Secular evolution in Mira variables

1
Secular evolution in Mira variables
Matthew Templeton Janet Mattei (AAVSO) and Lee
Anne Willson (Iowa State University)
93rd Spring Meeting of the American Association
of Variable Star Observers Berkeley, CA July 22,
2004
2
Secular evolution
The Miras and semiregular stars are in a very
rapid phase of stellar evolution the AGB. They
may undergo large changes on human (secular)
timescales.
The AAVSO International Database contains over a
centurys worth of data for some of these stars.
Can we use this data to study the real time
evolution of variable stars?
3
AAVSO Mira stars

• Over 1500 Mira stars in our validation file
• Over 500 stars for which substantial data are
  available
• Light curves several decades to over a century in
  length

Can we study evolutionary period change in the
Mira stars using AAVSO data?
4
Period variations in Miras

• cycle-to-cycle period fluctuations
• light curve shapes vary
• secondary periods

Some Mira stars have long-term period
changes. Why? Could they be caused by thermal
pulses?
5
Thermal pulses
(Lattanzio, J.C. Frost, C.A., 1998, IAUS 189,
p373)
6
Thermal pulses

• Built-up helium shell around core undergoes rapid
  burning
• Hydrogen-burning shell is extinguished
• Change in R,L,M
• Change in Period!

7
Period changes
Vassiliadis Wood 1993 (ApJ 413, 641)
8
Measuring period changes
We used weighted wavelet analysis (Foster 1996)
to measure how the period changes with time.

• Works best with several decades or more of data
• Period accuracy of 1-2 per window
• dln(P)/dt accurate to about 210-4 y-1

9
R Aquilae
10
T Ursae Minoris
11
Predicted and observed period changes

• monotonic period changes in timescales of
  centuries to 0.1 Myr
• short-term dlnP/dt variations of 10-2-3
• long-term dlnP/dt variations of 10-5 or less
• most stars have dlnP/dt smaller than our
  detection limit

12
Period changes Results

• About 10 (57 of 547) of the well-observed AAVSO
  Miras have statistically significant dlnP/dt
• Majority of stars have dlnP/dt too small to
  measure (limited span of data, random
  fluctuations, measurement accuracy of dlnP/dt)

13
RU Tauri
14
Long-term variations?
Stars like RU Tau and S Ori appear to have
periods that vary, but not monotonically.

• What causes them? Perhaps they are thermal (KH)
  oscillations?
• Are some stars undergoing monotonic changes
  just undergoing long-term variations on very long
  timescales? (LX Cyg, BH Cru)

15
Long-term variations

• Thermal pulse initiates large-scale changes in
  stellar structure
• Kick caused by shell ignition / quenching
  causes thermal oscillations on KH timescale

16
The future...
Longer spans of data will let us reliably measure
dln(P)/dt for more of these objects. Please keep
observing!