Recovered spectra fill a decade-long gap in Ap star monitoring
Ap stars are main-sequence early F to mid-B stars with large surface abundance anomalies and organised magnetic fields ranging from a few 10 G to a few 10 kG. In most of them the Zeeman splitting of spectral lines is smeared out by rotational Doppler broadening, so the field is diagnosed through spectropolarimetry. A few percent, however, have both a strong enough field and a low enough v sin i that their lines appear resolved into magnetically split components (RMSL). That makes the mean magnetic field modulus <B> — the line-intensity-weighted average over the visible disk of the modulus of the magnetic vector — measurable in a mostly model-free way.
The authors recovered 103 high-resolution spectra of Ap stars taken between 1999 April and 2004 July with the CES at ESO and the Gecko spectrograph at CFHT. Of these, 82 show RMSL and belong to 39 different stars. Most were obtained before observatories began systematic real-time archiving, and had been lost track of. Because the earlier large datasets span 1988-1998 (43 stars) and mostly 2001-2021 (36 stars), these recovered measurements bridge the gap between them and add constraints on how the studied stars vary.
One iron line yields both field strength and radial velocity
The measurement rests on a single spectral feature: the Fe II λ6149.2 Å line. Its magnetic splitting gives <B>, and its wavelength shift gives the radial velocity (RV). Both quantities were extracted from high-resolution spectra recorded in natural light, not in polarised light.
This matters for how the results should be read. The <B> values are averages over the visible stellar disk, not local field strengths, and they depend on the geometry of the observation only weakly — which is exactly why RMSL stars are valuable. The RV measurements are a by-product of the same analysis, but they are what make the binarity results possible. The paper combines its new measurements with published ones to characterise the variability of both quantities over time.
Two long rotation periods are now unambiguous
The study derived accurate, unambiguous rotation periods for the first time for two stars. HD 29578 has P_rot = 10.82 yr. HD 165474 has P_rot = 26.71 yr, which the authors note is the second longest accurately determined Ap star period, after HD 50169 at 29.04 yr. The <B> variation curves of both stars were also characterised by least-squares fits.
These are direct period determinations, not lower limits. That distinction matters because most of the paper's other rotation results are one-sided bounds, and a bound and a measurement answer different questions about a star.
Six stars get decade-scale lower limits, four show no change
For six stars the authors established reliable lower limits on long rotation periods from definite long-term drifts in <B>: HD 47103 (P_rot ≫ 6.8 yr), HD 116114 (P_rot ≳ 48 yr), HD 119027 (P_rot ≫ 9 yr), HD 150562 (P_rot > 21 yr), HD 177765 (P_rot ≳ 50 yr), and HD 216018 (P_rot > 26 yr). Shorter periods proposed for some of them in earlier work are confidently ruled out.
A different outcome applies to four stars, where no significant <B> variation was seen over the observed spans: HD 75445 (over 13 yr), HD 110066 (over 33 yr), HD 134214 (over 23.5 yr), and HD 137949 (over 19 yr). For HD 137949 the standard deviation of 49 <B> values across 19.3 yr is only 24 G. The authors cannot yet decide whether these stars are seen pole-on or simply rotate far more slowly than their observation span; they note that pole-on geometries should be statistically rare, so some likely have extremely long periods, and that linear spectropolarimetry could discriminate between the two possibilities.
HD 47103 is the first long-period Ap star above 7.5 kG
Earlier work concluded that in stars with rotation periods longer than 150 d, the average over a rotation cycle of the mean magnetic field modulus never exceeds 7.5 kG. HD 47103 is now confirmed as the first known exception: it is the first Ap star with a rotation period of years known to have a magnetic field significantly exceeding 7.5 kG, with <B>_av around 17 kG.
The authors are explicit that this does not overturn the broader statistical pattern. The rate of occurrence of very strong fields (<B>_av > 7.5 kG) remains much higher in stars with relatively short periods (P_rot ≲ 150 d) than in extremely long period stars. The finding is a single confirmed exception to a trend, not a reversal of it.
Nineteen of 34 stars are definite or probable binaries
The RV measurements support a substantial binary population in this sample: 19 of the 34 stars are definite or probable binaries. The paper determined orbital elements for the first time for three of them — HD 116114, HD 142070, and HD 144897. All three have orbital periods of the order of a decade, and their orbital eccentricities, about 0.65, do not significantly differ from each other.
The study also refined the orbital elements of six other binaries, revised the lower limit of the orbital period of HD 165474 from P_orb > 6 yr to P_orb > 33 yr, and set a lower limit for HD 200311 of P_orb ≫ 12.5 yr. Of the 14 spectroscopic binaries with a determined orbital period or lower limit, 12 have P_orb ≳ 2 yr. The two exceptions, HD 55719 and HD 61468, are also the only two binaries in the sample with eccentricities below 0.2. The large eccentricities of the long-period systems are consistent with previously reported behaviour for Ap stars generally.
Four additional stars — HD 110066, HD 119027, HD 157751, and HD 335238 — appear to belong to binaries based on the RV measurements, but the authors state that longer-term follow-up is needed for definitive confirmation.
The new data refine individual stars, not the statistical picture
The authors state directly that their batch of new measurements, while substantial, does not justify a systematic revision of the statistical trends derived in the earlier exhaustive studies. The results are consistent with those trends apart from the single HD 47103 exception. The value of the work lies in improved knowledge of individual stars: rotation period, magnetic field variability and structure, binarity, orbital period, and orbital elements.
The paper also reports the presence of RMSL in HD 12098 and HD 94427 for the first time. That detection is described as somewhat marginal — sufficient to determine <B>, but with rather large uncertainty. For HD 12098 the short rotation period of 5.460 d means Doppler broadening of the split components is not negligible relative to their separation, which limits achievable precision.
The authors frame the broader legacy value of these measurements as a reference for future long-term studies of the most slowly rotating Ap stars, and for understanding the physical processes behind the extreme differentiation of rotation rates among Ap stars, which can span factors of up to 10^5-10^6.
Separate measured periods from limits and non-detections
The most useful way to read this paper is to sort its results by what kind of claim each one is. Two stars have measured, unambiguous rotation periods: HD 29578 at 10.82 yr and HD 165474 at 26.71 yr. Six stars have lower limits derived from definite long-term <B> drifts, ranging from ≫ 6.8 yr to ≳ 50 yr. Four stars have non-detections of <B> variation over 13 to 33 years, which the authors explicitly cannot resolve into either pole-on viewing or extremely slow rotation without further observations.
That sorting determines what the results can support. A measured period can be used directly; a lower limit only rules out shorter values; a non-detection constrains nothing about the period except that it exceeds the observation span or that the geometry hides the variation. The single threshold result — HD 47103 above 7.5 kG at a rotation period of years — is the one place where the paper changes a previously stated boundary, and it does so for one star while leaving the underlying occurrence-rate trend intact.
Sources used in this answer
Ap stars with resolved magnetically split lines: Mean magnetic field modulus measurements 1999–2004
Recovered 1999-2004 spectra yielded the first accurate rotation periods for HD 29578 (10.82 yr) and HD 165474 (26.71 yr), decade-scale lower limits for six stars, four non-detections of <B> variation over 13-33 yr, confirmation of HD 47103 as the first long-period Ap star above 7.5 kG, and first orbital elements for HD 116114, HD 142070, and HD 144897.
