PNV J20233073+2046041 mag 6.8

Information about outbursts of eruptive stars, Be activity, ...
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Robin Leadbeater »

Francois Teyssier wrote:
Robin Leadbeater wrote:An update of my calculations of H alpha line strength (absolute flux, arbitrary units, click on the attachment to see the trend graphs)
The spectra are low resolution (LISA and ALPY) from the database and scaled to the continuum at H alpha
The H alpha EW is then converted to absolute flux (arbitrary units) using the AAVSO Vmag values (Flux = EW/10^0.4Vmag)
These values are then individually corrected for the ratio of the weighted mean flux in the Vmag passband to that in the continuum at the H alpha line.
(The weighted mean in the Vmag passband is calculated by multiplying the spectra by the normalised Bessel V response and calculating the mean of the result)

Cheers
Robin
Good work Rob in !
Could you compute the absolute flux ?

François
Thanks François, Christian

Using the same technique to calculate the absolute flux was my next step and was very simple. I used Christian's relative and absolute flux calibrated ALPY/LISA spectra as a test. The results are in the attached table

I used the AAVSO Vmag data and the formula that Christian uses in method 1 step 2 here
http://www.astrosurf.com/buil/calibrati ... ration.htm
to calculate the flux at 5556A and compared it with the calculated weighted mean relative flux in the Vmag passband from the relative flux calibrated spectra
The relative flux spectrum is then just multiplied by Flux5556A / (weighted mean relative flux) to give the spectrum in absolute flux.

The two methods (Christian's and mine) give very similar answers within a few percent. (probably dependent on the accuracy of the AAVSO Vmag data)

Cheers
Robin
Attachments
flux_cal_spectrophoto_v_Vmag.png
flux_cal_spectrophoto_v_Vmag.png (5.19 KiB) Viewed 7385 times
Last edited by Robin Leadbeater on Sat Sep 07, 2013 11:22 pm, edited 1 time in total.
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Robin Leadbeater »

Hi Martin,

I was typing at the same time as you :-)

You now see my method and I will look at your suggestion but I suspect we have come to very similar conclusions :-)

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Olivier GARDE
Posts: 1243
Joined: Thu Sep 29, 2011 6:35 am
Location: Rhône Alpes FRANCE
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Olivier GARDE »

A dynamic spectrum from all my data since August the 14th (30 spectras)

Image
LHIRES III #5, LISA, e-Shel, C14, RC400 Astrosib, AP1600
http://o.garde.free.fr/astro/Spectro1/Bienvenue.html
Christian Buil
Posts: 1431
Joined: Mon Sep 26, 2011 6:59 pm
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Christian Buil »

A really excellent and synthetic view of events, Olivier !

Christan
Christian Buil
Posts: 1431
Joined: Mon Sep 26, 2011 6:59 pm
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Christian Buil »

Robin, François et tous,

J'ai complété les méthodes spectrophotométriques avec le calcul des magnitudes BVR dans
un système photométrique. Voici les VRAIES magnitudes de la nova (dans le
système Johnson) et non plus en valeurs relatives :

I completed the spetrophotométrics method with computation of BVR magnitudes
ina photometric system. Here the REAL magnitudes of nova (in the
Johnson system) and not relative values​​:


Image

Cette figure est extraite de la page :

Extracted from:

http://www.astrosurf.com/buil/nova_del2 ... ometry.htm

La méthode employé est décrite ici :

Description of the method here:

http://www.astrosurf.com/buil/calibrati ... ration.htm

On y trouve les équations de rattachement au système photométrique et comment
les établir.

Je crois que c'est très précis. J'ai été surpris. On peut vraiment
faire de la bonne photométrie.

Je viens de mettre en ligne la version 5.3.1 de ISIS qui facilite ces calcluls.

I give photometric system equations and how establish them.

I think it is very accurate. I was surprised.


You can download ISIS 5.3.1 new version, just in-line, for a
full support of the methods.


