Urgent! NSv 6690 imminent SMBHB merger confirmation needed

[Simon de Visscher]

Hello,

I took 10x900s last night on NSV6690. Done with the Atik414ex on the ALPY600 and Newton 400f5.

I (approximatively) annotated the spectrum without the redshift correction. The instrumental response is based on a A8V star.

The red line is a smoothed version of the blue curve. Things seems to be quite aligned with Robin's data.
The .dat file is put in attachment.

Best,

Simon

[Robin Leadbeater]

Nice spectrum Simon ! Your 400mm aperture "light bucket" is working well. My H alpha line looks wider even allowing for the difference in resolution but I think that might be due to the Telluric bands either side of the red shifted H alpha which I removed in my spectra. Are the Tellurics left in your spectrum ?

I have not had any more news from the discovery team but there is an Atel about follow up X ray monitoring.
https://www.astronomerstelegram.org/?read=15225
There is some photometry in the AAVSO database but the light does not show any obvious trend so far.

Cheers
Robin

[Simon de Visscher]

Are the Tellurics left in your spectrum

That's a good question, and I think the following image partially answers it:

NSV_Nobgremoved.jpg (154.59 KiB) Viewed 11932 times

This is the spectrum without the background removal: the telluric lines do appear very clearly (do not mind the colors, that is just to make the image looking good ) thanks to the skyglow which was apparently quite strong that night.

And this is with the background removed:

NSV_bgremoved.jpg (101.87 KiB) Viewed 11932 times

So I *suspect* that the telluric lines are efficiently removed because of this*.

I have not had any more news from the discovery team but there is an Atel about follow up X ray monitoring.
https://www.astronomerstelegram.org/?read=15225

Really nice! I just hope the merging won't happen within 2 weeks because of the moon that is not going to help monitoring the event
Btw is there any db where the spectra can be stored?

Cheers,

Simon

*This might mean: capture the spectra before the target gets too high in the sky

[Robin Leadbeater]

Hi Simon,

I mean the telluric H2O/O2 absorption bands, not the sky glow lines. There are some strong ones either side of the redshifted H alpha and they will not be removed by the sky subtraction.
http://www.spectro-aras.com/forum/viewt ... 925#p16612

Here are the telluric bands (in red) superimposed on your spectrum. You can see the holes produced by the absorption bands

(from Christian Buil's Vega atlas http://www.astrosurf.com/buil/us/vatlas/vatlas.htm )

tellurics_NSV6690.png (18.74 KiB) Viewed 11917 times

The recommendation for the BeSS and most other databases is to leave them in and let the end user remove them but they are not always obvious and in this case will affect the result.

The traditional way to remove them is to make a template from your hot reference star and divide the spectrum by it. In my case I just included them in the response curve. Here is the response I used with the ALPY200 for example. You can see the Telluric bands left in.

response_inc_tellurics.png (14.38 KiB) Viewed 11917 times

The ARAS database only accepts spectra of particular objects but the BAA database accepts spectra of any object
https://britastro.org/specdb/
You will see my NSV6690 spectra there
They should be 1D fits files conforming to the BeSS standard and new users are moderated to check for quality

Cheers
Robin

[Simon de Visscher]

Ah no these ones I didn't do anything particular.

Here are the telluric bands (in red) superimposed on your spectrum

Ok indeed the effect is unfortunately very strong since it affects both the blue and red tails of the Ha region. I'll try to use the reference stars spectrum to remove them, but that's definitely a delicate data correction.

In my case I just included them in the response curve. Here is the response I used with the ALPY200 for example. You can see the Telluric bands left in.

Since the instrumental response is based on a blurred ratio of spectra, isn't there a risk to extend the correction beyond the "telluric "areas?
And aren't the high frequency variations of the telluric bands still there after the correction?

The ARAS database only accepts spectra of particular objects but the BAA database accepts spectra of any object
https://britastro.org/specdb/
You will see my NSV6690 spectra there
They should be 1D fits files conforming to the BeSS standard and new users are moderated to check for quality

Ok nice I'll have a look!

[Robin Leadbeater]

Since the instrumental response is based on a blurred ratio of spectra, isn't there a risk to extend the correction beyond the "telluric "areas?
And aren't the high frequency variations of the telluric bands still there after the correction?

Yes including them in the response is only an approximation, you have to use minimum smoothing (It is better than nothing though) It works reasonably well for my very low resolution ALPY200 where the lines blend into bands but is less accurate at higher resolution.

The "professional" way to do it is to produce a telluric line template from a hot reference star taken on the night, adjusting the intensity for any difference in air mass. This is then used to remove them from the target spectrum in a separate operation.

You can also use a pre-prepared template like the one I showed for Vega and "tune" it (scale and filter) to match the tellurics in the target spectrum and then divide the spectrum by it. This only really works well though if the tellurics are obvious, for example In high resolution high SNR spectra. It works well for H alpha profiles of Be stars for example and ISIS has a special function for this. Here it is used on Castor

ISIS_telluric_removal.png (52.01 KiB) Viewed 11887 times

[Robin Leadbeater]

Yes including them in the response is only an approximation, you have to use minimum smoothing (It is better than nothing though) It works reasonably well for my very low resolution ALPY200 where the lines blend into bands but is less accurate at higher resolution.

You can test how well it works by applying the response to the reference star. The tellurics should disappear.

telluric_removal_ALPY200.png (16.34 KiB) Viewed 11886 times

[Simon de Visscher]

The "professional" way to do it is to produce a telluric line template from a hot reference star taken on the night, adjusting the intensity for any difference in air mass. This is then used to remove them from the target spectrum in a separate operation.

Yes that was the "simplest" thing I had in mind. Obviously the airmass is the tricky part , but that is a second order effect, having the small over- or under-correction will be less a problem than no correction at all. The reference hot star was a A8V star at ~5 degrees from NSV6690, that should be fine.

I'll post discuss the result here when done.

The ARAS database only accepts spectra of particular objects but the BAA database accepts spectra of any object
https://britastro.org/specdb/

Thanks for pointing to that db, very useful, I'll try to send my spectrum/a once ready.

Best,

Simon

[Simon de Visscher]

Just played a bit with the telluric lines removal using a simple polyfit to extract the low frequency spectrum and then the HF one which I use as a divider to the uncorrected NSV spectrum. The correction is done without any loss of resolution. The lower right plots are blurred *after* calculating the correction, just to help the visualization. The main conclusion is that the SNR is so low at full resolution that the effect is sub-dominant.

Anyway, that's good to try. It may become relevant if I manage to make longer sessions on that object.