High time-resolution precision radial velocities with CRIRES: Probing uncharted territory in the spectra of peculiar A stars
Coordinator: D. Kurtz (University of Central Lancashire, UK)
Co-Is: C. Cowley, V. Elkin, L. Freyhammer, S. Hubrig, G. Mathys, A. Seifahrt, J.-F. Wyart
We propose observations of a bright roAp star in high time resolution to obtain precise radial velocity measurements on spectral lines of rare earth elements. Scientifically we will address the question of the atmospheric structure of this peculiar roAp star. On the technical side we will test the short term spectral stability of CRIRES against RV stable lines in the target and against atmospheric absorption features.
Allocated Time: 0.5 hours
|F_nu or F_line
|21 10 20.5
|+10 07 53.7
|23.4 Jy, H=4.1
Project description/scientific objective:
The rapidly oscillating Ap (roAp) stars are H-core-burning SrCrEu peculiar A stars with effective temperatures in the range of about 6600 K to 8000 K that are strongly magnetic and pulsate in high overtone acoustic modes with periods between 5.65-21 min. They are amongst the most peculiar stars known and even include among their numbers the extreme case of HD 101065, arguably the most peculiar star. As such, they are of great interest for the study of the inter-action of stellar magnetic fields with rotation, pulsation and atomic diffusion, all of which have implications for many fields of stellar astrophysics.
We (DK, GM, VE, LF) have been carrying out a pulsation survey of this class of stars with VLT UVES, obtaining 2 hr of observations at 1 min time resolution or better for all members of the class. A total of 44 hr service mode and 4 nights of visitor mode time have been devoted to this project so far with 3 more nights awarded in July 2006 and three more in March 2007. We (GM, SH) have also been studying the magnetic field structure of Ap stars for many years, including recent studies with VLT FORS.
We propose now a test run on the bright roAp star gamma Equ (V = 4.7) for 2 hr with 1 min time resolution with CRIRES. We expect from initial data from CRIRES to obtain individual spectra with S/N ~ 200, or better. The immediate goal is to test the radial velocity precision of CRIRES. We will observe gamma Equ for 2 hr at high time resolution with UVES in late July, so there will be a direct comparison of wavelength stability and radial velocity precision obtained between the two instruments. While gamma Equ is known to be multiperiodic, the multiple frequencies will not be resolved in two hours. Thus the radial velocity amplitudes of the pulsations may be different between the UVES and CRIRES data sets, but this does not affect the errors on those amplitudes and thus does not affect the test comparison of the precision of the two instruments.
The scientific goals of this study are manyfold: The IR spectra of the Ap stars are largely unknown, especially the important Rare Earth element lines; these first observations will chart new territory. We (J.-F.W., CC) have calculated wavelengths in the IR for some Rare Earth element ions. These are formed high in the atmospheres of roAp stars, hence their pulsation probes atmospheric levels (tau_5000 ~ 10e-5) that can be resolved in no other star but the sun. The best lines that we have found for our purposes are one of Nd II, an ion at 1626.81 nm, and a Ce III line at 1612.8750 nm. There are also other lines of Ce in this wavelength range. The ions of Nd and Ce are known to have high amplitude in roAp stars from our VLT survey. Taking into account the radial velocity of the star and barycentric correction, the above two lines do not overlap any strong telluric feature.
We (CC) have also determined from the VALD (Vienna atomic line data base) that within the range of the chosen setting are many Fe I lines. Fe I is known to have pulsation amplitude in gamma Equ below 20 m/s in HD 166473 we (GM, DK, VE) have shown the Fe lines have no pulsation amplitude above 2 m/s (they are formedat the level of a radial node in the observable atmosphere - a unique property of the roAp stars). Thus the Fe lines are a control on the precision of the instrument. We will also get the core of Br9 for H; H alpha shows an important core-wing anomaly in the visible and sample the pulsation at continuum optical depths in the range -5 tau_5000 -2. Br9 will probe somewhat deeper than this and give a good sample of pulsation at a range of optical depths.
The Nd II line we have chosen is magnetically sensitive with a Lande factor of g=1.015 and a nearly normal Zeeman triplet. Zeeman splitting increases as lambda2, so the Zeeman patterns are better separated in the IR, hence surface field strength measured to higher accuracy. gamma Equ has a known surface field strength of H_s = 3.9 kG. We (GM, DK, VE) have predicted that an improved pulsation diagnostic is possible by comparing the pulsation amplitudes between the pi and sigma Zeeman components, since they sample the nonradial pulsation over the visible hemisphere differently. So far in the visible we have found that the separation of the Zeeman components is not large enough to allow a test of this theory. The IR promises to provide that test and a new constraint on the pulsation geometry, hence mode identification, a fundamental step in asteroseismology.
Ultimately, we expect IR spectra of the most peculiar of the Ap stars may allow a cleaner study of their abundances. In the visible, for example, HD 101065 hassuch a bizarrely rich spectrum of Rare Earth element lines that there may be no true continuum. Arguments that have raged over the last few decades are settling down, but may be much better resolved in the expected less-crowded IR spectra of these stars.
The radial velocity amplitudes for Rare Earth element lines for gamma Equ are of the order of 400-800 m/s from studies already done.If CRIRES can reach precision of tens of m/s for individual lines, we will have good S/N for the pulsation studies. We have the expertise to carry those out (DK, VE, LF); we have expertise in Ap and roAp stars (DK, GM, SH, VE, CC); we have the expertise for the magnetic studies (GM, SH); we have the expertise in laboratory understanding of the Rare Earth elements (J.-F.W.) and the astronomical study of the line spectra (CC); we have the expertise in abundance analysis (DK, VE, LF, GM, SH, CC); and, of course, we have the expertise in obtaining the spectra (AS and the CRIRES team).
For data reduction we will use standard packages for long slit data in MIDAS and IRAF as well as a self-written IDL pipeline that has been successfully tested with data obtained during CRIRES commissioning run in June (AS).
Readout and nodding overheads are 30sec for a single exposure with 60s DIT (see data taken in commissioning). Hence for 2 hours (-15 min for preset and acquisition) at least 35 AB cycles should be taken with DIT = 60s and NDIT=1. (70 exposures of 60s x 1 NDIT = 90s x 70 = 105 min + 15 min for preset + AO acquisition = 120 min).
Use 0.2" slit!
Adaptive optics notes: Target = Guide star, R=4.4 mag, B-V=+0.3 mag
NOTE: If the throughput is high enough that the spectra runs into saturation the DIT should be adapted accordingly and the number of AB nodding cycles should be increased to guarantee a temporal coverage of ~2 hours.
If the throughput is much less than expected open the slit max. to 0.3" and enhance the exposure time to max 120s.
Keep nod throw reasonably small to avoid distortion problems (suggestion: 5")
Take at the beginning and end of the OB a well exposed ThAr reference spectra.