N. R. Tanvir
Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge, CB3 0HA, United Kingdom.
Keywords: Hubble constant, Cepheids, Leo-I group
The long-running debate over the value of the Hubble constant has centered largely on the calibration and use of secondary and tertiary distance indicators. Such indicators are required to extend the distance ladder to a regime where peculiar velocities are small compared to recession velocities. By greatly increasing the range of Cepheid distance determination, HST is beginning to provide much more secure calibration of the secondary indicators.
The Leo-I group is the nearest group containing a mix of early- and late-type galaxies and is, therefore, a good location to set about the calibration of the various early-type galaxy distance indicators. It is relatively compact, has a small range in velocity and is at high galactic latitude. Most importantly, it contains a unique ring of intergalactic HI gas which is orbiting the central two E/S0 galaxies (NGC3379 and NGC3384) and exhibits a tidal tail apparently due to the passage of the spiral galaxy M96 (Schneider 1989). This fortuitous circumstance provides good evidence that M96 is indeed near the center of the group, and gives confidence that its distance can be taken to apply to the group as a whole.
To date, three Leo-I spirals have been studied for Cepheids. The result for M96 itself gave a distance modulus of mag (Tanvir et al. 1995; hereafter TSFR). In passing we note that this is somewhat higher than the 30.07 mag and 29.90 mag inferred from Ciardullo, Jacoby & Tonry (1993; hereafter CJT) for the PNLF and SBF methods, respectively, when calibrated consistently in the M31 system, but is in excellent agreement with the new SBF distance to NGC3379 of mag determined by Sodemann & Thomsen (1996). The other two spirals being studied for Cepheids are UGC5889, a small galaxy close to NGC3377 (Tanvir et al. 1996, in prep.) and M95 which is being observed as part of the HST distance scale key-project (Freedman, this volume, p. ). Between them, these galaxies should provide an extremely good distance to the group, with some indication of the depth of the late-type members.
Figure: This flow diagram shows all the steps used by TSFR in deriving . Each box represents a difference in distance modulus between particular objects, shown as ellipses. The two routes from Leo-I to Coma rely mainly on various early-type galaxy distance indicators, namely the surface brightness fluctuation (SBF), planetary nebula luminosity function (PNLF), globular cluster luminosity function (GCLF), color-magnitude relation (CMR) and methods. Although the formal error on the weighted mean in the Leo-I to Virgo step is only 2% TSFR allowed a larger uncertainty to account for possible systematic effects. such as those described in section 2.2. The Coma cluster is assumed to be sufficiently remote that the corrections to its velocity for peculiar motions are small. The reader is referred to TSFR for bibliographic details of each indicator.
Stepping from Leo-I to Coma, TSFR infer a value for of (see figure 1). A detailed discussion of the Cepheid photometry is the subject another paper (Tanvir et al. 1996, in prep.). Here we consider some of the issues concerning the secondary indicators which can be calibrated in the Leo-I group.
The peculiar velocity of the Leo-I group is difficult to determine, so to evaluate we must use secondary distance indicators to step to more remote clusters. Various secondary indicators can be calibrated in the Leo-I group. TSFR considered five secondary indicators based on early-type galaxies, as is illustrated in figure 1. The argument for using only the early-type galaxies as secondary calibrators is that to use the Leo-I spirals, other than M96, requires us to face again the difficulty of assigning group membership to the late-type galaxies. Early-type galaxies being more highly clustered suffer much less from background and foreground contamination, and in the particular case of the Leo-I group there is independent evidence placing at least NGC3379, NGC3377 and NGC3384 at a common distance (Ciardullo, Jacoby, & Ford 1989, Tonry, Ahjar, & Luppino 1990). Of course, M96 is itself another calibrator of the Tully-Fisher relations. Below we consider in more detail some of the issues concerning type Ia supernovae, which TSFR did not use, and the surface brightness fluctuation (SBF) and planetary nebula luminosity function (PNLF) methods, which they did.
Normal type Ia supernovae have been observed in two galaxies in the Leo-I region, namely SN1967C in NGC3389 and SN1989B in NGC3627. The estimated peak, extinction-corrected V-band magnitudes of these supernovae are very different being 13.3 and 10.84, respectively, (Leibundgut et al. 1991, Wells et al. 1994). Even considering the large, and therefore uncertain, extinction correction in the case of SN1989B, of , the difference between peak magnitudes alone shows that both galaxies are not at the same distance. In fact, while both have been classed as group members in some previous surveys of the region, neither galaxy is in the set of ``high-confidence'' Leo-I members used by TSFR, as listed in Tanvir (1996). Let us consider the evidence for and against the group membership of each galaxy:
This galaxy is located very close on the sky to the center of the Leo-I group, forming a tight triplet with the central two E/S0 galaxies, NGC3379 and NGC3384. Its heliocentric velocity is v=1270 compared with for the average of NGC3379 and NGC3384 (data from de Vaucouleurs et al. 1991). Some catalogues (e.g., Huchra & Geller 1982) have included NGC3389 as a Leo-I group member but latterly it is generally thought to be in the background (e.g., Garcia 1993). In fact, if NGC3389 were a group member then SN1967C would have to have been underluminous by mag for any reasonable SNe Ia calibration.
This galaxy forms a well known triplet, with NGC3623 and NGC3628, some on the sky from the center of the Leo-I group. Tully (1987) distinguishes the two groups, calling them the `M66 group' and `M96 cluster', respectively, and suggests they form part of a larger, more amorphous structure dubbed the Leo Spur. Branch, Romanishin & Baron (preprint) have noted that if NGC3627 is assumed to be at the TSFR Leo-I distance, then the magnitude of SN1989B agrees very well with the current Sandage et al. \ SNe Ia calibration (e.g., Sandage 1995, Tammann, this volume, p. ). However, using the Tully-Fisher relation distances given by Bottinelli et al. (1984) we find the M66 group to be % closer than the Leo-I/M96 group, although the uncertainties are such that the possibility of both groups being at the same distance is not ruled out. For the present, then, we suggest that SN1989B only be used as a SNe Ia calibrator with some degree of caution.
