30 replies to this topic

[robin_astro]

A low resolution spectrum of SN 2024gy in NGC4216  at ~mag 15 here (so an easy target for the ALPY200) and still climbing

 

The guider image

 

 

The spectrum (black) compared with a matching spectrum from the supernova identification program SNID  (red)

 

 

Also uploaded to the Transient Name Server website
https://www.wis-tns.org/object/2024gy
(note the good agreement there with the professional spectrum used to classify it 2 days earlier)

 

Cheers

Robin
 

 

[StupendousMan]

Very impressive!

[mborland]

Robin,

 

I'm going to try to get a spectrum of this supernova late tonight with my ALPY 600. Can you tell me which reference star you used?

 

The reason I ask is that when I search the reference list at http://www.astronomi...tsSpectro0.html the closest star is HD111397. It's not really very close. Also, it isn't in the ISIS database. I'm thinking of just using a nearby star with a known spectral type.

 

Thanks--Michael

Edited by mborland, 14 January 2024 - 12:27 PM.

[robin_astro]

I used HD103578  A3v 

 

 

Cheers

Robin

[mborland]

Here's my spectrum compared to the others from https://www.wis-tns.org/object/2024gy. I'd say mine looks similar to Robin's and the professional spectrum, but with clear differences. I suppose some of this could be actual evolution of the supernova spectrum. I hope to get another chance to measure the spectrum a week or so from now to see if it is changing.

 

 

Some details of my acquisition:

* ALPY600 with calibration module mounted on Esprit 120 ED.

* ASI533MM science camera at -20C and gain of 400.

* Used HD103578 for response calibration, per Robin's recommendation above. Recorded 100 spectra with 6s exposure. These were taken with the star >25 deg above the horizon.

* Recorded 19 spectra of SN2024gy with 720 s exposure. These started at about 35 deg above the horizon and ended at transit (about 4 hours later).

* Processed mostly with ISIS. The instrumental response correction for SN2024gy was done separately since I needed to smooth out some spikes and it was easier that way.

 

One concern I have is that I measured the instrument response with the star pretty low in the sky compared to the measurements of the SN. It was either that or stay up all night to take the HD103578 frames in the middle of the run. Unfortunately, plate solving isn't accurate enough to allow me to switch targets unattended; I need to manually place the target in the slit.

 

Comments and suggestions welcome.

 

--Michael

 

[robin_astro]

 I'd say mine looks similar to Robin's and the professional spectrum, but with clear differences. I suppose some of this could be actual evolution of the supernova spectrum. I hope to get another chance to measure the spectrum a week or so from now to see if it is changing.

 

 

yes the spectrum will evolve significantly with time (both in the features and in the wavelengths as the velocity decreases. (The velocity does not actually decrease significantly but we see deeper into the explosion as the material expands and cools.) By comparing spectra with a database using programs such as SNID

https://ui.adsabs.ha...07001B/abstract

and GELATO

https://gelato.tng.iac.es/

for example you can estimate the age and redshift as well as the supernova type 

 

(Not the same type as this one but there are some animations of  the evolution of SN 2023ixf in M101 last year (I am still following it spectroscopically)

https://britastro.or...f9ec183a85f2b58

 

Cheers

Robin

[mborland]

Robin,

 

Thanks for the tips as always. According to GELATO, my data shows the SN is "100% type Ia". I feel as if I've won a prize.  

 

Those animations are very interesting and show impressive dedication. Fortunately, SN2024gy is going to get easier to monitor as the months go by, so I hope I can attempt something similar.

 

--Michael

[robin_astro]

I managed another spectrum last night (with the ALPY600 this time as it is much brighter).  Plotted together here to show  the evolution over 10 days

https://britastro.or...14785&multi=yes

 

Note how several of the features have shifted to the red, particularly the broad absorption around 6100A due to Si II, a defining characteristic of type Ia. The wavelength at rest should be at 6355A but it is blue shifted by the velocity of the ejecta. Once the redshift of the galaxy is taken into account, the velocity was initially ~15,000 km/s slowing to ~11,000 km/s, typical values for a type Ia supernova explosion

 

Cheers

Robin

[mborland]

Robin,

 

Is your .dat file available for the latest measurement? I'd like to overlay it with mine, which was taken just a few days earlier. By eye they seem pretty similar.

