2 Target selection and observations

Since the main aim of our program was a statistical exploration of the roAp phenomenon, we attempted to bias our target selection towards the detection of new roAp stars. Most of these objects are cool SrCrEu stars. Regrettably, systematic spectral classifications for Northern Hemisphere objects are still missing in the literature. Therefore we chose a different approach.

Nelson & Kreidl (1993) explored the positions of pulsating and nonpulsating Ap stars in several color-color diagrams for the Strömgren and Geneva photometric systems. While they did not find a clear separation of the two groups, the Strömgren [m₁], [c₁] and Geneva $\Delta, [g]$ diagrams appear useful to select promising roAp candidates. Consequently, we searched existing Strömgren (Hauck & Mermilliod 1998) and Geneva catalogues (Rufener 1989) for potential target objects.

Since this procedure does not allow one to distinguish between Am and Ap stars, we attempted to find spectral classifications for the objects of interest in the literature. Consequently, priorities were assigned to the different stars and observations were performed according to them.

Three telescopes at two observatories were employed for our survey: the 0.9 m and 2.1 m telescopes at McDonald Observatory, Ft. Davis, Texas together with a standard two-channel photometer (Grauer & Bond 1981) and one of the twin 0.75 m Austrian Automatic Photoelectric Telescopes (APT, Strassmeier et al. 1997) at Fairborn Observatory, Washington Camp, Arizona with a single-channel photometer.

  

Table 1: Journal of the observations

The observations (see Table 1 for an overview) were carried out as continuous 10-second integrations through a Johnson B filter. Large apertures ($\gt 35\tt ''$) were employed to minimize the contributions of seeing and guiding.

The reductions were performed using the standard technique for roAp star observations (e.g. Martinez 1993), including some low-frequency filtering in the presence of sky transparency variations by low-order polynomial fits or involving the Channel 2 comparison star, if found appropriate. The final reduced light curves were searched for rapid oscillations by means of a standard Fourier technique (Breger 1990). Amplitude spectra of our runs are displayed in Figs. 1-3 (see below for more details).

 

Figure 1: Upper panel: a portion of our light curve of HD 99563. Lower panel: The amplitude spectrum of the whole run

 

Figure 2: Upper panel: the discovery light curve of HD 122970. Lower panel: The amplitude spectrum of this run

 

Figure 3: Amplitude spectra of our null results. Note the larger ordinate scales for the first run on BD+40 175A and for the run on HD 83965B

 

Figure 3: continued

2.1 Comments on individual stars

BD+39 654A: This star was our most promising roAp candidate. It has been classified at SrCrEu by Bidelman (1983). Schneider's (1986) Strömgren photometry yields (b-y)₀ = 0.136, $\delta m_1 = -0.095$ and $\delta c_1 = -0.267$, while Geneva photometry (Burki et al. 1998) results in $[\Delta]$ = 0.163 and [g] = 0.162. All these features are typical for roAp stars. However, BD+39 654A did not show any rapid variability, but ironically, its Channel 2 comparison star, HD 17892, turned out to be a new $\delta$ Scuti variable (Handler 1998). In any case, we consider it worthwhile to continue testing BD+39 654A for rapid oscillations.

BD+40 175A: This is the only object for which we have no color photometry available. However, its spectral classification (SrCrEu, Bidelman 1985) and its effective temperature estimate ($7100~\pm~400$ K, Babel & North 1997) make it a good roAp candidate.

We took a short run under non-ideal photometric conditions, where this target seemed to show rapid oscillations with an amplitude of 6 mmag and a period of 9.1 minutes. Two further runs, one with the McDonald 2.1 m telescope considerably decreasing scintillation noise, did not confirm this possible variability. We suggest further observations of BD+40 175A as well.

HD 99563: Dorokhova & Dorokhov (1998) announced the variability of this object with a period of about 11.2 minutes. We reobserved this star with the APT. The amplitude spectrum of our data shows a dominant peak with an amplitude of 2.0 mmag with a period of 10.7 minutes (Fig. 1). Therefore we consider the roAp nature of this star to be confirmed.

HD 122970: This is a new roAp star. We discovered rapid oscillations with an amplitude of 2.0 mmag and a period of 11.1 minutes in the first night we observed it (Fig. 2). This behaviour was confirmed during two subsequent runs. This star was followed up with multisite photometric and single-site spectroscopic observations; the results of this effort will be reported in a subsequent paper.