The O-type star HD 93521 (O9.5 III) was featured in a recent Sky & Telescope weekly bulletin because it appears to have been born where O stars aren't supposed to be, namely in the galactic halo. I thought it would be interesting to compare its spectrum with a disk O star, namely Na'ir al Saif in the sword of Orion. I used a LISA spectrograph with R ~ 1100 which covers the whole optical spectrum from 3800 to 7200 angstroms. I've put the halo star's spectrum in the AVSPEC database, but unfortunately the database won't accept the Orion star. But the two spectra are indeed very similar, with 3 obviouse differences: (1) Na'ir al Saif is somewhat bluer (as expected from its slightly earlier spectral type), (2) Na'ir al Saif shows a strong & narrow C III emission line at 5696 angstroms, and (3) HD 93521 has a broad bump in its spectrum between about 5350 and 5500 angstroms. The first two differences fit into the range of variability expected for ordinary O-type giants, but I have no explanation for #3. Does anyone else have an idea?
I could not find your spectrum in AVSpec. There are several high resolution spectra of HD93521 in the ELODIE archive (though many of them are low SNR)
http://atlas.obs-hp.fr/elodie/fE.cgi?n=e500&c=o&of=1,leda,simbad&nra=l,simbad,d&o=HD93521
They do not appear to show a broad hump in the 5350-5500 region though, just the expected He II absorption line at 5412A
(I think the AAVSO database just covers known variables (in brightness) but you could add your spectrum to the BAA database which covers all objects)
Cheers
Robin
Thanks, Robin, for the link to the Elodie archive. My spectrum of HD 93521 is now published in AVSPEC (https://app.aavso.org/avspec/obs/9631), and the "bump" between 5350 and 5500 angstroms is quite evident. I used the nearby A7 V star HIP 49593 to determine the atmospheric/instrumental response, so this spectrum is in units of "normalized flux". I've made a nice chart (PDF file) comparing the spectrum of HD 93521 to the disk O9 III star Na'ir al Saif, but unfortunately I don't know how to post it here.
Thanks, I see your spectrum now. I had a look at this star the other night with the ALPY600 using the same reference star as you (HD87696, though this is actually a Del Sct variable so perhaps not ideal). There is no sign of the hump at 5350-5500 and my continuum deviates significantly from yours at shorter wavelengths being steeper and and more closely matching an early B giant. Perhaps there is a problem with your response correction ? If you apply your response to your measured reference star spectrum does it closely match the Pickles A7v reference ?
Both my HD93521 and HD87696 spectra can be downloaded from the BAA database. Here are the plots
https://britastro.org/specdb/data_graph.php?obs_id=11646%2C11647&multi=yes&legend_pos=ne
Cheers
Robin
Robin - I finally got around to re-observing this halo O-star with my LISA spectrograph. This time I used a MILES A2V star (Chertan in Leo) for the response curve. The result matches your spectrum (and an O9V synthetic Pickles spectrum) much better than before--no bump and a lot more blue continuum. I've put the new spectrum in the AVSpec database, but it's not yet "published". I don't know what went wrong the first time, but perhaps my previous response reference star (HD87696) was not a good choice, as you suggest. As to the original question, there now does not appear to be any significant difference between this halo star and O stars in the disk, at least at R~1000. Perhaps the HIRES folks could see a difference in metallicity?