LHiRes Stability

[Mike Potter]

I've seen several mentions regarding the sensitivity of the LHiRes III to slight bumps, or even touching. And have experienced, first hand, some often random shifting of calibration lines. So I decided to conduct a few tests of the "stability" of the LHiRes wavelength calibrations. What I found really surprised me - it's actually quite good. In one of the tests I measured a star with a well-established radial velocity (a proposed Vr standard) hd82885 - 11 LMi. My Vr values on two separate nights yielded standard deviations of 1.1 km/sec and 0.4 km/sec. However, the mean value was off by about +3 km/sec in each case. For those interested the article is http://www.beverlyhillsastronomer.org/s ... ctrograph/

Mike Potter
http://www.beverlyhillsastronomer.org

[Robin Leadbeater]

Hi Mike,

My LHIRES is one of those which can give significant changes in the neon line positions, also one of my two grating holders is less stable when moving around the sky than the other for some reason. For the BRITE project RV measurements on the Si II lines I used a fixed NeAr lamp at the telescope aperture and ran it during star exposures to superimpose calbration lines on the spectra and avoid any potential flexure/drift/instability issues. I am just sorting out all the results to send off. My 1 sigma RV error based on repeat measurements of Vega from June to December was 0.5km/s See the attached graph.

I cannot shed much light on your RV offset. The figure for Vega given by SIMBAD of -20.6 km/s appears to be incorrect though. Looking in the literature I found figures like -13.9, -13.5 which is close to my -13.8 mean

Cheers
Robin

[Mike Potter]

Thanks for the comments, Robin:

I finally finished reducing and measuring all of the Deneb Si II spectra. I know - I'm slow! First of all I surely get fewer clear nights than you do! 27 nights stretching from May to November. Mostly after July as I had knee replacement surgery at the end of May. Anyway, the first data I started to reduce were from mid September and for reasons unknown those data show by far the worst scatter in the measured Vr. On each night I acquired anywhere from three to seven spectra - then determined the standard deviations in heliocentric-corrected Vr based on the Si II 6347 & 6371 lines for each night. When I had just done a few nights' data I was pretty disappointed, but it turns out that those were the worst if the bunch. I'm revisiting now to see if the problem is in the reductions. The plot shows the standard deviation for each of the 27 nights for which I had spectra in the 650 angstrom region. Without the six nights between September 15th and October 2nd, indicated, the mean standard deviation is 0.76 km/sec.

The second plot is the result from the whole data set. It's hard for me to tell just looking at your plot but trying to compare nights where we both got measured Vr it would seem the +3 km/sec bias is in these data as well.

Mike

PS - the vertical axis is in Km/sec

[Robin Leadbeater]

Hi Mike,

Yes we had a good summer, though I only took one spectrum (each of perhaps 10-20 sub exposures) a night so 30min clear skies was generally enough. For the RV spectra, with Deneb being so bright, I even observed through quite thick cloud sometimes.

I will email you the data and spectra I sent to Noel so you can compare. There is no guarantee that I am right though of course ! (I find this is a drawback of the "Convento" way of running campaigns compared with the more open "ARAS" way. You dont get the chance to compare results at the time to spot any anomalies early (or even know who else is actually taking part to coordinate your observations)

Cheers
Robin