Nova in Sco - ASASSN-21lk

Alerts and Monitoring of Novae
Peter Velez
Posts: 146
Joined: Tue May 12, 2020 7:36 am

Nova in Sco - ASASSN-21lk

Post by Peter Velez »

On 22 June ASASSN reported a possible galactic nova in Scorpio - details here:

https://www.wis-tns.org/object/2021quw

Its magnitude was up to about 13 at its peak and has come off a little since then. It was confirmed as a galactic nova on 30 June 2021 - see ATel#14744.

Hamish Barker, Terry Bohlsen and I collected some spectra in late June/early July and shared it with the ASASSN team. The results were posted in ATel#14768 as follows:

ATEL #14768 ATEL #14768

Title: Optical spectroscopic follow up of the classical nova ASASSN-21lk
Author: E. Aydi, A. Kawash, J. Strader, L. Chomiuk, K. V. Sokolovsky
(MSU), H. Barker (Rutherford street observatory), P. Velez (ARAS),
T. Bohlsen (SASER), K. Z. Stanek, C. S. Kochanek (OSU), and B. J.
Shappee (Univ. of Hawaii)
Queries: aydielia@msu.edu
Posted: 9 Jul 2021; 02:08 UT
Subjects:Optical, Binary, Cataclysmic Variable, Nova, Star, Transient,
Variables

We report on spectroscopic follow up of the classical nova ASASSN-21lk,
which was discovered by the All-Sky Automated Survey for SNe (ASAS-SN,
Shappee et al. 2014, ApJ, 788, 48) on 2021-07-22.14UT, and was classified
spectroscopically by Karambelkar et al. (ATel #14744).

We obtained several low- and medium-resolution spectra on the nights of
2021 June 28, 29, 30, July 02 and 04, using a variety of instruments, including
the Goodman spectrograph (Clemens et al. 2004, SPIE, 5492, 331) on the
4.1 m SOAR telescope. The spectra taken on the nights of June 28, 29, and
30 showed emission lines of Balmer and Fe II, but we do not resolve absorption
features in these spectra due to low S/N and resolution. In the later spectra,
P Cygni profiles of Balmer, Fe II, and O I can be resolved, with absorption
troughs at blueshifted velocities of around 1400 km/s and 1900 km/s. The
Balmer P Cygni lines are superimposed on top of a broader emission with
wings extending to 3500 km/s. The spectra are consistent with a classical
nova after optical peak. The latest ASAS-SN light curve (Kochanek et al.
2017, PASP, 129, 4502): https://asas-sn.osu.edu/sky-patrol/coor ... ab70fa6407

I attach a plot of my data from 4, 6 and 7 July. All spectra were taken with UVEX 600 lines/mm grating on PW CDK 12.5 with an Atik 460EX camera. Imaged from the iTelescope facility at Siding Springs Observatory Australia. Total integrations times varied from 4.5 hours (4 July) to 3 hours (7 July). I've had a stab at identifying a few of the more prominent lines but would welcome any suggestions/comments.

I plan to adjust the grating to access H alpha early next week - once the clouds leave!

Pete
Attachments
2021quw_20210707_397_Peter Velez.png
Francois Teyssier
Posts: 1520
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen
Contact:

Re: Nova in Sco - ASASSN-21lk

Post by Francois Teyssier »

Hi Peter,

Nice results on such a faint target (mag V = 12 to 13)

Identification
Not easy in a relatively fast nova (Vmax ~2200 km/s) with broad lines.
Balmer lines: OK - Always the first step for most or objects

Now, one must consider the state of the evolution.
The nova is about 1 or 2 mag under the max luminosity (not a robust estimation as AAVSO light curve shows only one point for the max.)
Thus, the nova is in the first decline (here, smooth, very typical)
We are still in a) optical thick and high density ejecta b) low temperature about 10 000 K of the photosphere (near the external part of the very extended remaining envelop - 1 UA as an order or magnitude).
Thus: from a) no fordidden lines which require low density and from b) no high excitation lines: all lines << 20 eV.

