find them:
- Go to the Stardial archive of JPG images:
- Choose to look by Right Ascension by clicking on the
- Scroll down the list of RA positions to
2115, which stands for Right Ascension = 21 hours, 15 minutes.
- You will see a choice of years: 1996 - 2006. If you select one
year, you will be presented with a list images from individual dates
from that year; for example, 0620.jpg
is the image taken on June 20.
- Each image has a Size listed with it. Good images, free from
clouds and other problems, are about 20K to 30K in size.
Bad data will be obvious.
You can look at several images of other sizes to get an idea
for the range of weather conditions in Chambana.
When you select any particular image, your WWW browser should show
a picture something like this (but without the red box):
Finding the variable star
The red box in the picture above outlines an area in the constellation
Aquarius which contains our variable star RR Aqr. Your first
job is to figure out which of the stars is the variable.
Here are two Stardial images of the field, one taken when the variable
was at its brightest, the other when it was at its faintest:
Can you find the star which changes its brightness?
Astronomers use a method called "blinking" to search for
variable stars: they alternate rapidly between two images,
displaying registered versions of each for just a second or two.
You can "blink" the two images above
by selecting the link below.
You will see one image by itself; click on it to display the other
image. Keep clicking (or using the Alt-Left and Alt-Right buttons)
to alternate the images until you find the variable.
Click on the text at the bottom to return to this exercise
when you have found it.
Click here to reach blink pages
Note that the star RR Aqr is varying while the other stars are not
varying. (If all the stars varied together in synchronism, we'd
suspect that clouds and/or equipment malfunction were the cause.)
Making a light curve
You will to construct a "light curve" for RR Aqr, in which you make a
graph of apparent magnitude plotted against time. Magnitude measurements
are to be made using the oldest device, the human eye.
Now that you've found the variable star, you must measure its
change in brightness over an entire season.
You can compare the appearance of RR Aqr against those of
stars nearby to gauge its apparent magnitude.
Below is a chart of stars inside the red box
(perhaps you'd prefer a
of the chart.)
The chart has reversed the usual color scheme: astronomers
often use a "negative" portrayal of the sky as white and stars as dark
in publications, both because it's easier on printers, and
because most people can discern faint details more clearly.
The chart provides apparent magnitudes in the "V" passband (i.e. visible)
stars in the area.
As you examine images from the Stardial archive,
compare the appearance of RR Aqr to these eight stars.
If it looks just like one of them, then
its magnitude must be close to that of the star;
so if RR Aqr looks just like star "D", it would have a magnitude of 8.1.
On the other hand, if RR Aqr seems to fall between two of
the marked stars, its magnitude must be between those of the
two stars; for example, if it looks fainter than star "E", but
brighter than star "F", then its magnitude must be between 9.4 and 10.0:
RR Aqr is
found in the field with the right ascension of 21 hours and 15
minutes, or 2115. Go to the
find the archive that is organized by location (the right ascension, RA,
mode, under "web" and "jpg")
and find your the "2115" field (the one that
contains your variable star). Pick
data from 2006, 2005, and 2004 (and 2003 too, if you wish to add more
Also be sure to pick clear nights
(note: if the other stars look very different in an image,
that means it was not a clear night).
This is not a project with canned data. We do not know what the star
will do, whether it will brighten, fade, or how fast it will change.
You will determine that by your observations.
You should have a minimum of 30 observations
spaced throughout the three years (or more than that for four years).
This is actually the way
variable stars are examined, except that professionally, electronic
measuring devices are used; generations of amateurs, however, have
made useful observations using the "naked eye" technique at the
For each image you choose,
- calculate a time for the image: convert the date of the image to
T = number of days since your first data point.
It might help to have a calendar ...
- compare the brightness of RR Aqr against those of the stars
with marked magnitudes
- write down in a table the date of the image, the time T,
and your estimate for the apparent magnitude m of RR Aqr
Now, turn your table of measurements into a
observe, record your data in a table and then plot your data on a
graph with the date (i.e. day) on the bottom axis (the x-axis) and the magnitude
increasing downward (so fainter is down) on the side (the y-axis).
Plots should be neat, and the use of a computer
graphing package to plot your data is strongly recommended.
At the end, connect your observations with a smooth curve;
this may be done by hand.
You will notice that
there are several months of each year with no data recorded
for RR Aqr.
Explain why this is the case. Hint: this is not
an error in the system, or an issue of computer disk space,
but rather to do with issues related to the celestial sphere.
What do you
estimate the period of the star's oscillations to be?
Note that the data gaps mentioned in Question 1
complicate this estimate. Is there a way to get around
Based on this period, estimate how bright RR Aqr will be
on August 3, 2006.
What is the amplitude of the variation in RR Aqr -- that is,
by how many magnitudes does it vary from minimum to maximum?
Convert the amplitude from magnitudes to intensity:
how many times brighter is RR Aqr at maximum light than minimum?
Pretend for a moment that the outer layers of RR Aqr don't
heat up and cool off, but maintain a constant temperature
as they expand and contract.
If that were the case,
then the brightness of the star would depend simply on its
surface area: the larger the diameter, and larger the surface
area, the brighter the star.
If RR Aqr varied only in size, what would be the ratio of
its diameter at maximum light to its diameter at minimum light?
Would a citizen of a planet around RR Aqr
find the change in the angular size of his/her/its star
measuring device is perfect, and all measurements necessarily contain
errors. When you plot your graph you will see that the points do not
exactly follow a smooth path. The deviations of your points from the
smooth curve indicate your typical "error of measurement," which
should always be cited. Estimate your typical errors from the graph
before turning it in. There are formal ways of assigning errors, but in this
assignment you need only make your best estimate.
Reports are due on April 12th in your discussion class.
Include in your report:
- a table of measurements that includes the date and the
- a graph with magnitudes increasing downward;
- your error estimate (the size of a typical error);
- your answers to questions 1 through 7 above.
Fri Apr 8 14:49:22 CDT 2005