spectro-aras.com

I'm a relative newcomer to spectroscopy using a LISA on a C11 and I'm gradually learning about recording and processing spectra using ISIS.

One area I'm interesting in exploring is the degree to which it is possible to measure relative radial velocities with a low resolution device like the LISA. I know ISIS is able to calculate relative RV between two spectra using cross correlation. What sort of RV precision/resolution have LISA users been able to achieve in practice? Anyone willing to share their experiences?

Thanks,
David

Hi David,

I did some work on this using the ALPY which were quite encouraging
http://www.spectro-aras.com/forum/viewt ... ?f=8&t=618
but it needs more work.
This was comparing different stars. The results might be even better where the same star is used eg binaries as the template would be identical
Spectrograph stability is important of course. I dont know the LISA stability but given the differences in mechanical design it is hard to imagine it being as good as the ALPY which appears to be rock solid (Though hopefully better than the LHIRES)

Good to see your spectra in JBAA and BAA VSSC BTW

Cheers
Robin

Hi Robin,

Thanks for your response. I don't know if the following is a valid method to test the RV measurement stability with the LISA but I thought I would investigate it.

I've made several measurements of Nova Del over the last 3 months and with each one I have a set of spectra (usually 10 x 60s) of HD 196544 and immediately afterwards a set of 5 x 20s neon lamp spectra. I processed these in the usual way with ISIS and derived a wavelength calibrated spectrum of HD196544 response corrected with the MILES spectrum. As I was learning the process in the early days I found looking back that some of the early spectra were not so well focused so I've eliminated these from the analysis.

For the remaining 12 spectra, I cross-correlated 11 of them individually using ISIS with the final spectrum I've taken so far on 9th Dec. This gives a fitted radial velocity difference in km/s for each measurement of HD196544. Ideally these should all be zero. I used the fitted RV difference given by ISIS rather than the observed value (is this correct?). This plot shows the result. The std dev of the 11 RV difference measurements is 7.1km/s with a mean of 5.2 km/s. If you ignore the two outlying values (not for any good reason!) the std dev is 3.3 with mean 2.4. Why the differences start consistently positive and then tend towards zero I don't yet understand.

On the face of it however this appears to be an encouraging result, but is it a valid assessment of RV stability with the LISA?

Regards,
David

Hi David,

I think it gives a good indication of what I would say is the excellent (sigma ~1/40 of the resolution) repeatability of the measurement (a combination of the equipment performance and the procedure used), which is what counts at the end of the day. Taking lamp spectra at the time of each observation and using them for calibration takes care of any long/medium term spectrograph stability issues. The only additional suggestion I have would be to take lamp spectra immediately before and after each target measurement to check for and quantify any drift during the observation. ISIS also has a rather clever system for tracking and correcting for any drift in multiple exposures using cross correlation which could be useful if you are not already using it. The ultimate precision/accuracy will depend on the spectral content of the actual target(s) measured/compared but these results bode well I would say. Do you have project involving RV measurement in mind ?

Robin

Hi David,

you observe a drift of radial velocity. Did you apply the heliocentric correction to your spectra? It varies by about 10 km/sec during the time of your observations. With the precision of your measurements you should be able to measure the velocity of the earth around the sun.

Regards, Martin

Hi Martin,

Yes, I do apply the heliocentric correction in ISIS before computing the radial velocity. The source of this apparent drift is still unknown to me. It's trending towards a mean of zero which is encouraging but I need to make more experiments like this to understand how good these RV measurements really are.

Thanks for your comments.
David

Hi Robin,

Thanks for these comments.

You say " ISIS also has a rather clever system for tracking and correcting for any drift in multiple exposures using cross correlation which could be useful if you are not already using it." Can you tell me more about this. I don't think I've come across it in what I've read about ISIS so far. It sounds useful.

David

Hi David,

You tick the "Wavelength Registation" box. There is a thread on this subject here
http://www.spectro-aras.com/forum/viewt ... ?f=8&t=458

Cheers
Robin

Hi Robin,

Thanks. I've investigated the wavelength registration option in ISIS and it doesn't work well with the sort of CV spectra I've been working with. It produces a degraded output spectrum, I guess because there is not the well-defined line structure in these spectra which the cross-correlation needs to work effectively.

I'm interested to find out if the LISA is stable enough to measure RVs in some high inclination CVs. More experiment needed!

David

Hi Robin,

I am very interested by the radial velocity measures with Alpy and ISIS and would like to try them.
I saw your topic
http://www.spectro-aras.com/forum/viewt ... ?f=8&t=618
There is one thing i don't understand about your work with HD 157214 and HD 160693.
In ISIS -> Misc -> Radial velocity, how did you find these two -0.6 and +0.9 Vh values ?
I thought (surely wrongly) one had to put the Simbad radial velocity for HD 157214 (so -79 km/s) and +40km/s for the other one.
As i don't understand the meaning of 1/2 CC interval and step, what values do we have to put there ?

Thanks in advance for your help,

Cheers,