Nova Sco 1992
Nova Sco 1992 (V992 Sco) was discovered on May 22 (TJD 8753) by Paul Camilleri (IAUC 5526) at a visual magnitude of 8.2. It soon reached a maximum of V = 7.26 on May 26 (TJD 8768). The following decline was slow but peculiar, with a series of peaks and dips, some having delta-m > 1 mag as well as smaller changes (Figure 1). The progenitor appeared to be a star of V = 18. Early optical spectra (IAUC 5529) showed the usual forest of lines, primarily H Balmer and Fe II. H-alpha was in emission, with a weak P Cygni profile at 700 km/s. On May 26, H, H, and H were in absorption, but by May 29, these had become emission lines. Strong O I and Ca II triplet emission showing P Cyg profiles were also observed on May 28 (IAUC 5553). Other emission lines also showed FWHM = 700 km/s.
The early stages of the nova were thus indicative of strong but relatively slow outflow.
We began obtaining infrared observations of the nova approximately a month after outburst, on June 20 (TJD 8793). The JHKL'M photometry and spectra were taken at the Siding Springs 2.3 meter observatory using the Infrared Photometry System (IRPS) and the Cooled Infrared Grating Spectrometer (CIGS). The spectra were obtained on June 21, while the photometric data were obtained on June 20, 21, 22, 23, 24; July 11, 13, 17; August 12; and September 11. Conditions were photometric and drier than normal.
We calculated the interstellar reddening to the nova based on the B-V colors at maximum and assuming that the bolometric correction is 0 (generally true for a nova in the early stages). From this, we calculated E(B-V) = 0.60, which implies (assuming that R = 3.1) Av = 1.85. We used this value to correct that data used in the modeling of the spectral energy distributions (SEDs) discussed below.
The Infrared Photometry
The full set of (observed) infrared photometry and colors is shown in Figure 2 with photoelectric V magnitudes from the IAU circulars for comparison.
The infrared colors up to TJD 8820 are roughly constant, though both K and V are undergoing large changes. Therefore the large variations in V in the early part of the nova's evolution cannot be tied to dust formation episodes. The optical colors published in the IAU circulars also showed little variation in these early stages, up until TJD 8840. The variations could be due to either density variations in the ejecta or further minor outbursts from the nova.
After TJD 8820, there is a slow reddening of the IR colors and increase in the K magnitude (as V correspondingly decreases). At this stage we are seeing the formation of dust and the gradual thickening of the circumnova dust shell. In particular, the very red colors in August and September (TJD 8846 and 8876) are indicative of a very thick dust shell.
With this in mind, we modeled 3 dates for which we have infrared photometry, and for which there is good UBVRI photometry (all from A.C. Gilmore).
JD 2448820: These data were impossible to fit with any model. We present the best one here. The central source temperature had increased to 8000 K, and the optical depth had increased to tau = 0.9, though the grain size was till 0.2 microns. The shell radius for this model had decreased to delta = 0.25. However, there is clearly a near-IR excess. It was not possible to fit all the infrared data with any one model. We must conclude that the structure is more complex than a single shell. Since it was already apparent from the lightcurve and optical spectra that the nova was ejecting more material past the initial outburst, this is no surprise.
JD 2448876: These data were obtained well into the dust formation time of the nova outburst. The colors are very red, with the flux still rising at M. The model still used a central source temperature of 8000 K and the grain size was still 0.2 microns. The optical depth (obviously) was much larger: tau = 2.5. The shell thickness had also increased to delta = 0.95. The R magnitude in this, as in previous observations, was affected by the strong H-alpha emission.
In general, we see from the models that Nova Sco 1992 had a long, gradually increasing dust formation episode.
The Spectra
We obtained a set of J, H, K, L', and M spectra on June 21 (TJD 8794), during the early stages of the evolution (and 3 days before the first model). These are presented in Figures 3 through 7.
The most curious thing about these spectra is the apparent "reverse" P Cyg profiles on the hydrogen lines. This appears in all the strong H emission lines. Many of these lines also show a double-peaked profile, which is common among novae. It is difficult to understand the reverse P Cyg profile: if it is due to underlying absorption, one would expect some sign of it on the blue side, as well. Reverse P Cyg profiles have been observed in other objects such as the peculiar B supergiant HD199478(3), the Be star FY CMa (IAUC 4391), the interacting binary W Ser(4), and the M star HD 139216(5). If the V variations were due to pulsations, then our spectra may have been taken during an "infall" period. We will be analyzing these lines in detail, in order to try to fit the profiles.
Besides the perplexing H I lines, we see emission from of He I, C I, O I, and N I. There may be lines of He II as well but the ID is less certain.
We also see emission lines of CO, particularly a strong line at 4.65 microns. We see additional lines at 1.58, 2.293, 2.374, and 2.385 microns. Other lines may be blended with the plethora of other emission lines. The 4.65 microns line in particular affects the SED we modeled on June 24 (TJD 8797).
The appearance of fundamental CO emission lines usually portends the formation of dust. We of course see this occur later in this nova's evolution.
Nova Sco 1992 is an unusual dusty nova, with an odd spectrum not generally seen in these objects.
1. Computer Physics Communications 48, 271
2. MNRAS, 277, 259
3. Markova and Valchev, 2000, A&A, 363, 995
4. Strupat et al., 1998, ESO Msng, March, p. 40
5. Sato et al., 1978, PASJ, 30, 557