Christian
Martin Dubs
Posts: 141
Joined: Fri Jan 06, 2012 9:16 pm
Location: Maienfeld, Switzerland

Re: PNV J20233073+2046041 mag 6.8

Post by Martin Dubs »

Hello Christian,

you have done an excellent job of implementing these methods in ISIS and explaining them in detail. There is one small point I would like to add. Most of us do not have a photometric sky nor an ALPY spectrograph. Therefore the absolute calibration is not very reliable, considering also guiding and focusing errors. Therefore it would be useful to calculate absolute flux from the observed relative flux and the observed V magnitude. You have a flux calculation in Tools - spectra processing 3, but this is only useful for standard stars with a smooth continuum and not for Novae and other stars with irregular spectra. It would be very useful if you could include there your method described in
http://www.astrosurf.com/buil/calibrati ... ration.htm
but using the magnitude and the measured relative flux spectrum to calculate the flux conversion constant (by integrating over the filter weighted spectrum, running backward your calculations as described in the tutorial).
This would give a one-click conversion to absolute flux, of course with the precision limited by the V mag which may be different for different observers depending on the filters used and depending on how careful the atmospheric extinction has been corrected. But it would certainly be very helpful for a fast evaluation of flux and comparison of the time evolution of the nova in particular.
With the new version you can also do this, but not quite as comfortably.

Thank you for considering this.

Martin
Christian Buil
Posts: 1431
Joined: Mon Sep 26, 2011 6:59 pm
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Christian Buil »

An English version of BVR conversion page:

http://www.astrosurf.com/buil/calibrati ... ion_en.htm

Martin and all, a detail about the observationnal point of view. For non-Alpy users, for
example LISA users, effective and quality spectrophotometry can be
made by substitue momentarily a wide slit to a narrow
slit (a one minute operaion, sure and easy on LISA/LHIRES, not possible on Alpy, it is
the primary reason of a specific double slit for Alpy ). This requires two sets of
slit and two slit supports, which has a price of course. For LISA (and LHIRES) a
standard 100 microns slit can be used with success I think.

Note that my sky is rarely photometric, I work in suburban condition. But the method
used is essentielly differential for compute the sensitivity curve and application. So under
bad condition you can expect some good precision!

The secret is really the use a relatively wide slit when doing spectrophotometry.
It is the only difficult point. The rest of the operation is trivial!

The goal is hight precision, otherwise it is not worth the trouble. It is better to work
with a standard photometer. And the important information is the independant direct
extraction of fux (erg/cm...) in the profile, not the B, V, R magnitude (anecdotic for me).

Note that I have a moderate confidence in the magnitudes given by
the AAVSO for example. The challenge and the interest of spectrophotometry is to make
independent measurements. The good side of the analysis is :

Flux profile (high astrophysics value) -> magnitude (less informative about physical phenomena)

not

magnitude -> flux profile

Christian
Francois Teyssier
Posts: 1520
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Francois Teyssier »

Christian, Robin, Martin,

J'ai réalisé le même travail que Robin sur le spectre du 29-08 et ai trouvé une différence de -3% (PhotometryV/Spectrophometry), donc trés proche de la valeur déterminée par Robin.

Sur la comparaison des deux méthodes
1. La méthode photometry V :
- il faut une trés bonne valeur de mag V (ce n'est pas le cas de beaucoup de mesures AAVSO sur la nova) ; il faut donc obtenir sa propre photométrie en parallèle.
- la méthode est "opérateur dépendante" sur le choix assez arbitraire du continuum pour déterminer l'apport des raies en émission. A ce sujet, il faut que je modifie ma page, pour tenir compte du conseil de Steve : il ne faut pas faire de lissage, se contenter de couper les émissions. VSpec le fait trés bien. Avec ISIS, le double clic est parafois difficile à maîtriser
- elle est facilement accessible.
- on peut remonter dans le temps, puisque nousa vons les spectres et la photometrie AAVSO
2. La méthode spectrophotométrique est la méthode recommandée par Steve. Celle qui est la plus robuste sous réserve d'avoir, comme le dit Martin, un ciel photometrique.
On peut également utiliser LISA avec la fente 100 mcm option. A ce sujet, pour de futures productions de fentes ce serait bien de remplacer la fente 15 mcm, totalement inutile par une 100 mcm.
Je suis d'accord avec la demande de martin d'un traitement automatisé dans ISIS, mais cela risque d'être assez difficile de déterminer les zones des raies en émission à découper.

