NGC 3597 nova?

[geckzilla]

Just doing my usual thing, searching the HLA for something to process. NGC 3597 is a galaxy undergoing some kind of merger so while trying to get a wider view of it I decided to combine two data sets to see how they'd look even though it didn't seem promising because they were too different. Of course, I quickly spotted this blip. I searched around to see if anyone had made mention of it before but could find nothing related to any nova, supernova, light echo, or whatever else it could be. Maybe I didn't look hard enough or don't know where to look. Well, anyway, there it is. I find it strange that no one would have noticed.

Click to view full size image 1 or image 2

...

[/hide]

[BDanielMayfield]

If the object is part of NGC 3597 it would have to be a Supernova, wouldn't it? If it was a nova it would have to be from a foreground star in our galaxy, I think.

When were the two images taken?

Bruce

[geckzilla]

I didn't want to say supernova because there are potentially some other things that could be that bright (I think? My ignorance is showing... help!) so I went with the more generic word nova. It's almost as bright as the core of the galaxy but not quite. The object is present in 1997 and absent from observations 1994 and 2000. Those posted images are from 1994 and 1997.

[Ann]

Very interesting, geckzilla. I have posted a question to David Bishop, the man running the site http://www.rochesterastronomy.org/supernova.html, which keeps track of the latest supernovae. Let's see if David replies to my question, and if he knows something.

Ann

[saturno2]

I think that this object is a nova

[geckzilla]

Owlice told me to write to CBAT about it, so I did. I hope I did it right.

[Ann]

I just received a reply from David Bishop. He had no information about a possible supernova in NGC 3597.

David Bishop wrote:
Nothing. Not even in the boneyards (where I put objects that are not supernovae).

Better luck with CBAT, geckzilla.

Ann

[Ann]

The point of light can be seen right next to a dust lane in NGC 3597, exactly the kind of place where interesting things may happen. Also, the object has the right kind of luminosity to be an exceptional but transient phenomenon in this galaxy. That is one reason why I, from my complete amateur perspective, find it hard to disregard this point of light as product of some kind of photographic flaw or mistake rather than a real phenomenon.

So for what it is worth, I believe that it was real. But what was it?

Let's assume that it was in fact a supernova, which exploded in 1997 (when it was detected by Hubble and was recorded in the data you used to process this image, geckzilla). How likely is it that a supernova in galaxy 3597 might have gone unnoticed?

NGC 3597 is located in the constellation Crater, quite well placed for most telescopes in the world. It is hard to believe that a supernova would go unnoticed because it was seen in the direction of the constellation Crater. But how thorough were the supernova searches in 1997?

It wasn't too long ago that the supernova searches in general were extremely "deficient" due to a lack of good telescopes. Consider supernova 1987A. It was discovered on February 24, 1987, yet it was the first supernova discovered that year! Things had improved in 1993, when 1993J was discovered on March 28 as the tenth supernova of that year. But things are very different today. Supernova 2014A was discovered on 2014/01/01.580.

My point is that it might just be possible that a supernova in NGC 3597 might have gone unnoticed in 1997. Isn't it possible that the directors of the Hubble telescope might have ordered the telescope to photograph NGC 3597 without realizing that the galaxy might contain a recent supernova? Isn't it true that Hubble has been ordered to photograph enormous numbers of galaxies? Is it really so strange if there was an as-yet unrecorded supernova in one of them?

As a color commentator, I note that the transient light source in NGC 3597 looks slightly red. This may mean nothing. Supernovae type Ia are blue when they are young, but if they are partly obscured by dust, they will be reddish. The mysterious light source in NGC 3597 is near a dust lane, so it might of course be dust-reddened. And other types of supernovae, notably types II and Ib, may well be yellowish right from the start. In any case, all supernovae turn yellow and red with age.

Another interesting possibility is that this light source might perhaps not have been a supernova at all, but a luminous red nova. Such an object would have been less luminous than a supernova, and it would have faded faster.