These two methods, suited primarily to application in early-type (gE/S0) galaxies, are claimed to be of very high precision, and hence are highly weighted in the TSFR determination. The PNLF method is based on measuring the bright end cut off in the planetary nebula luminosity function, which is assumed to be universal. The SBF method relies on the measurement of fluctuations in brightness across the face of elliptical galaxies which are due to the counting statistics of individual stars in each resolution element. The evidence for high precision is based on the good internal agreement of each indicator for galaxies within groups and clusters (Jacoby, Ciardullo & Ford 1990, Tonry 1991). Both methods were also claimed to agree well with each other, when calibrated in the M31/M32 system (CJT), but more recent work places this calibration in doubt (Méndez et al. 1993, Sodemann & Thomsen 1996).
However, Bottinelli et al. (1991) have pointed out that there is even a small but significant difference between the two indicators in their measurement of the distance ratio between the Leo-I group and the Virgo cluster. They suggest that this reflects a dependence of PNLF distances on parent galaxy luminosity, which, if accounted for empirically, changes the relative Leo-I to Virgo distance to . Bottinelli et al. argue that such an effect could be produced by a high luminosity tail to the PNLF, however, CJT counter that this should actually produce a dependence on surveyed luminosity rather than total luminosity of the parent galaxy, something which is not seen.
In fact, because the PNLF galaxies in Virgo all have brighter absolute magnitudes than those in the Leo-I sample, there is an equally good correlation between and SBF distance. Such a bias could be the result of problems in determining the completeness level for the very faint PNe in Virgo. Both possibilities are illustrated in figure 2. We note that any problem is rather less likely to be with the SBF distances, since the SBF method has been shown to agree fairly well with other indicators out to higher redshifts (Jacoby et al. 1992). It is worrisome that such a discrepancy should arise in galaxies of the same type (i.e., giant E/S0) and in the same distance regime, however, we note that both correlations are reduced somewhat if the three Fornax cluster galaxies (CJT) are added. Clearly further observations are necessary to resolve this conflict.
Figure: Comparison of SBF and PNLF distances to individual E/S0 galaxies in the Leo-I group (closed symbols) and the Virgo cluster (open symbols) taken from results compiled by Ciardullo, Jacoby & Tonry (1993). The PNLF method finds the Leo-I to Virgo distance to be systematically less than the SBF method. The discrepancy is significant at the 97% level given the quoted errors. Empirical correlations are seen with (a) parent galaxy magnitude (assuming SBF distances), as previously noticed by Bottinelli et al. (1991), and (b) SBF distance modulus. Distinguishing which of these effects, if either, is fundamental requires more data.
We have argued that the Leo-I group offers a particularly good stepping stone to , by allowing calibration of the E/S0 galaxy secondary distance indicators. In general, the E/S0 indicators agree well with each other and benefit from the predominance of early-type galaxies in cluster cores. TSFR have used the HST Cepheid distance to M96 in the Leo-I group to obtain . Further, HST observations of Cepheids in another two Leo-I galaxies will refine this estimate further.
There are some outstanding problems with the SBF/PNLF indicators, in that they appear to give small but significant differences in the Leo-I to Virgo relative distance. This discrepancy is important for the Leo-I group and also the wider distance scale controversy, but should be resolved by further observations. Finally, we urge caution in using SN1989B as a calibrator for the SNe Ia, given the line of sight uncertainty in the position of its parent galaxy NGC3627 relative to the core of the Leo-I group.
Bottinelli, L., Gouguenheim, L., Paturel, G., & de Vaucouleurs, G. 1984, A&AS, 56, 381
Bottinelli, L., Gouguenheim, L., Paturel, G., & Teerikorpi, P. 1991, A&A, 252, 550
Branch, D., Romanishin, W., & Baron, E. preprint
Ciardullo, R., Jacoby, G.H., & Ford, H.C. 1989, ApJ, 344, 715
Ciardullo, R., Jacoby, G.H., & Tonry, J.L. 1993, ApJ, 419, 479 (CJT)
Garcia, A.M. 1993, A&AS, 100, 47
Huchra, J.P. & Geller, M.J. 1982, ApJ, 257, 423
Jacoby, G.H., Ciardullo, R., & Ford, H.C. 1990, ApJ, 356, 332
Jacoby, G.H. et al. 1992, PASP, 104, 599
Leibundgut, B. et al. 1991, A&AS, 89, 537
Méndez, R.H. Kudritzki, R.P., Ciardullo, R., & Jacoby, G.H. 1993, A&A., 275, 534
Sandage, A. 1996, in Particles, Strings and Cosmology, WSP (Singapore), in press
Schneider, S.E. 1989, ApJ, 343, 94
Sodemann, M. & Thomsen, B. 1996, AJ, 111, 208
Tanvir, N.R., Shanks, T., Ferguson, H.C., & Robinson, D.R.T. 1995, Nature, 377, 27 (TSFR)
Tanvir, N.R. 1996, in Clusters, Lensing and the Future of the Universe, eds. Trimble, V.& Reisenegger, A., ASP conf. ser., in press
Tonry, J.L, Ahjar, E.A., & Luppino, G.A. 1990, AJ, 100, 1416
Tonry, J.L. 1991, ApJ, 373, L1
Tully, R.B. 1987, ApJ, 321, 280
de Vaucouleurs, G. et al. 1991, Third Reference Catalogue of Bright Galaxies, Springer-Verlag (New York)