Step by step

1. OI 4368
1.1. If a line exists at this wavelength, it should be [OIII] 4363. But [OIII] requires an ionizing source > 30 000 K (to produce O2+ ions) and low density ("forbidden" transition from metastable level). Conclusion: it cannot be [OIII] at this stage of development.
Note: the bright OI lines in the first stages of nova are 7773 (clearly seen in the spectra secured by Terry) and 8446.
1.2. More over. Yes, there's a shoulder in the red part of H gamma. But is it a line?
This method is comparison with other lines of the same ion or atom. Here, recombination line of HI. We compare the profile with H beta.
Sco2021b_1.PNG
Sco2021b_1.PNG (18.71 KiB) Viewed 5213 times
The profiles (at least in the red part) are similar. We deduce that there's no emission line. So the identification is simply H gamma. OI 4368 should be removed.
Note: as the ejecta are optically very thick at this stage, the blue edge of Balmer lines can differ significantly due to self absorption (the physical level of the pseudo-photosphere is strongly dependent of the wavelength)

2. The series HeI 4922 [OIII]5007 Fe II 5154 Fe II 5169
2.1. The series is not homogeneous both in ionization level (He I, [OIII] in comparison to Fe II and density ([OIII] according to "permitted" lines).
Thus very improbable at this stage + see above for a) forbidden lines b) high ionized lines.[OIII] will appear later in the so called "nebular stage" when the expending ejecta will becam optically thin in the optical range.
2.2. If we have HeI 4922, we should have also He I 5016 AND He I 5876 (most often the brightest of He I lines). And He I lines requires an excitation energy > 20 eV.
2.3. When we have a series of 3 lines in the range 4900-5200, the most probable identification is the mutiplet Fe II (42) which is the strongest (or at least one the strongest) of Fe II multiplets. The nova is still in the thick Iron Curtain stage.
Thus the id. is Fe II 4922, 5016 and 5169 - Wavelength which must be known "par coeur" for Novae Studies near maximum luminosity

The identification is confirmed by a comparison of the three lines:
Sco2021b_2.PNG
Sco2021b_2.PNG (23.29 KiB) Viewed 5200 times
An interesting emission feature appears in the blue part of Fe II 5016 line, correlated with a stronger absorption of this line in comparison to Fe II 5169 (Fe 4922 is usefulness due to its proximity with H beta).
It is consistent from the time-series and also appears in Terry's spectra. When a "special" feature appears, it is important -if possible- to assure its reality by a comparison with other spectra and - if possible- acquired with other setups and observers. French: "une hirondelle ne fait pas le printemps".

It is need further investigation.

Well at this state, one can consider that the spectrum is dominated by Balmer and single ionized metals. This assumption will be a guide line for further identification: a. "permitted" lines only b. excitation level ~ 10 000 K.

My id. at this step:
Sco2021b_3.PNG
More latter (4650 and 5700 ranges)

F.
Last edited by Francois Teyssier on Sat Jul 10, 2021 12:14 pm, edited 3 times in total.
Peter Velez
Posts: 146
Joined: Tue May 12, 2020 7:36 am

Re: Nova in Sco - ASASSN-21lk

Post by Peter Velez »

Francois

this is extremely helpful. Forbidden lines at this stage are unlikely to be correct and I need to consider the energy levels for emission lines.

I need to do some more homework.

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

Re: Nova in Sco - ASASSN-21lk

Post by Francois Teyssier »

Thanks Peter,
Very interesting post, because I think that it can help other observers interested by "what in my spectrum".
I will go on in the same post (more during the afternoon of my time or tomorrow morning).
Cheers,
François
Peter Velez
Posts: 146
Joined: Tue May 12, 2020 7:36 am

Re: Nova in Sco - ASASSN-21lk

Post by Peter Velez »

Francois

Please keep going with your comments - this is the information that its so hard to find, how to "unpack" a spectrum.

This may be a basic question but I'm not embarrassed to ask it - it looks like you are using Plot Spectra to present data. How do you superimpose different parts of the spectrum eg presenting the H beta and H gamma lines together?

Pete
Peter Velez
Posts: 146
Joined: Tue May 12, 2020 7:36 am

Re: Nova in Sco - ASASSN-21lk

Post by Peter Velez »

Francois

you can disregard my last request - I hadn't appreciated that Plot Spectra has a Multi Line Plot feature for velocity. I have that working now and have been able replicate your work.

I have also rewatched your video on Nova Per 2020 which is extremely helpful. I was planning to ask about the energy required for ionisation of particular elements but found a great table early in your video. The worked example with David Boyd's spectrum is also very useful. I have a better understanding now of how I have gone off piste.