 

Here's mine:

https://drive.google...iew?usp=sharing

 

--Michael

 

[robin_astro]

Hi Michael,

 

All my spectra are available in the BAA spectroscopy database as 1D fits files which should be readable by any astro-spectroscopy software. If you go here

 

https://britastro.org/specdb/data.php

 

Enter SN 2024gy in  the search box, click "filter object list" and select it in the "Object" list. If you then click "Fetch Spectra" you will be presented with a list of spectra for that object in the database where you can view the details, view the spectra or download them.

 

Cheers

Robin

[mborland]

Robin,

 

I compared your two spectra with mine. Over the full range, there's quite a bit of disagreement on the intensity. My result seems to have relatively higher signal at short wavelengths.

I'm normalizing each curve to a maximum value of 1.

 

However, if I eliminate data below 4750 A, the agreement becomes better.

(The normalization changes and uses only the data shown.)

 

I'm surprised by the differences because we used the same calibration star, which I would have thought would compensate for differences in location and equipment. Perhaps I did the response calibration incorrectly.

 

--Michael

 

[robin_astro]

Hi Michael,

 

I normalise all my low resolution spectra to 1 at 5500A (Actually the mean 5400-5500).

 

We can discount the early spectrum because it will have evolved significantly but If you divide your spectrum by mine for the similar date you will see there is a roughly linear large slope difference between them so there is likely a problem with the response correction. I will run some more checks on my data but I did not notice any obvious issues.

 

 

It is difficult to troubleshoot on what is a varying target like this but problems like this are not uncommon and there are many possible causes  which  can range from the raw data to the data reduction. Have you seen any problems like this with your data before, for example when recording targets with known spectra eg by doing exercises like these?

 

http://www.threehill...troscopy_21.htm

 

Also the document at the top of the page describing how to calculate the response has some useful tips.

 

Cheers

Robin

 

 

 

 

[robin_astro]

One quick  thing you can check is how much the spectrum changed  during the run due to air mass or changes in the wavelength sampling at the slit due to atmospheric refraction or a combination of chromatism and focus.  (At 25 deg the atmospheric extinction due to airmass will be high)

 

On a long run like this which things can change I would normally take reference star spectra at start and end.  (On this occasion SN 2024gy rose from 34 deg to 47 deg and I took a single reference star spectrum part way through the run at 42 deg)

 

Robin

[robin_astro]

One quick  thing you can check is how much the spectrum changed  during the run 

ISIS stores all the individual spectra ( @pro ) if you have it set to save the intermediate files.

 

Here are my first and last. You can just pick out the effect of reducing air mass at the blue end  (~10-15%) but it is much smaller than the difference between our spectra

 

[robin_astro]

  (At 25 deg the atmospheric extinction due to airmass will be high)

 

 

We can use Francois Teyssier's "reference star finder" spreadsheet to make a rough estimate. It depends on the atmospheric conditions (AOD) but if we use the default figure of 0.08 and assume the reference was measured at 25 deg and the target at a mean 45 deg altitude then the result relative to 5500A would be ~10% low at 7500A and ~35% high at 4000A

 

 

ISIS also has a tool to generate atmospheric transmission curves which could be applied to the spectrum to estimate the effect ("Tools" "Atmosphere")

[mborland]

Robin,

 

Thanks for all the analysis, information, and ideas. I'm still pretty new to spectroscopy, having only started in late November, so I'm probably doing something wrong. I'll play with some of these ideas over the weekend and see what I can learn.

 

Last night I took data for two reference stars near NGC 4216, one being significantly higher in the sky than the other. I'll look at these and see if I get similar response curves.

 

One thought I had is to find a reference star that is near the middle of the nightly path that will be followed by the object I'm interested in. That way, I can take the reference data at the beginning of the evening, set up to follow the object of interest, and go to sleep. The downside is that if conditions change during the night, I'll introduce another type of error, but that's always the case to some extent.

 

This could be accomplished approximately by a simple modification of the reference star search. Instead of searching for the reference star closest to the (RA, DEC) of the object, I'd search for (RA-H/2, DEC), where H is the number of hours of data I'm planning to acquire.

 

--Michael

[dhferguson]

Cheers,

 

FYI, here is a heavily cropped image of NGC 4216, pre SN, taken about ten months ago. Bortle 6.5, C8 Edge x0.7, ASI2600MC, 114 min total exposure.

