Doppler shift attempt!

[Gasman]

Hi guys

Crude attempt at detecting doppler shift using the Halpha lines of Vega and beta Lyra

I suppose its fairly selective on where you actually pick the high/low points but I`ve tried to select the highest and lowest points of the Halpha peak and trough on my plot.

So it appears allowing for the accuracy of these images that using the Ha emission line of beta Lyra at 6562.3 and the Ha absorption line of Vega of 6565.3 gives a shift of about 3 angstroms. I`m not quite sure how to relate that to speed??. I noticed one of John`s post`s over on the Inside astro forum and he arrived at a figure of 0.29 nm (eg 2.9 angst)  so maybe I`m not too far out of the ball park as John`s spectroscope has a better resolution than my Alpy .

cheers

Steve

[jnp]

Hi Steve,

It' a good start into a complex area.
 
I wasn't sure if the Alpy had sufficient resolution, but it shows the Helium 6678 region nicely. This changes quite dramatically over the 13 day period.

You may need to recheck the calibration. The dips before 6900A are telluric Oxygen at 6869A and should line up (as unaffected by doppler shifts). Were you using Neon lamps?

The doppler equation calculates speed from wavelength. The velocity you measure depends on positions relaive to the Earth solar orbit and the Earth's rotation. BASS will show velocity on axis and report values if you setup the xaxis tab (in Chart Settings) to Show doppler shift and enter 6562.85 as the Rest wavelength. There is no heliocentric calcuation in BASS yet.

You can increase the number of decimal places in the Measurement Options tab of Measurements screen to help with rounding issues.

Can you confirm how you are taking the 'measurements'. Using the Measurements screen, you select a range before and after the Ha region (like during a calibration). The app will then report values it detects. For Vega read against 'Min' result values, or betaLy read against 'Max' values.  These will show doppler shift if enabled. No need to pick the points yourself.

cheers
John

[Gasman]

Ah! right cheers for that John. I didn`t give a thought to the telluric Oxygen regions lining up . Must admit I do flit between ISIS and BASS a lot and maybe the calibration differs slightly??.  Yes I used an Argon/Neon lamp for the calibration with ISIS and transferred both Vega and beta Lyra calibrated fits to BASS, methinks maybe best stick with the same software from start to finish. Its mainly a calibrating convenience thing why I use ISIS with the Arg/Neon lamp. Didn`t know BASS had the Doppler shift option!

Thanks again

Steve

[Gasman]

HI John

Just done a calibrate on beta Lyra using the Balmer values Halpha,beta and gamma in BASS and noticed that the telluric areas of the betalyra profile and the reference B8i profile (no B8ii-B8iii in the library) don`t line up also??

Have I missed something please?

regards

Steve

[jnp]

Hi Steve,

I think the Pickles reference spectra dont include the tellurics.

Interesting Helium activity at 5875A BTW

The calibration in BASS and ISIS differ in method but, should be the same once converted to ID profile.

regards

John

[Gasman]

Hi John

Must confess I haven`t used any of the new stuff in BASS before (doppler etc ) didn`t even know it was there so I`ve just printed off the whole ruddy handbook to bring myself up to speed !!

Theres`s dedication for ya!

cheers

Steve

[Merlin66]

+1 from me for BASS Project!

John has recently added the capability of preparing Spectroheliograph mosaics - Majic!!

[robin_astro]

Hi Steve,

As a rule of thumb, with care  it is relatively straightforward  to measure doppler shifts down to the equivalent of around 1/10 the of the resolution of the spectrograph

(eg ~1A/50km/s for the ALPY. and ~0.03A/1.5km/s for the LHIRES) A stable spectrograph helps (which the ALPY is) and you need a good consistent technique for measuring the line centres (You can measure relative changes in the same star much more precisely than making an absolute velocity measurement)  Dont forget to correct for the velocity of the earth too which can be ~+-30km/s.