I keep thinking that this light source was for real, and something went on in galaxy NGC 3597 which was photographed by Hubble in 1997, even though it went unnoticed and unrecorded by the astronomical community.

Ann

[geckzilla]

No belief necessary. It's a real thing. I checked and it was on all eight exposures. By lucky coincidence, the galaxy was being imaged deeply when the event occurred.

[Ann]

No belief necessary. It's a real thing. I checked and it was on all eight exposures. By lucky coincidence, the galaxy was being imaged deeply when the event occurred.

Congratulations on this discovery, geckzilla! It's really very well done!

I believe there are tons and tons of Hubble data that is pretty useless until someone bothers to look into it, like you have done.

Again, congratulations!

Ann

[geckzilla]

Reading this paper is kind of driving me crazy.
http://articles.adsabs.harvard.edu/full ... 3.000.html

They repeatedly mention both Hubble data sets which existed at the time it was written as well as some other data but it's all about the evolution of globular clusters and the evidence for a second core which indicates that it could have recently merged with another galaxy. I'm finding it impossible to believe that for all this scrutiny, no one has noticed the so-called nova. Clearly the article is all about globular clusters but it seems strange to think that it's not even worth mentioning the transient object. I'm starting to think I am the one missing something. I mean, are they so common that only certain ones are noteworthy enough for mention?

[Ann]

geckzilla wrote:
I mean, are they so common that only certain ones are noteworthy enough for mention?

Absolutely not. No way.

How often do really bright lights appear in the Andromeda galaxy? How often do we see them in M33? In the Magellanic Clouds?

These things are rare! They are not commonplace at all.

They repeatedly mention both Hubble data sets which existed at the time it was written as well as some other data but it's all about the evolution of globular clusters and the evidence for a second core which indicates that it could have recently merged with another galaxy.

Bahh!!! The Andromeda galaxy has a second core (not really, but sort of). And globular clusters are all over the place! You even find them in some itty bitty dwarf galaxies. Why do they have to look at them in NGC 3597? And what do they mean about the evolution of globular clusters? How much do they think that a twelve billion year old globular will evolve in four years or so?

Don't give up on this, geckzilla. Somebody has to pay attention to that bright nova, or whatever it was.

Ann

[neufer]

Click to view full size image 1 or image 2

Just doing my usual thing, searching the HLA for something to process. NGC 3597 is a galaxy undergoing some kind of merger so while trying to get a wider view of it I decided to combine two data sets to see how they'd look even though it didn't seem promising because they were too different. Of course, I quickly spotted this blip. I searched around to see if anyone had made mention of it before but could find nothing related to any nova, supernova, light echo, or whatever else it could be. Maybe I didn't look hard enough or don't know where to look. Well, anyway, there it is. I find it strange that no one would have noticed.

Are you quite sure that your images are of NGC 3597

[Ann]

That's a thought. Could the galaxy with the nova possibly be another galaxy than NGC 3597?

I checked out NGC 3597. It is moderately small, about 2.2' X 2.1'. The galaxy in the image you have processed appears to be too large to fit into Hubble's field of view, which is presumably the reason why we only appear to see the central part of it.

Would a galaxy the size of 2.2' X 2.1' fit into Hubble's field of view so that we could see all of it?

Ann

[geckzilla]

Are you quite sure that your images are of NGC 3597

Quite. Here's a widefield view of it. The previous image was of just the core.

[BDanielMayfield]

Are you quite sure that your images are of NGC 3597

Quite. Here's a widefield view of it. The previous image was of just the core.

ngc3597.jpg

When you inlarge this widefield veiw you can recognize the same structure as in the first two images, and you can also see geckzilla's "nova" object too.

Is this larger veiw from the same data set with the "nova" or was it taken at a different time?

