Measuring the magnitude of a Supernova

[rogerg]

Hi all,

Last night I was photographing NGC 1954. I (for once) decided to compare the photo's I was capturing with those I took previously, in this case December 2009. I was surprised to see what looked like a stereotypical supernova in the arm of the galaxy! A quick google search confirmed it to be a supernova discovered almost exactly a month ago 05/12/2010.

So, I set about trying to work out more of this opportune example.

I have images which I'm putting through Astrometrica and CCDSoft in an attempt to measure the magnitude of the supernova. Astrometrica is giving me reasonable results I think, but I'm having trouble (newbie alert) determinig the validity of the results.

I have attached a series of images. Hopefully one of the experienced supernova hunters could give me some feedback.

From what I can assertain (by reading the astrometrica group and the astrometrica help file) the "Fit RMS" of 0.041 for the supernova star is a "good" value, adding weight to the calculated magnitude of 16.3. I am more unsure of the other values.

Having 7 images open in Astrometrica the magnitude calculated is 16.3. This matches reasonably well with the other observations (http://www.rochesterastronomy.org/sn2010/sn2010ko.html) so far, except for one, the most recent (5th Jan) which is actually prior to my image but brighter than my image at 16.9 (the others indicate the supernova to be increasing in magnitude, if mine is correct it is now decreasing).

Anyone care to comment?

Its the first time I've photographed one so am quite enjoying spending the day going through trying to process the image and get something useful from it.

Thanks,
Roger.

[higginsdj]

The fit RMS is simply the fit to a stellar psf curve. Astrometrica, in general, does not use photometric catalogues (it does have one that is being tested for photometric accuracy but it has very limited coverage). If you look on the second image you will note that the dmag values for the images range from 0.2mag to 0.3mag. (This gives you an indication of the level of accuracy in your images using whatever catalogue you are using). Some areas in the astrometric catalogues are known to be well off in magniutde (up to 2 mags worse case).

How well have you accounted for linearity in your images? How about saturated stars? You need to realise that Astrometrica was not designed with photometry in mind.

Cheers

[rogerg]

Thanks for your reply David,

There is one star that is saturated. Would that be excluded or can I manually exclude it in some way? I don't understand what you mean by linearity, I will read up on it tomorrow (run out of time right now). The images are quite consistent in terms of overall image quality, and al have a slight amount of RA drift evident, if that is what's meant by linearity (I doubt it). I will look up what programs othe than Astrometrica are suitable too (more suitable), and continue trying to work with CCDSoft. Think I need to work out what a "good" number for dmag is too. It's great that Astrometrica help has explanations of these terms but it (probably intentionally for good reason, like context) never says what value is expected/good :)

Thanks,
Roger.

[higginsdj]

Linearity is about how linear (consistent) your CCD chip is at gathering photons. ie my ST-8E is linear from 7K to 38K even though it saturates at 65K. So, as an example, if I imaged twice as long, I get twice as many photons 'counted' in my linear section. This is NOT the case outside the linear section. This means both my target and ALL my comp stars must lie within the linear band of my CCD.

So if I use stars who's peaks are >38K or <7K then my photometry will be off.

Image quality is not terribly important. I can still get 0.03mag precision from mildly trailed images it's the SNR of the target and comp stars that is important.

MPO Canopus is designed for precision photometry work (both asteroids and Stars) be it differential or absolute. The most important thing, after your images, is the catalogue that you use. UCAC2 and 3, USNO A2 and B1 et al are astrometric and not photometric catalogues.

Whilst you use astrometrica and astrometric catalogues you aren't going to get a good dmag value. If you want to measure and report Sn brightness then you need to be able to determine the magnitude reliably within +/- 0.1mag at the very least. Its just a case of expand your toolset and skillset.

Cheers

[rogerg]

Thanks David, it'll take me a little time to progress from here and digest the information you've given me properly, so won't try responding to everything right now.

Presumably by 7K you mean 7k ADU, so where objects are between a birghtness of 7000 and 38000 ADU on your chip? If convenient to explain how I could measure the linearity of my ST7 that would be apreciated otherwise I'll go hunting the information later when I have time.

MPO Canopus, I think you have mentioned that to me bofore, I'll check it out. So many software titles, can't remember them all.

Thanks,
Roger.

[rogerg]

I just realised this is probably as easy as matching a star field of a large number of reference stars and then looking at how the calculated brightness compares to the theoretical brightness of stars across the full ADU range.

[higginsdj]

No. You want to take 'flats' (dark and bias adjusted) without optics and using a stable light source (ie non varying). You then image for 2s, 4s, 8s, 16s, 32s, 64s, 128s, 256s etc until saturation (your light source needs to be reasonably faint to do this else you will saturate too early).

You then plot the average ADU of each image (you might want to take say 5 images at each integration time and average the results) and see where the CCD plot stops producing a straight line at both lower and upper limits.

Cheers