Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 364, 517-531 (2000)

Previous Section Next Section Title Page Table of Contents

5. Discussion and conclusions

With the main goal to investigate systematic differences between early-type and late-type galaxies - as for colours, redshift distributions, and ages of the dominant stellar populations - we have analyzed a morphologically-selected complete sample of 52 spiral and irregular galaxies in the Hubble Deep Field North with total K-magnitudes brighter than K=20.47 and typical redshifts from [FORMULA] to 1.5. The sample makes use of total photometry in the UBVI bands from HST and the JHK bands from ground, all carefully tested with an extensive set of Monte Carlo simulations.

The present sample exploits in particular the ultimate imaging quality achieved by HST in this field, allowing us to disentangle among galaxy morphologies, based on accurate profiles of the surface brightness distributions.

Our analysis makes also use of an exhaustive set of modellistic spectra accounting for a variety of physical and geometrical situations for the stellar populations, the dusty ISM, and relative assemblies. The high photometric quality and wide spectral coverage allowed us to estimate accurate photometric redshifts for 16 objects lacking a spectroscopic measurement.

We have also carefully evaluated all plausible systematic effects of the selection, in particular the redshift cutoff implied by the limiting surface-brightness achievable in the reference K band image.

A warning is in order, in any case, about the general conclusions derived from our sample of K-selected galaxies: they should be treated with caution, due to the very small field of view and modest spatial sampling of the present survey. Ferguson et al. (2000) and Eisenhardt et al. (2000) estimate that the number of [FORMULA] galaxies in the total HDF co-moving volume between z=1 and z=2 is only a few dozens. Considering also the strong clustering inferred for Lyman break galaxies (Adelberger et al. 1998), statistical fluctuations imply large uncertainties on any conclusions based on samples like the HDF, until more substantial surveys to similar depths will be made available.

Three the main results of our study.

  • The sample galaxies are distributed in redshift up to [FORMULA], but appear to be significantly missing above, compared with evolutionary models assuming standard recipes for the luminosity evolution and a substantial redshift of formation. We reported a similar finding in our previous study of early-type galaxies in the same area (Franceschini et al. 1998). Our conclusion is that, either the area has some peculiarities, or the underlying mass function for galaxies of all morphological kinds has a global decline at these high redshifts. Confirmation of this result will require more substantial sky areas to be surveyed to similar depths by large telescopes.

  • Differences between early- and late-types are apparent in the rest-frame colour distributions and the evolutionary star-formation rates per unit volume. In particular, the short-wavelength rest-frame colours B-J, once dust reddening is taken into account, appear quite significantly bluer for late- than for early- galaxy types. On the contrary, the longer-wavelength colors V-K appear to be very similar for the two morphological classes. We interpret this as an indication that, while the stellar mix in spirals/irregulars includes young newly formed populations, less apparent in E/S0, the underlying older age component traced by the V-K colors has quite a similar origin and age distribution for the two galaxy categories.

    We warn, however, that the complication in spectro-photometric modelling introduced by dust-extinction in the gas-rich systems prevents to reach conclusive results on the source by source basis. Only future long-wavelength IR observations, from space (SIRTF, FIRST, NGST) and from ground (10 m class telescopes in the mid-IR and interferometers in the sub-mm), will allow to break down the age/extinction degeneracy.

  • We found that an integrated quantity like the comoving-volume star-formation rate density as a function of redshift [FORMULA] is much less affected by the uncertainties related to the dust distribution. The reason for this is mostly in the fact that our analysis is strongly constrained by the evolutionary baryonic mass function in stars traced by the near-IR galaxy luminosities, the estimate of the baryon mass at any redshifts being much more robust than that of the instantaneous rate of star formation (see Sect. 4.5).

    By combining this with the early-type galaxy sample previously studied by FA98, we find a shallower dependence of [FORMULA] on z between [FORMULA] and [FORMULA] than found by Lilly et al. (1996), i.e. [FORMULA] rathen than [FORMULA] as in Lilly et al. (1996). In this redshift interval our observed [FORMULA] turns out to roughly agree with results published by Cowie et al. (1999) and Treyer et al. (1998).

    Our present results, based for the first time on a careful modelling of the whole UV-optical-NIR Spectral Energy Distributions of galaxies, then support a revision of the Lilly-Madau plot at low-redshifts for UV- and K-band selected galaxies. UV-selected and near infrared selected galaxy samples display a remarkably similar evolution of the comoving SFR density [FORMULA], at [FORMULA].

The three above findings seem to favour the general scheme of hierarchical assembly for the formation of bright galaxies, envisaging their progressive build up during a substantial fraction of the Hubble time. After all, this is the most physically motivated present description.

In this context, a warning is in order concerning some published specializations of the Cold Dark Matter cosmogonic scheme, predicting that spheroidal galaxies in the field form at low redshift ([FORMULA]) from merging of spirals (Kauffmann & Charlot 1998). This prediction is not supported by our results in Fig. 12, where Ellipticals and S0s appear to have been mostly formed at [FORMULA], whereas spirals/irregulars keep some sustained SF activity at [FORMULA]. This result suggests that merging of spirals to form ellipticals at low redshifts cannot be a dominant process, quite in agreement with what found by Brinchmann & Ellis (2000).

It is conceivable, however, that minor modifications of the CDM hierarchical scheme (e.g. in terms of different assumptions about the cosmological parameters [FORMULA], [FORMULA]) can explain the observed differences between morphological types. In our view, these differences in the population histories are quite probably related to the presence of different environments at different densities (and consequently different cosmic timescales of formation) in what we call the "field". In particular, moderately high-density environments (typically galaxy groups, as found very numerous in the spectroscopic survey by Cohen et al. 1999), with an accelerated cosmic timescale of evolution and fast gas consumption, mix with truly low-density environments, where the transformation of primordial gas into stars slowly progresses during the whole Hubble time. We believe that the two galaxy morphologies analyzed in the present paper and in FA98 trace such different environments in the universe.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: January 29, 2001