Bruce

[geckzilla]

The left side is the older image which is less deep (you can see how much more grainy it is) and the right side is the newer, deeper image which also contains the nova. Since they don't completely overlap there is that blank spot.

[Ann]

Yes, now I can see that the galaxy in the image you have processed is small enough to be NGC 3597, and I also recognize the outline of NGC 3597 from my software. The right side of the galaxy is clearly the same as my software shows me. If the nova is in the same dataset as the one that shows the right side of the galaxy, then the nova is definitely there.

Could you begin to estimate the brightness of the nova, geckzilla? My software says that the overall brightness (V magnitude?) of the galaxy is 12.7. But it also says that the B magnitude of NGC 3597 is 13.689 ± 0.089, and moreover it says that the B-V index of the galaxy is 0.610. (I don't really get that, because if the V magnitude is 0.610 magnitudes brighter than the B magnitude, then the V magnitude should be about 13, not 12.7.)

Anyway, a B-V index of 0.610 means that there is definitely quite a bit of star formation in this galaxy (since a "red and dead" galaxy will have a B-V index of about 0.90 or more), and the presence of star formation makes a nova more likely. Your picture shows the graininess of star formation in the center of the galaxy.

Ann

[geckzilla]

I have no clue how to determine the magnitude of the thing. The best I can describe it is just slightly dimmer than the brightest pixel of the galaxy nucleus. I did look up the magnitude of the galaxy at some point and consider describing it based on that, but then I decided that since I don't how exactly galaxy magnitudes are calculated that it might not make any sense to do so.

[geckzilla]

I got around to processing this galaxy last night and the core this morning if anyone is interested. Nothing new but it's less rough. For the core there are some notes you can hover on the Flickr page if you click through.

[Ann]

Any progress on this, Geckzilla? You deserve to be recognized by the astronomical community for your discovery.

Ann

[geckzilla]

Yeah, they gave me a gold medal and a trip to Tahiti. I'm typing this on the fancy computer that automatically connects to the Internet from anywhere that they gave me. This yacht is awesome.

[neufer]

Yeah, they gave me a gold medal and a trip to Tahiti. I'm typing this on the fancy computer that automatically connects to the Internet from anywhere that they gave me. This yacht is awesome.

And no one deserves it more.

[geckzilla]

Joking aside, Alice emailed someone who was interested in analyzing the images for us. Here is what Professor Schaefer had to say about it:
(I felt it prudent to remove email addresses and the phone number.)

