Nova Cen 2013 - the brightest nova this century
On Monday, December 02, Australian amateur astronomer John Seach (Chatsworth Island, NSW) discovered a nova in Centaurus, not far from beta Centauri.
A nova is a massive nuclear explosion on a dying star. These stars - white dwarves - are the final evolutionary stage of Sun-like stars. A typical white dwarf has a mass similar to the Sun but its size is similar to the Earth: a teaspoonful weighs a ton. The nearest known white dwarf is Sirius B.
About 40 stars go nova each year in our Milky Way, but only about 10 of these are actually observed. Once a nova is observed, a light curve can be compiled, and based on the curve, it is classified as either a fast, slow, very slow, or recurrent nova. A fast nova drops by 3 magnitudes from maximum within 100 days, while it takes 150 days or more for a slow nova to dim by the same 3 magnitudes. Very slow novae remain at maximum light for a decade or more! Recurrent novae have at least two observed outbursts, separated by decades.
The brightest recent nova was Nova Cygni 1975, which reached mag 2.0. Recently (2013 August) Nova Delphini reached mag 4.3, making Nova Cen 2013 (so far) the brightest nova this century.
See the ASSA "Nova Centauri 2013" webpage for more details.
This morning, 2013 Dec 19 at 00:56:51 SAST, I estimated the nova as V=5.1, from a single image (ISO 1600, 20-seconds) analyzed in APT (green channel only). Comparison stars were HR 4888, HR 5241, HR 5349, and HR 5140.
Peter Wedepohl of Somerset West let know this morning that he bagged Nova Cen this morning as 5.0.
Bruce Tomalin of Durban writes: "Saw it again thsi morning 19/12/2013 in between wispy clouds form 3h20 to 3h50 am. The reference star I referred to above appears to be 334G. Cen at 5.22 mag. This morning the nova was not visible to naked eye but in binoculars in same field of view as 334G. cen appeared to be the "same" brightness. So to me, the nova seems to have dimmed a bit since last seeing it on the 13/12/2013 (I missed the brightening in between!). Good forecasts for next few days so excited to see what it does..."
Dramatic clouds last night, giving a sombre sunset, gave way to whispy rubbish littering the sky later in the evening. By 03:30 things were much better and I got a series of images of the nova. Thanks to failing entirely to understand how my intervalometer works, I got series of 4, 10, and 22-s images, instead of 6, 12 and 24-s as I had planned. These were shot at ISO 640 with a 22-mm lens at f/5.6. Naturally, I took the dark frames at ISO 320, which led to surprising vocabulary shortly after 04:30.
In IRIS and APT I analysed the green-only 10-second images (a single RAW frame, a 15x10-s stack without darks, and a a 15x10-s stack with (proper) darks). The three sets give similar results; my conclusion is that the nova was V = 4.4 +- 0.1 on 2013 Dec 16 @04:00 SAST. (Or 4.42+-0.06).
This is a sharp drop in brightness!
An e-mail from variable star observer Peter Wedepohl from just around the corner confirms this: Peter writes: "Brightness has started dropping very suddenly. At 03h30 today it was down to 4.5. Fascinating - it will be interesting to see what the nova experts make of this." Exciting times!
Ridiculous and horrible little clouds were loitering about last night, just aimlessly obscuring the sky without any clear purpose or life plan. Fortunately at some point in the morning they found direction and the four-o-clock alarm call wasn't a waste, revealing the nova - bright! - shining above the distant mountains and on a rapidly brightening sky.
I took a series of 6-second exposures (50-mm f/5.6, ISO 320) and, using darks, stacked them in Deep Sky Stacker (2'03" total). The green channel was split out and imported into IRIS, as well as Aperture Photometry Tool (v.2.4.2), for aperture photometry. Eight stars were selected as comparison stars, with V magnitudes taken from SIMBAD. My conclusion is that the nova was V = 3.3 +- 0.1 on 2013 Dec 14 @04:38 SAST.
