All Figures

Figure 1: Bright cluster galaxies: top row, left: photometric redshift vs. apparent magnitude for the galaxy sample in the A901 field with the cluster at . Top row, right: redshift vs. absolute magnitude. The lines show the selection for the cluster sample used in the following analysis. Bottom four panels: redshift histograms in apparent magnitude bins. The dashed lines show the redshift interval for the adopted cluster sample. At R>21 the errors in photometric redshift start to broaden the cluster distribution and bias it to lower redshift. A rough estimate for the field contribution is shown in grey.

Open with DEXTER


In the text

	Figure 2: Rest-frame colour-magnitude diagrams: cluster ( left) versus field ( right) with cluster red-sequence fit (grey line) and Butcher-Oemler style red-sequence cut (black line).
Open with DEXTER


In the text

	Figure 3: Selection of field galaxy sample: luminosity-redshift diagrams of red-sequence galaxies show rich groups and clusters in all three COMBO-17 fields. We exclude them to select a field sample (shaded area) from the low-redshift domain of all fields.
Open with DEXTER


In the text

Figure 4: Observed-frame colour-colour digrams: top row, left: two observed-frame colour-indices, each straddling the 4000 Å-break, show galaxies as a 1-parameter family. The grey line roughly indicates the location of the gap in the colour bimodality. Top row, center: when one colour index does not encompass the 4000 Å -break, the galaxies fan out into a wider range. Top row, right: a medium-band colour index probing the 4000 Å-break. Bottom row: template colours (black lines) are plotted over the galaxy sample (grey points). The left-most line is an age sequence from 50 Myr ( bottom left) to 15 Gyr ( top right) without dust reddening. The four other lines show colours for the same age sequence with reddening of EB-V=[0.2,0.3,0.4,0.5] omitting the line at 0.1 for clarity.

Open with DEXTER


In the text

	Figure 5: Three galaxy types: left: stellar age vs. EB-V from estimated template parameters. Center/Right: broad-band/medium-band colour. Top row: dust-free old clump in the red-sequence. Middle row: dusty population in the red-sequence. Bottom row: the blue cloud.
Open with DEXTER


In the text

Figure 6: Colours and spectra: top: the mean template of the dusty red-sequence galaxies is redder at the red end and bluer at the blue end than that of the dust-free old clump galaxies. Middle: the dust-free old clump reflects older age in a stronger 4000 Å -break and Ca H and K-lines than the dusty red galaxies. The spectral differences in these templates match up with the mean colour differences of the two SED classes. Bottom: similar trends can be found in the mean observed spectra (we note that the flux calibration is incorrect towards the red).

Open with DEXTER


In the text

	Figure 7: Top: mean 2dF spectra for three galaxy types (restframe). Vertical offsets are used for clarity. The hashed area is affected by night-sky emission lines. Bottom: the mean dusty red spectrum divided by the mean old red spectrum.
Open with DEXTER


In the text

Figure 8: Cluster ( top) versus field ( bottom): left: the age  dust-parameter plot shows great enhancement of red-sequence galaxies in the cluster, as well as a shortage of young (<1 Gyr) populations. The cluster-specific enhancement applies to both, dust-free old galaxies ( center) and dusty red-sequence contaminants ( right). The number of objects in the cluster is roughly twice that in the field (795:385).

Open with DEXTER


In the text

Figure 9: Distribution of galaxy types in (x,y,z): top row: sky maps. The dust-free old galaxies are concentrated towards the cluster cores, while the other two types are more scattered over the field. Middle row: the cluster cores form two main velocity components, a North group and a South group separated by  km s-1 in rest-frame velocity. The dusty red galaxies have a similar velocity distribution as the old red objects. Blue galaxies tend to avoid the locus of red galaxies. Bottom row: redshift distributions reflecting mostly internal dynamics.

Open with DEXTER


In the text

Figure 10: Type-density relation: the cluster environment covers projected galaxy densities of . The field points lie at a density of 0.67, but old red (9%) and dusty red (11%) galaxies are slightly offset for clarity. The fraction of intermediate-age dust-reddened galaxies in the red-sequence increases from low-density fields to the outskirts of a cluster environment, but then decreases gradually with density inside the cluster.

Open with DEXTER


In the text

Figure 11: Top row: luminosity distributions: the dust-free old populations follow almost a Gaussian, while the blue cloud could more easily be described by a Schechter function. The cluster shows an excess of dust-free old populations, especially fainter than L*, and of dusty red-sequence contaminants, mostly brighter than L*. The field and cluster samples are each normalised by the number of their galaxies.

Open with DEXTER


In the text

Figure A.1: Redshift quality: 404 bright galaxies observed with 2dF on three COMBO-17 fields, mostly containing galaxies from Abell 901/902. Left panel: comparison of vs. zspec. Center panel: redshift error vs. redshift . Right panel: redshift error vs. R-band aperture magnitude. The 1- redshift error is <0.01, and the outlier rate is <1% (3 out of 404 objects with errors around 0.06).


In the text

Figure B.1: Observed-frame colour-magnitude diagrams at : Left: an observed-frame colour-index straddling the 4000 Å-break makes dust-free old populations (black dots) appear as red as possible in comparison to dust-reddened galaxies. A Butcher-Oemler-style red-sequence cut (line) below the colour-magnitude relation selects some younger dustier galaxies, which can be identified with the photometric information in COMBO-17. Right: a colour index longwards of the 4000 Å-break moves the red-sequence bluewards, further into the bulk of the galaxy population, making an efficient selection of dust-free old populations very challenging.


In the text