Hi;
In looking it over, I found that the original HST paper on these images (Carlson et al. 1999, AJ, 117, 1700) did not mention the new star, but it appeared odd (as in color separated) in their color composite. I also checked all the HST images and came to the same conclusions, that the image looks real, is a point source, did not move over the course of four hours, and (without real photometry) is apparently fairly red in color. I estimate that the B magnitude is 22, but it will take real photometry to get a an accuracy better than a magnitude or so. I cast around for any other deep photos, but I soon realized that at B~22, that HST is the only game in town. I checked and the galaxy was near enough to opposition, so there is no visibility reason to have missed a peak in the Spring/Summer of 1997. The Asiago catalog has no SN in the galaxy at any time.
I have a vast experience at checking out 'guest stars', starting in 1983, so I know the usual variety of things to check. Here is a list of possibilities:
***Image artifacts. These come in many varieties. The existence of good PSFs at the exact same location on many images through two filters over 4 hours completely rules out artifacts.
***Solar System objects. Very faint asteroids are a dime a dozen and not catalogued. But all (even KBOs) would have easily recognizable motion (the galaxy was far from any stationary point) over the four hours between the two orbits with images. Even the parallax from the HST motion would have been visible in one image alone.
***Flare star. OK, so it is a real astrophysical object outside our Solar System, but it could still be a foreground star in our Milky Way. Such a candidate would have to flare in brightness, and the first try is a flare star of some type. But this star has a flare of many (>~5) magnitudes, and this would make it a very rare flare star. A near-killer for the flare star idea is that the star had comparable brightness over a time separated by four hours, and flare durations are ~2 orders of magnitude smaller. A killer for the flare star idea is that the apparent color is reddish, while flares are super-blue. So this idea is dead.
***Dwarf nova. Dwarf novae peak at M~4.5 mag in eruption, while the new star has B~22, so the distance modulus is around 17.5. This corresponds to a distance of 32 kpc. This is pretty far outside our Milky Way. Faint DNe are a penny a dozen, but this is too faint, forcing the system to be far outside our galaxy and making this solution very unlikely.
***Nova. Novae peak at M of -8. This makes for a distance modulus of 30 mags (or smaller if not at peak). The corresponding distance is 10 Mpc, which is far outside our Local Group, even with the nova not being at peak. Well, what about the nova being in NGC 3597? The galaxy is at 50 Mpc (NED), with a distanc modulus of 33.5. So a nova in NGC3597 would appear at peak as 25.5 mag, with no possibility of being much brighter. I think that this is greatly and significantly fainter than the observed new star. So the nova idea is ruled out for it being either in our Local Group (it is much too faint) or in NGC 3597 (it is much too bright).
***Supernova. Depending on the type, they can peak at M from -20 to -16 (with rare ones outside this range). For a distance modulus of 33.5 (for NGC 3597), this corresponds to peak magnitudes of 13.5 to 17.5. This range at peak is much brighter than the observed magnitude, which I think is ~22. A possible resolution is that the SN is far past peak, with half a year being adequate. (Another resolution is that the SN is highly reddened, but this is incompatible with the observed color not being very red.) If the SN is half a year after peak, this would easily explain why the SN at peak was *not* discovered, simply because NGC 3597 would have been around conjunction at the time. (Also, the vast efforts to check many galaxies each night were not going fast at that time, for example with KAIT starting in 1998.) Supernova at late time are moderately red, so this all ties together.

So in all, the only remaining explanation, and this is an ordinary one, which should be quite common, is that the new star in NGC 3597 is a supernova of unknown type in that galaxy, with the peak having occurred roughly half a year earlier.
So what to make of this? Well, if it had been caught at peak (and had been observable), then this would have been one of the brighter events and well observed and hence important. Not catching it at peak, having no light curve, and even having no SN-type means that there is little of use that can be recovered now. The one possibility is that we do have an HST image of it and so can get a very accurate error circle, and other HST images can be used to look very deeply for any progenitor. But there are no prior deep HST images and the distance to NGC 3597 is so large that constraints would be largely useless, not to mention that the lack of a known type would cripple any conclusion. So I conclude that there is not useful science to be done with this discovery.
Nevertheless, it should still be correctly reported to the IAU people. Your initial contact with the CBAT should start things off OK. Likely, you should have real photometry (for F450W and F702W). The best source is Matthew Carlson, as he is first author on the HST paper, and he certainly already has perfect photometry that he only need spend five minutes looking up for you. He is at New Mexico State University in Las Cruces, at [email address]. You could also try Jon Holtzmann at [email address]. Both should be happy to help.
There is some utility just in the whole story of the discovery. It is a nice one, and it does have the moral lesson that substantial numbers of SNe are missed. An appropriate forum would be APOD, where perhaps some image blinking between the two would be good. This would be doubly good for APOD, both as a galaxy picture, but also as a fun story with things changing.
I would be happy for Judy to use this email with Dan Green at the CBAT to provide supporting analysis for a SN identification. I am happy to have you post this email (or part of it) to the Starship Asterisk page, and indeed, I request that you do so because my analysis provides detailed calculations, long experience, and a likely real answer.

Bradley E. Schaefer
Professor, Department of Physics and Astronomy
Louisiana State University
Baton Rouge, Louisiana 70803, USA
[email address]
[phone number]

[Beyond]

Way to go,