7 Optical colours

The estimate of galaxy colours could be biased towards the selection band and affected by an aperture effect. In order to test for this, we first determined colours simply by using SExtractor in the so-called double image mode on the I₈₁₄-selected catalogue, we measured B₄₅₀ magnitudes within the 1$\sigma$ isophotal area determined in the I₈₁₄band ( $B_{{\rm col}}$). The (B-I)_I colour is then the difference between $B_{{\rm col}}$ and the 1$\sigma$ isophotal magnitude in the I₈₁₄ band. We then measured the (B-I)_B colour for the B-selected catalogue using the same method, but with the B₄₅₀ band 1$\sigma$isophotal area as a reference for both the I₈₁₄ ( $I_{{\rm col}}$) and B₄₅₀ magnitudes.

For every source the B-I colours measured on the basis of the B₄₅₀ or I₈₁₄ area are different. In Fig. 8 we show the residuals ( B-I)_I-(B-I)_B vs. I₈₁₄. (B-I)_B is slightly redder than (B-I)_I), probably because the B₄₅₀-band isophote underestimates the I₈₁₄ band flux (broadly speaking I₈₁₄ isophotal areas are larger than B₄₅₀ isophotal areas).

In order to measure colours unbiased by the selection band we chose a different approach, that is we created a meta-image UBVI by summing all four frames, after normalizing each to have the same rms sky noise. We then run SExtractor in the double image mode: detection and isophote boundaries were measured on the combined image, while isophotal magnitudes were measured on U₃₀₀, B₄₅₀, V₆₀₆, I₈₁₄ images individually. Using this procedure (Moustakas et al. 1997) both object detection and isophote determination are based on the summed image, and isophotes are not biased towards any of the bands.

We finally cross-correlated the catalogue obtained from the combined image with the I₈₁₄ sample selected according to our criteria (see Image Analysis). We assigned a lower limit in magnitude to sources undetected in any of other bands (that is, missing in our final U₃₀₀, B₄₅₀, V₆₀₆ catalogues). The limiting magnitude is the 5$\sigma$ isophotal magnitude within the isophote measured in the combined image.

Figure 8: Difference between (B-I) computed in I814 or B450 bands vs. I814. Colours are measured having as a reference the 1$\sigma$ I814 and B450 isophotal areas respectively. ( B-I)I-(B-I)B correlates with the I814 magnitude: (B-I)B is slightly redder than (B-I)I at faint magnitudes, probably because the B450-band isophote underestimates the I814 band flux

Figure 9: Asymmetry index vs. I814 magnitude: sources with both B450-V606 and V606-I814 redder than local ellipticals are shown as stars, while sources bluer than local irregulars are shown as triangles

Figure 10: Central concentration index $C_\eta$ vs. I814 magnitude: sources with both B450-V606 and V606-I814 redder than local ellipticals are shown as stars, while sources bluer than local irregulars are shown as triangles

Figure 11: Central concentration index $C_{{\rm abr}}$ vs. I814 magnitude: sources with both B450-V606 and V606-I814 redder than local ellipticals are shown as stars, while sources bluer than local irregulars are shown as triangles