Notez que j'utilise la méthode 1 depuis plus de deux ans pour CI Cygni : voir mesures flux absolus Ha, Hb, He II, [Fe VII]
sur la page : http://www.astronomie-amateur.fr/feuill ... Cyg_2.html
FeVII_5721_Intensity.png
FeVII_5721_Intensity.png (5.07 KiB) Viewed 7312 times
Abscisse = orbital phase

avec notemment calcul de la température de la naine blanche à partir des flux absolus Hb, HeII

Enfin, last but not least, Steve travaille en ce moment sur une méthode pour calibrer nos spectres à partir des flux absolus qu'il a calculé sur les spectres NOT et Ondrejov.

Si cette campaigne nous fait avancer sur ce sujete de calibration absolue, ce sera un grand pas en avant.

François
Martin Dubs
Posts: 141
Joined: Fri Jan 06, 2012 9:16 pm
Location: Maienfeld, Switzerland

Re: PNV J20233073+2046041 mag 6.8

Post by Martin Dubs »

François,

je suis d'accord avec presque tout de ce que tu dis. Mais il y a une chose que je ne comprends pas. Pourquoi veux tu decouper les emissions? Les emissions comptent aussi que le continuum a la Magnitude V, proprement ponderé vers lambda. Alors je ne vois pas ce problème. Prendre deux spectres et changeant la fente me semple toujours plus compliqué, il faut aussi le double effort de calibration, parce que les fentes ne sont pas exactement a la meme position apres le changement. Avec le ciel imparfait d'ici ca ne vaut pas l'effort. Moi je vais utiliser la methode 1, avec l'integration ponderée comme décrit dans le tutorial de Christian et le calcul du facteur de calibration a la main. En outre, j'aimerais aussi calibrer des spectres ou j'ai seulement le flux relativ.

Cordialement, Martin
Francois Teyssier
Posts: 1520
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen
Contact:

Re: PNV J20233073+2046041 mag 6.8

Post by Francois Teyssier »

Martin, parce que c'est *la* méthode correcte pour calibration en flux absolu avec la magnitude.
Tout le reste est approximatif, d'autant plus approximatif que l'intensité des raies est importante par rapport au continuum

Je l'a décrite sur : http://www.astronomie-amateur.fr/feuill ... ement.html

Pour nova del 2013, la correction est de 17% environ ces derniers jours.

Elle est décrité dans :

Correction of UBV Photometry for Emission Lines
Skopal, 2003

We investigate the effect on U,B,V magnitudes of the removal of emission lines from the spectra of some symbiotic stars and novae during their nebular phases. We approach this problem by the precise reconstruction of the composite UV/optical continuum and the line spectrum. The corrections Δ U, Δ B and Δ V are determined from the ratio of fluxes with and without emission lines. We demonstrate the effect for symbiotic nova V1016 Cyg during its nebular phase. We find that about 68%, 78% and 66% of the observed flux in the U, B and V filters is radiated in the emission lines. The effect should be taken into account before using the observed color indices of emission-line objects for diagnosis of their radiation in the continuum.

http://adsabs.harvard.edu/abs/2003BaltA..12..604S

and

Skopal, 2007
On the effect of emission lines on UBVR photometry


We investigate the effect on the U, B, V, RC and RJ magnitudes of the removal of emission lines from a spectrum. We determined Δm corrections from the ratio of fluxes with and without emission lines, transmitted from the object through a photometric filter. An exact and simplified approach for operative use was applied. The effect was demonstrated for classical symbiotic stars, symbiotic novae and the classical nova V1974 Cyg. It was found that about 20 30%, 30 40%, 10% and 26/20% of the observed flux in the U, B, V and RC/RJ filters, respectively, are radiated in the emission lines of the investigated classical symbiotic stars. The largest effect was found for symbiotic novae (RR Tel and V1016 Cyg) and the classical nova V1974 Cyg at 210 days (an average of 74%, 79%, 56% and 66/60%), because of their very strong emission line spectrum. In all cases, the line corrected flux points fit the theoretical continuum well. The difference between Δm corrections obtained by the accurate calculation and that given by our approximate formula is less than 10%. Deviations up to 30% can exist only in the U passband. Examples for practical applications are suggested.

http://adsabs.harvard.edu/abs/2007NewA...12..597S

Amicalement

François
Post Reply