 

Happy observing always,

 

Don

 

Attached Thumbnails

• 

[mborland]

Cheers,

 

FYI, here is a heavily cropped image of NGC 4216, pre SN, taken about ten months ago. Bortle 6.5, C8 Edge x0.7, ASI2600MC, 114 min total exposure.

 

Happy observing always,

 

Don

Don,

 

It is definitely cool to just look at this galaxy and see that there's a star visible where none was before.

 

--Michael

 

[robin_astro]

 

One thought I had is to find a reference star that is near the middle of the nightly path that will be followed by the object I'm interested in. That way, I can take the reference data at the beginning of the evening, set up to follow the object of interest, and go to sleep.

Yes Francois Teyssier's spreadsheet has a tool for this. (The "trajectoire" tab). You can enter the observing times of reference  and target and match the times to minimise the altitude error. Downloadable here

http://www.astronomi...tsSpectro0.html

 

As you are using a refractor you also need to consider the possibility of chromatism in the telescope optics which will make the spectrum sensitive to focus drift, affecting the wavelengths sampled by the slit and hence the continuum shape. Do you see any evidence of this eg any sign of fishtailing at the blue end ? Try taking a series of spectra at different focus settings and see if the spectrum shape changes.   See also this post

https://www.cloudyni...600/?p=13200412

 

Cheers

Robin

[mborland]

Robin,

 

Following your suggestion, I looked at the spectra of SN2024gy over the night. There is indeed quite some effect from altitude

The effects at the short-wavelength end are quite significant, probably exceeding a factor of 2!

 

I took a spectrum of HD97633 at the beginning of the session when it was at about 34 deg altitude, which gives me a response function for that altitude. I think I can use my SN2024gy data to create a model of how the intensity for each wavelength changes with altitude, then use that to map everything back to 34 deg. Once I've done that, I can use the HD97633 response function to correct for instrumental and atmospheric effects at 34 deg altitude.

 

I didn't see any obvious sign of effects from chromatic aberrations in my scope.

 

--Michael

 

[robin_astro]

I am pretty confident my spectrum is OK but as a rough sanity check of the slope in my spectrum to make sure there is nothing seriously wrong I had a look at the multiband photometry in the AAVSO database. For the date of my observation B-V was ~ +0.25 so similar to a late main sequence A star. Overlaying the A7v pickles standard roughly  matches the broad slope suggesting all  is OK (but a much different spectrum in detail of course)

 

[mborland]

Robin,

 

I performed the above-described correction to my SN2024gy data from 1/18. I started by heavily smoothing the SN2024gy spectra, which allowed me to fit the altitude dependence to a quadratic function for each wavelength. That allowed me to map all the spectra back to the same altitude, that at which I took the reference star data.

 

Since I took data for two reference stars (HD103578 at 23 deg) and (HD97633 at 34 deg), I did this twice, obtaining the following altitude-adjusted spectra

 

You can see that the adjustment is not perfect, but it is a big improvement over the original. With these in hand, I summed over all the spectra in each set and then applied the response function for the appropriate star. This allows me to compare my final spectra to yours

As you can see, my results are much closer to yours now. The comparison for "23 deg" (i.e., mapping to 23 deg altitude and correcting with HD103578) is closer at the blue end. I tend to trust the "34 deg" data more, however, because the altitude adjustment gave more consistent results. In some regions, the agreement is really remarkable.

 

Comments welcome.

 

--Michael

Edited by mborland, 21 January 2024 - 08:55 PM.

[mborland]

I had another opportunity to take a spectrum last night. I applied the same correction technique to this data as well as my older data from 1/14. The spectrum seems to show some trends over the last three weeks, both in terms of the position of various dips/bumps (broad lines) and the overall intensity in various regions.

 

 

--Michael

 

[robin_astro]

Hi Michael,

 

You might be interested in this thread in the ARAS forum following another bright type Ia SN 2019ein.

https://www.spectro-...opic.php?t=2308

 

Even though they are used as "Standard Candles", type Ia are not all the same

 

You can plot the velocity of the ejecta against time 

https://www.spectro-...?p=12679#p12679

and use the relative strength of the lines at maximum to classify them according to the "Branch" classification system

https://www.spectro-...?p=12716#p12716

 

Cheers

Robin

[robin_astro]

and SN 2020ue another bright Ia

 

https://www.spectro-...opic.php?t=2457