Astron. Astrophys. 318, 908-924 (1997)

2. Review of distance determinations of V 1974 Cygni

In our previous paper (Chochol et al., 1993) we tried to determine the distance of the nova by measuring the equivalent widths of the Ca II H and K interstellar lines. This procedure led to an unplausibly low value of d = 0.6 kpc. Similarly, Annuk et al. (1993) using the same method obtained the distance of d = (1.4 0.1) kpc. The obvious disadvantage of this procedure is its sensitivity to contamination by the circumstellar lines of Ca II and H I. Thus we prefer to reject the data as unreliable.

The most widespread distance determination method for novae is based on the estimate of absolute magnitude at maximum light (e.g. by the MMRD - relation) and on the evaluation of interstellar extinction. Both parameters are, of course, sensitive to various sources of errors. However, there are several well established techniques to estimate these errors and therefore proper analysis may lead to reliable determination of the nova distance.

Most authors estimated the absolute magnitude of the nova in maximum by employing various statistical relations between absolute magnitude in a given colour of a broadband colour system and the time of the decay of the light curve by two (t₂ ) or three (t₃ ) magnitudes below maximum light. A review of published data is given in Table 1.

Table 1. Review of the colour excesses and distance estimations from absolute magnitudes at maximum light of Nova V 1974 Cyg

As it is easily seen from Table 1, the values of absolute magnitude and distance of Nova V 1974 Cyg published by various authors are in reasonable mutual agreement. There are, however, exceptions, namely the results of Shore et al. (1994), Rafanelli et al. (1995) and Della Valle & Livio (1995). They all derived the absolute magnitude using the evidence for novae in near neighbour galaxies (LMC; M 31). Della Valle & Livio (1995) derived an absolutely calibrated MMRD relation:

and calculated the absolute magnitude of Nova V 1974 Cyg at maximum as = -8.3. The relation can undoubtedly serve as the best tool for the estimation of the absolute magnitudes of novae at maximum. However, the weak point of the relation is that it connects absolute visual magnitude with the rate of decline in V colour ( = 0.125 mag/day) for the decay time only. It is easy to calculate using the data in Table 1, that the daily rate of decline (given by t₂ /2 and t₃ /3) significantly depends on decay times as well as on the colour. The daily rate of decline calculated from t time is = 0.071 mag/day. If we wish to employ the more reliable time (instead of t₂ time) in relation (1), we have to transform t₃ to t₂ time. Capacciolli et al. (1990) published the following transformation formula. If t₃ 80 days (this is certainly true for Nova V 1974 Cygni), then

By applying this relation we easily find = (23.9 2.1) days. The corresponding = (-7.72 0.13) is then in full agreement with other calculations of .

The low extinction coefficients in the direction of Nova V 1974 Cyg were determined rather artificially - we believe that the authors were a priori influenced by the distance determination of the nova from the first HST imaging (see Paresce, 1994). Nova V 1974 Cyg is located near the galactic plane (b = 7 8); so the assumption of = 0 introduced by Shara (1994) and accepted by Della Valle & Livio (1995) is quite absurd. Mathis et al. (1995) critically evaluated the reddenings published by different authors and concluded that the possible absorptions are within the interval 0.59 - 1.0 (reddenings E(B-V) being between 0.19 and 0.32). As was pointed out by Paresce et al. (1995), the spectrum of the nova shows a prominent broad absorption feature centered at 220 nm caused by the interstellar extinction. They removed the feature assuming a value for E(B-V) = 0.35.

Although the accurate determination of the interstellar extinction as well as of the absolute magnitudes of novae is difficult, we conclude that the true distance of the nova can hardly be larger than 2.2 kpc unless one assumes unrealistically low interstellar extinction in the given direction and/or an extremely bright absolute magnitude of the nova.

However, advances in observational techniques gave us the opportunity of true distance determination based on nebular expansion parallaxes. When the angular diameter (and shape) of the expanding envelope is resolved and if we knew the true projected expansion velocity of the envelope from high-dispersion spectra, the determination of the distance of the nova is almost trivial.

In the meantime several measurements of the diameter of the nova envelope have been published and the distance of the nova was determined under specific (sometimes possibly poor) values for the average expansion velocity (see Table 2.)

Table 2. Angular radius of the nova envelope and related distance of the nova

Contrary to expectations, the spread of distances is large and the a priori most reliable distance determination through the optical imaging of the envelope by the FOC camera of the HST lies well above the limit set by previous indirect methods. As noted by Paresce (1994), the possible weakness of the geometric method lies in the uncertain selection of the "proper" expansion velocity from the early spectra of the nova. We believe that this is a crucial problem, affecting profoundly the distance determination in all similar cases, for novae and possibly also for supernova remnants.

It is well-known from the classical work of McLaughlin (1943) that the velocity development in the spectra after outburst is complex, and several systems with widely different velocities develop in the course of time. The ratio of different velocities easily exceeds 2:1, and thus it is not easy to predict correctly which velocity values correspond to the observed angular size of the nova envelope. Moreover, there is ample evidence that the velocities of given segments of the envelope change with time due to the changes of driving force and/or the interaction of the envelope with the circumstellar medium.

Wade (1990) found that most of 26 resolved nebular remnants of classical novae are prolate in outline, with substructures that can be characterized as consisting of polar blobs and equatorial rings. Boyarchuk and Gershberg (1977) found these substructures spectroscopically in the expanding envelope of Nova V 1500 Cyg.

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de