The challenge I have (and I expect I am not alone) is taking account of the expected strengths of lines. For example, you comment that 5876 is most often the brightest of the He I emissions. Where do I access this information? I could trawl through the academic publications (and contributions such as yours) but it would be good to know if there were a convenient location where this could be readily accessed.

Looking at the ASASSN data, the peak magnitude was around 12.71 about 13 days ago and its come off by about 1 magnitude now.

I would welcome any other comments you might have - I plan to keep following this one for a while. My Uvex is centred on Ha now so it should yield some interesting results

Pete
Hamish Barker
Posts: 226
Joined: Thu Jun 20, 2019 12:11 am

Re: Nova in Sco - ASASSN-21lk

Post by Hamish Barker »

Great posts! I have data from 28-30 June and 2 and 3 and 8 July. Where would be a good location to upload them on aras beam or elsewhere?
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Nova in Sco - ASASSN-21lk

Post by Robin Leadbeater »

Hamish Barker wrote:Great posts! I have data from 28-30 June and 2 and 3 and 8 July. Where would be a good location to upload them on aras beam or elsewhere?
To Francois for the ARAS database
http://www.astrosurf.com/aras/Aras_Data ... taBase.htm
and for the long term so they are accessible to future generations of astronomers, the BAA database
https://britastro.org/specdb/

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
admin
Site Admin
Posts: 72
Joined: Wed Feb 16, 2011 1:36 pm

Re: Nova in Sco - ASASSN-21lk

Post by admin »

Hamish: you can send your fit files to francoismathieu.teyssier at gmail.com and copy to arasdatabase at gmail.com.

https://aras-database.github.io/databas ... 2021b.html

Note: I don't know who invented this wrong idea that Aras Database could be ephemeris and in what intent. This not friendly.
Thanks to the pioneer involvment of Christian Buil, Valérie Desnoux, François Cochard, ... BESS Database for Be and Aras database for eruptive stars and other monitoring projects are gathering spectra since 2007. BESS is hosted by LEIDA Observatoire de Paris Medon. ASDB archives will soon be hosted by an International Institution.
admin
Site Admin
Posts: 72
Joined: Wed Feb 16, 2011 1:36 pm

Re: Nova in Sco - ASASSN-21lk

Post by admin »

The identification in this kind of spectrum is not easy due to the width of the lines and blends.

From the identification of Fe II (42), the knowledge of the development of a nova spectrum just after maximum luminosity (and the physics of the nova event), the guidelines for further identification are:
- lines from single ionized or neutral atoms. (= low ionization levels ~10 000 K)
- no forbidden lines (high density of the ejecta, = optically think)

The identification in this kind of spectrum is not easy due to the width of the lines and blends.

From the identification of Fe II (42), the knowledge of the development of a nova spectrum just after maximum luminosity (and the physics of the nova event), the guidelines for further identification are:
- lines from single ionized or neutral atoms. (= low ionization levels ~10 000 K)
- no forbidden lines (high density of the ejecta, = optically think)
The special profile of Fe II 5018 is clue for further identifications.

Two prominent blends appear in the range 5600-5900.
In this range, Fe II lines are very rare (Fe II 57 for instance), thus we must have other species.

We begin by 5800-5900 range, which is easier.
From 1) the stage of the first decline (mag = mag max +1 ) -thanks for the input, Peter- 2) the absence of He I 6678, 7085 (see Terry's spectrum), we deduce that He is still neutral, thus He I 5876 doesn't appear.
At this stage, the very low ionized Na I D (IE = 5.1 eV) is often detected in emission. Thus It is very probable.
We retain Na I as a component of the blend.

Other lines which can be detected in the range 5600-5900 and at this stage, especially in fast novae, are NII, O II (but also other metals such as Si II or Mg II)
N II: 4643, 5001, 5176, 5180, 5478, 5667, 5731, 5940, 5942, 6137, 6341, 6357, 6482
NII 5001 is especially interesting, because it could be the blue shifted component of Fe II 5018.

We can try these identification as a temporary attempt.
It will be easier to decide when futher spectra will be secured by an examination of the "trajectories" of the lines in a stacked graph.

More later.
François
Post Reply