I was wondering if there's an explanation for the apparent unusual V-Rc and Rc-Ic 'colors' in the AAVSO Database for this nova ? I've only looked at multi-color observations in our DB going back to September 2014 which evidence the unusual (for me) negative V-R values of between -0.1 and -0.3. The R-I colors are even more negatively expressed (from about -0.7 to -1.0).
Wanting to 'see' this for myself, I observed this nova several nights ago in V,Rc,Ic with the resultant V-R near negative 0.6 while the Rc-Ic indice was near negative 0.9.
The B-V color (perhaps) is more unremarkable, near +1.0 as per recent BV submissions to our database. It would also be interesting to see what the NIR colors (ie: redward of Ic) are like.
My recent V obs near 12.8 on June 25.067 2015 (UT) using 122, 129, 148, and 152 comps as the ensemble (error near 0.05) is several tenths brighter than other recent V obs by other AAVSO observers however several visual observers have submitted data closer in time to my own that also suggest a minor brightening around June 25.
Most of the stars I look at have brighter mags as you go through the B-V-R-I bandpasses. V339 Del seems an interesting breed apart.
This late in its
This late in its evolution, the nova is in its "nebular" phase. Almost all the light reaching us is from the ionised shell of ejected gas that surrounds the binary pair. You're not actually measuring the star(s), but the gas it throws off. Early on, this gas is a fireball, and creates a spectrum that approximates that of a regular star. But as this gas expands and becomes optically thin, the spectrum of the nova starts closely to resemble that of a planetary nebula.
By far the strongest contribution to the overall flux comes from the [OIII] emission line at 5007 Ångstrom. In a recent spectrum I took of Nova Cen 2013 (about four months younger than V339 Del) the [OIII] line measured 400 times stronger than the continuum, and twenty times stronger than the next nearest line, H alpha at 6563 Å.
So, in short, the colour indices you are getting are accurate, because - unlike stars - the nova is a bright blue-green colour :-) For the same reason, any photometry transforms based on a linear colour index will be way off. If it were me, I'd suggest submitting untransformed observations of late novae, but I'm sure Arne or some others will have the best advice. It was quite interesting watching my BVRI observations of Nova Cen a year ago. They had been bang on with those of BSM South, but then started to diverge. I was quite worried until I realised what was happening.
I have attached an image of the spectra I took last year. It hasn't changed much since then. It shows how most of the light now emitted in very narrow bands. The green [OIII] is the brightest with Ha the next brightest. This makes the colour indices behave as you have noticed.
Thankyou Jonathan and Terry for your very informative explanations.
I suppose a gradual normalization of colors might be expected as the immediate nova environment settles (relatively). With the symbiotic nova V745 Sco last year, I don't think the negative color indices were noted - admittedly the outburst environments of NC types are, I imagine, quite different from those of so-called classical novae.
Jonathan, you did mention your nebular-phase observations with N Cen 2013. Have you (Terry, this is posed for yourself as well) recorded these types of transient spectral data for other novae as well ? For example, V1280 Sco (NB:2007) is still quite bright, even though we're 8 years on now. Although I expect the multicolor coverage might not have been as forthcoming as it was for N Cen 2013 and N Del 2013 I will look for evidence of nebular-phase\negative indices in our database for this 2007 event.
Lastly, I suppose that by the time we get to our DQ Her 1934 or DI Lac 1910 environments, colors have pretty much settled until ....
Again, thankyou for your contributions to this topic.
Johnathan is right - once a nova enters the nebular stage, transformation makes little sense. Untransformed data, especially where you can provide either a wavelength response curve for your filter or provide a link to a vendor's curve, along with your camera make/model, is probably the way to go. Wide-band photometry with the Johnson/Cousins filters works wonders in many circumstances, but dealing with pathological objects like emission-line stars and nebular novae, is best done with spectra.