This morning, despite weather predictions, I stepped outside to be pleasantly surprised by the Pointers poking up behind the neighbour's rooftop. [ Insert gear-up montage sequence ]. The nova was obvious on the first photos, so I took a series at difference ISOs (to control the noise) and exposure times (to reduce trailing), making sure the field was wide enough to include useful comparison stars at approximately the same altitude (to reduce the gradient problem). The 50-mm at f/5.6 did the trick nicely, with the most useable images taken at ISO 320 for 6 to 8 seconds. I used the 8-s exposures for the analysis, but in retrospect should have used the 6-s ones.
Before completing the analysis, however, I thought I'd lost my mind. The numbers jsut weren't adding up, neither were they multiplying or dividing, for that matter. The last I'd heard, the nova was V=4.4, so I was expecting it to be fainter today. After all, most novae drop quite smoothly in brightness after maximum (a recent study shows that 38% of known novae have smooth light curves - defined as following a power law decline with no major fluctuations).
Something was odd - was I using incorrect comparison stars? Is my camera broken? Have I been abducted?
The truth is far more exciting - the nova has brightened over the past few days!
After peeking at the latest AAVSO light curve (which confirmed it was brightening thus restoring my previous levels of sanity) I eagerly estimated its brightness. Two frames were used for analysis: the first (#8209) included stars westward of the nova (e.g. tau Cen, gamma Cen) while the second (#8212) included stars eastward of the nova (e.g. eps TrA, beta Cir). The RAWs were converted to TIFFs and dropped into Photoshop. Comparison stars were selected that were, by visual inspection, slightly brighter or slightly fainter than the nova. In this way, 11 suitable stars were identified (see diagram 1), and their V magnitudes taken from SIMBAD.
Each star was then, in turn, placed in a 6x6 selection region such that the "Mean" Luminosity value reported by Photoshop's "Histogram" was a maximum (see diagram 2).
These mean pixel values were then correlated with the V magnitudes; a linear fit gave correlation coefficients > 0.99. This procedure was repeated using a 8x8 selection region.
Using these correlations, the nova's magnitude turns out to be V = 3.6 +/- 0.2 [2013 December 10 at 23:25 ].
Carol Botha imaged the nova this morning (Monday, December 09). The little rectangle shows a zoomed-in view, which includes beta Cen at lower right. More of Carol's images will be found in her "Nova Cen 2013" Facebook album.
By my estimate, comparing it to HR 5089 (HD 117440, V=3.88) and HR 5241 (HD 121474, V=4.71), I make the nova as magnitude 4.4.
This morning, the nova was still a naked-eye spot just to the left of beta Centauri, despite the terrible light pollution (and the onset of day).
From an 8-second photo of the region (50-mm, f/5.6), the nova is slightly fainter than Upsilon-1 Cen (HR 5249, HD 121790, V=3.87) but brighter than Upsilon-2 Cen (HR 5260, HD 122223, V=4.34). Using Iris's aperture photometry routine, the nova has V=4.1 +/- 0.2.
The brightness of a nova can be used to gauge approximate distance. The distribution of their absolute magnitudes shows twin peaks, one at -8.8 and a second at -7.5. Buscombe and de Vaucouleurs showed in 1955 that the mean magnitude of novae 15 days after maximum light was a constant; more recent work (1995) shows that the absolute magnitude 15 days post-max is -5.60 ± 0.14. There are, however, exceptional objects that suggest this method be used with caution.
Let's do some back-of-the-envelope calculations. Assume the V magnitude at maximium is 3.7, guessed from the AAVSO light curve.
The relationship between absolute (M) and apparent (m) magnitude depends on the distance (d, in parsecs):
M = m + 5 - 5 (log(d))
for M = -8.8, m = +3.7: d = 3162 parsecs.
for M = -7.5, m = +3.7: d = 1738 parsecs.
So, with enthusiastic hand-waving, we can say the nova lies between 5,700 and 10,000 light years from Earth.
nothing more to see. please move along.