There is an example here using the ALPY and the cross correlation technique in ISIS

http://www.spectro-aras.com/forum/viewtopic.php?f=8&t=618

which can potentially be used to measure much smaller shifts providing the system is stable enough (<1/100 of the resolution  in the case of exoplanet shifts for example)

Cheers

Robin

[robin_astro]

Steve, 

Dont forget that Beta Lyrae is a very complex system (an eclipsing binary with mass transfer and an accretion disc, which is where the H alpha emission comes from.) so the radial velocity will not be constant.  If we look in Simbad we find that the difference in the mean radial velocity between Vega and beta Lyrae is only ~22 km/s (~0.5A) but with beta Lyrae there will be a varying orbital velocity component and the shape of the H alpha line also changes dramatically during the orbit as the secondary component eclipses the disc, affecting the measurement of the line centre.

Robin

[Owmuchonomy]

Very interesting topic even with my superficial understanding. Thanks for sharing.

[Gasman]

Hi Chaps

It was a somewhat crude attempt to try and detect some shift in the spectrum of the eclipsing binary star beta Lyra and maybe give some indication of whats happening as one component eclipses the other. The more I read about beta lyra the more interesting it becomes, apparently some of the Helium lines also go from emission to absorption over an 8day period see here- http://www.lafterhall.com/beta-lyrae_4200-5200_alpy-c9_201407-02-06-09ut_002.jpg a very complex system indeed!. All interesting stuff

Steve

[Gasman]

Right just to try and get the practical bits of ISIS in my head and note that the Simbad RV value for beta Lyra of 2.20 km/s

Calculating the Heliocentric velocity (Vh) for Vega and beta Lyra on the night of 020715 I get 1.393 and 2.669 respectively , I used a CCF 1/2 interval of 2600 to show a reasonable curve (not sure about that??) and step of 1km/s. Putting these values into ISIS thus-

This shows an RV difference (observed peak ) of 246 km/s, does this indicate the difference in radial velocity between Vega and beta Lyra is 246 km/s?.

So I now ask

1- How do I determine the correct value to use for CCF 1/2 interval

2- As ISIS seems to calculate the difference in RV between an object and reference star then Simbad gives RV for beta Lyra 2.20 km/s and RV for Vega -20.60 km/s implying a RV diff of 20.60+2.20 = 22.80 km/s so my value of 246  km/s is an order of magnitude out but am I on the right path

Need a lie down now !!

cheers

Steve

[robin_astro]

Hi Steve,

This ISIS screen is for calculating the radial velocity difference between two spectra using cross correlation. To work it needs the two spectra to be very similar with the same features in each - typically the same star at a different time.  It works by measuring the correlation between the two spectra, then moving one of them by a small step (in km/s)and recalculating the correlation. It keeps doing this until the correlation reaches a maximum ie when the two spectra line up best and reports the shift (in km/s) needed to bring them into line.  It will not work properly comparing Vega with beta Lyrae though as they have totally different spectra.   

You need to measure the centre wavelength of some lines in both spectra (eg in ISIS using FWHM tool in the profile display)  and compare them with the rest wavelength for that line. You can then calculate the radial velocity 

rv = c *(measured wavelength-rest wavelength)/rest wavelength

Cheers

Robin

[robin_astro]

There are some other useful related tools in ISIS under the Misc tab

Compute the heliocentric velocity correction  (your observatory coordinates need to be set up correctly)

Apply an velocity shift to a spectrum (eg to apply the heliocentric correction)

Generate a spectrum calibrated in velocity relative to a rest wavelength from one calibrated in wavelength

Also ISIS will calculate a heliocentric corrected spectrum during data reduction if the box in the General tab is ticked  

Cheers

Robin

[Gasman]

Thanks Robin I was perhaps assuming too much from ISIS . I was under the impression that you simply input an object and a reference spectrum and ISIS would come up with an RV difference. Shame there isn`t an ISIS user manual to follow. I`ve printed out a few ISIS tuts. using Google translate as they are all in French but not ideal and some of the tools not very intuitive (at least with my limited intuition )

Steve

[robin_astro]

Hi Steve,

Yes it can be frustrating tracking down descriptions of some of the features in ISIS. A useful link can be the change log for ISIS

http://www.astrosurf.com/buil/isis/new/release.html

You can find details of the cross correlation radial velocity measurement under

ISIS version V5.1.0 (01/12/2012)

Cheers

Robin