Seismologic study of the solarlike star Procyon A

1

Università degli studi di Catania Facoltà di
scienze matematiche fisiche e naturali
Silvio Leccia

• Seismologic study of the solar-like star Procyon A

Napoli, 22/03/2005
2
Collaborations and Pubblications

• Padova group R. Gratton, R. U. Claudi, S.
  Desidera

• Aarhus group H. Kjeldsen, J. Christensen
  Dalsgaard

• Catania group L. Paternò, A. Bonanno, R. Ventura
  

Leccia S., Kjeldsen H., Bonanno A. et al.
2005 in preparazione Claudi R.U., Bonanno A.,
Leccia S. et al. 2005 AA 429L17 Claudi R.U.,
Bonanno A., Leccia S. et al. 2004 soho14125C
3
The Talk
Conclusions
4
Stellar oscillations

• The stellar oscillations are small perturbations
  of the stellar equilibrium structure

5
Stellar oscillations

• Determined by the restoring forces pressure
  forces caused by compression and rarefaction and
  bouyancy forces caused by horizontal density
  differences being affected by gravity.
• p modes When the pressure dominates the
  restoring forces (Acoustic waves ? the Sun)
• g modes when restoring is dominated by
  buoyancy (standing gravity wave)

6
Stellar oscillations

• A mode of stellar oscillation is characterized by
  three wave numbers n, l, m.

l2, m2
l1, m1
l2, m1
7
The asymptotic approximation for the frequencies
of p-modes
Large separation
8
The asymptotic approximation for the frequencies
of p-modes

Small separation
9
HR asteroseismic diagram

10
An example the Sun
107 modes 0 lt l lt100
Peak frq. 3.05 mHz ??0 135.0 ?Hz ??02 9.6
?Hz
Kjeldsen Bedding, 2004
11
The starsObservational problems

• Requested resolution ? several nights of
  observations
• Measurements of very small amplitude (some parts
  per million, less than 1 m/s)

12
Solar-like stars
modes few tens l ? 3
13
Procyon A
14
Procyon A the state of the art 2004
15
Observations
Observations 6 nights (01/ 2,3,4,6,7,9/2001) In
strument TNG SARG Grism yellow (462 792
nm) Resolution 144000 ltS/Ngt 300 ltSeeinggt 0.88
arcsec
P.I. and observer Dr. A. Bonanno
16
Procyon data set

• lttexpgt 10s
• Readout time ? 120 s
• ? 950 spectra collected

17
The pipeline

• The pipeline is composed by IRAF scripts launched
  by a C-shell
• Working and temp directories and lists for the
  IRAF tasks are generated automatically
• Trimming, overscan, bias subtraction and
  flatfielding of ? 900 spectra in ? 10 hours
  machine time (PC)

The pipeline realized is a fundamental instrument
for future asteroseismic campaign with SARG
18
The data set

• Each spectrum can be
  divided in two parts
• In the blue part (from 462 to 620 nm) there are
  the absorption lines of the Iodine cell
• Radial Velocities measurements
• The red part (from 622 to 792 nm) free from the
  absorption lines
• EW measurements for lines that are T -
  sensitive

19
The Iodine cell technique
SUPERPOSITION OF REFERENCE AND STELLAR SPECTRUM
POSSIBILITY OF REMOVING INSTRUMENTAL SHIFTS
IODINE SPECTRUM forest of very narrow lines
blends even at very high resolutions A FULL
MODELLING OF COMPOSITE SPECTRUM IS
REQUIRED Spurious shifts between the narrow
iodine lines and the stellar lines when the IP
changes
Simultaneous model of instrument profile (using
iodine lines themselves) is mandatory to reach a
RV precision below 5 m/s (AUSTRAL Endl et al.
2000)
20
Radial velocity results
int. err. 1.38 m/s r.m.s. 4.48 m/s
21
Power spectrum
22
The analysis of radial velocity

• We high-filterd the PS
• We were left with a time series of residual
  velocities ri that reflects the noise propreties
  of the measurements
• We examinated the ratio ri/?i wich we expect to
  be gaussian- distribuited, so the outliers
  correspond to suspect points

Log N (gtr/?)
f (gt r/?)
23
Power spectrum
24
No weights
Modified weights
25
Frequencies analysis

• Noise level measurements
• A first measurement of the large separation
• Extraction of the frequencies
• Echelle diagram ? mode identification
• Measurements of large and small separation (if
  it is possible)

26
Search for a comb-like pattern

• CR(??)PS(?max-1/2 ??)PS(?max1/2 ??)PS (?max-??)
• PS (?max??)PS (?max-3/2 ??) PS(?max3/2 ??)
• PS (?max-2 ??)PS(?max2 ??)0.5 (Kjeldsen et al.
  1995)

?? 56 1 ?Hz
lt??gt 55.7 1.4 ?Hz
27
Identification of Oscillation Frequencies

• We identify the strongest peak in the PS in the
  frequency range 0.5-1.5 mHz and we subtract the
  corresponding sine wave from the time series.
• We recompute the PS and define a new search area
  centered at distance 56 ?Hz.

28
Echelle diagram
29
Results 1
30
Equivalent width measurements
Red part of the spectrum 21 orders, but
... only 4 orders can be used for EW measurements
31
Measurements technique
32
Granulation
EW/dT7.8 (Bedding et al. 1996) MOST
dI/dT4.5 (Kjeldsen, private communication) To
convert EW to I we multiply the PSD by
(4.5/7.8)2
33
Result 2
MOST
SARG
34
Conclusions

• Realized automatic pipeline for data reduction
  from SARG
• Identified 25 p-modes from radial velocity
  measurements
• Found ??55.69 ? 0.14 ?Hz and
• ??5.4 ?0.5 ?Hz
• Obtained upper limit for the granulation and
  compared with MOST data