5. The spectral energy distribution of P1724, and its location in the H-R diagram
In order to investigate the spectral energy distribution (SED) of P1724 and derive its bolometric luminosity, we have made use of all available UBVRIJHKL photometry from the literature (Walker 1969, Penston et al. 1975, Warren & Hesser 1977, Breger et al. 1981, Preibisch et al. 1995, and Cutispoto et al. 1996). Our own BVRI observations obtained at CTIO are also useful (Table 2), but the MTSB and RW measurements cannot be reliably placed on the standard system. In addition, we obtained new data at JHK to supplement the previous observations.
Our new observations in the JHK bands were obtained on 20 Feb 1997 at the CTIO 1.5m using the CIRIM array, and are listed also in Table 2. They were obtained in a raster pattern, with 15 arc sec offsets. Three frames were exposed and co-added at each raster position. The total integration time on P1724 was 7.2 seconds through each of the K, H, and J filters. The data were flat-fielded using dome-flats, co-added, median-filtered, sky-subtracted, then co-aligned and summed. Five or six standard stars (Elias 1981) were observed nightly. Uncertainties in the magnitudes, based on the scatter in the standard star solutions, are mag. All CIRIM data reductions were performed using IDL.
Because the observations collected in Table 2 are not simultaneous, they are significantly affected by the photometric variations in P1724 due to rotational modulation by spots, which is much larger than the precision of these measurements. In addition, there are probably also long-term variations as discussed in the previous section. Further evidence of these changes is given by the range of spectral types reported for the star over the years: G8 (Herbig & Bell 1988 and Tagliaferri et al. 1994), K0 (Van Altena et al. 1988), K2 (Terranegra et al. 1994), and K3 (Preibisch et al. 1995). Variations in the spectral type of this order are typical for spotted stars (e.g., Bouvier et al. 1993).
As a compromise, we have adopted the average between the maximum and minimum values in each band for the determination of the SED and the calculation of the bolometric luminosity. This SED is thus an average one. We compared the observed SED of P1724 with those of standard stars with spectral types in the range from mid-G to mid-K. We find the best match for a spectral type of K0 (see Fig. 6).
The SED in Fig. 6, corrected for extinction, shows that P1724 has neither significant UV nor significant infrared (IR) emission in excess of the standard star. Hence, we may classify it as a naked TTS. Because the equivalent width (0.7 to 2Å; cf. Sect. 7) and flux are very weak, it may also be considered a wTTS. 3
The extinction and bolometric luminosity were estimated as described in detail by Alcalá et al. (1997). We obtain , an extinction of mag 4, and from these a radius of derived from the Stefan-Boltzmann law. If we use instead the Barnes-Evans relation (Barnes & Evans 1976), we obtain a radius of , again adopting as the distance. The IR photometry is intrinsically less precise than the optical data, but at the same time less sensitive to the effect of spots. We repeated the luminosity and radius calculations using the optical and IR bands separately, and obtained internally consistent results as when using all the photometry together. This indicates that the results are quite robust. For the remainder of the paper we adopt for the radius.
By comparing the locus of P1724 in the H-R diagram with theoretical tracks and isochrones, we can estimate its age and mass. From both the Forestini (1994) and the D'Antona & Mazzitelli (1994) models, we derive an age of years and a mass 5 of . P1724 appears to be slightly more massive in the Palla & Stahler (1993) model for intermediate-mass Herbig Ae/Be stars. Hence, as far as the mass is concerned, P1724 appears to be a star near the borderline between TTS and Herbig Ae/Be stars. P1724 is clearly a pre-MS star; post-MS stars with spectral type G0 or later show much weaker lithium than P1724 and/or rotate slower than (Gray 1989).
The mass and radius of P1724 from the SED yield a surface gravity () in excellent agreement with our spectroscopic gravity determination (, Sect. 3). The surface gravity obtained from Strömgren photometry by Terranegra et al. (1994) yields according to Tagliaferri et al. (1994), also consistent with our results within the errors. However, activity in general is known to affect Strömgren photometry (Morale et al. 1996, Alcalá et al. 1998). Our high-resolution spectroscopy yields a gravity independent of the distance (Sect. 3), which is consistent with the above distance-dependent measurements. This can be taken as an additional indication that P1724 is most likely located at the distance of the Orion nebula.
© European Southern Observatory (ESO) 1998
Online publication: June 2, 1998