Transcript スライド 1 - STScI

Variable star survey in the Magellanic Clouds
-- Ten-years-long NIR monitoring observations -Yoshifusa Ita & IRSF/SIRIUS variable star survey team
email: [email protected]
Abstract
We started a near-infrared (JHKs) variable star survey in the Large and Small Magellanic Clouds in December 2000. Since then, we keep
monitoring an area of 3 square degrees along the bar in the LMC, and also an area of 1 square degree in the central part of the SMC. In the
last 10 years, we observed these areas about 80-90 and 100-110 times for LMC and SMC, respectively. As a result, we obtained time series
data with more than 3,000 days baseline for both clouds. Our survey is the world's first and only one that provides near-infrared time-series
data with such a long baseline.
Why in the NIR?
Our survey is operated in the near-infrared bands, so it can detect infrared very red variables that are difficult to detect in existing optical
surveys, such as OGLE (Udalski 2003). Figure 1 shows an infrared variable detected in our survey. It is obvious that the star is faint in the
optical, but is very bright in near and mid-infrared wavelengths. The mid-infrared image is taken by AKARI satellite.
Discovery Spaces / Finding limits
Figure 1
A panchromatic image of an infrared variable
detected in our survey.
1. What can we discover with our infrared survey?
→
I
J
K
11um
Figure 2 is the I-K vs. K color-magnitude diagram of variable stars in the LMC. Relevant data is taken from Ita et al. (2004). The horizontal line
shows the 10 sigma detection limit for our monitoring survey at K band, which is about 15.5 magnitude. The right diagonal line stands for the 10
sigma detection limit of the OGLE survey at I band, which is about 19.5 magnitude. Also, the left diagonal line shows the saturation limit of the
OGLE survey at I band, which is about 13.0 magnitude. Therefore, all of the variable stars in the middle area are not only detected by ours, but
also by OGLE optical survey. Our discovery spaces are red and bright part of this diagram. Our survey can discover infrared variables and also
bright variables. However, most of the bright variables may have been discovered by the IRAS-oriented studies.
2. Specifications of our survey.
Figure 2
A I-K vs. K color-magnitude diagram of
optical variable stars in the LMC
showing our discovery spaces.
Observations are made with 5 sec * 10 dithering mode
• 10 sigma detection limit is about 15.5 mag in K-band.
How bright/faint variables can our survey detect?
•Bright detection limit: saturate at 8 mag in K-band
•Faint detection limit: depends on the amplitudes of variable stars, and also detection limits.
+We certainly detected classical cepheids with ΔK of about 0.3 mag that are about 15 mag at K (with pulsation
period of a few days).
-Very sadly, we could not detect extremely red objects with tau_v ~ 200 (discoveried by Spizer/SAGE survey,
Gruendl 2008) whose K band luminosities are fainter than 17 mag.
Tentative results from the SMC survey
We developed a period determination software that uses lomb-scargle periodgram. We calculated
æ
1
ç
P(w ) =
2å1/s i2 çç
è
Figure 3
Tentative Period-K magnitude relations for
SMC variables detected in all JHK bands
with their false alarm plobability is less than
1 percent.
K-band
i
i
2
2
cos
w
t
/
s
å
i
i
i
i
2
i
2
2
sin
w
t
/
s
å
i
i
2
ö
÷
÷
÷
ø
for trial periods (2π/ω) from 2 to 2000 days for each detected variable. We took the 1st, 2nd, and 3rd most
powerful signals (P(ω)) if their powers have false alarm probability of less than 1 percent.
Figure 3 is the tentative Period-K magnitude diagram made by using the software. This is the result from fully
automated analysis, and we did not eliminate any spurious observational data and so on. The K band
magnitudes in the ordinate are calculated by 0.5*(K_min + K_max), and not the time-averaged ones. We will
calculate the time-averaged mean magnitudes and that would improve the quality of the P-L relations.
Discovery! I&J-drop variable
in the SMC
H-band
(å X coswt /s ) + (å X sinwt /s )
2
i
2
Variability in the HR diagram.
Figure 4
CMD of the all source detected in the 1 sq. deg survey
area in the SMC(left panel ) and their chance of
variability(right panel). Note that red AGB stars are
100% variables.
Figure 4 shows a color magnitude diagram of all sources
detected in our 1 square degree survey area in the SMC
(left) and their chance of variability (right). As you can see,
light variation phenomena is ubiquitous in the HR diagram.
The color of the left figure shows the percentage of
variability in the 0.05mag x 0.05mag bin. I counted all
sources in a bin, and then counted the number of variable
stars in that bin.
It is quite interesting to see that much of the luminous and
blue sources are variables.
We discovered an infrared variable
star that is not detected in the
OGLE-I band survey. The above
Many
interesting
variable
sources
are
detected.
figures show the light curve of the
pulsar
unknown
unknown
B0
infrared variable. It is detected in H
and K-band, and is not detected in J
band, probably because of the heavy
circumstellar extinction.
We detected all kind of variable sources within the survey areas. The figures shown above are some
There are only 1 star like this in our
examples. The upper panel shows the light curve in J(bule), H(green), and K(red) bands, and the lower
1x1 square degree survey area in
panel shows the time variation of near-infrared color in J-H(blue), H-K(green), and J-K(red) colors.
the SMC. The number of infrared
variables will be compared with that
We will submit the catalog paper as soon as possible, and the whole data will be published after the
in the LMC.
acceptance of the paper.