2 Photometric observations and data reduction

The most uniform CCD photometry of the inner square degree of the Coma cluster is provided by Jørgensen & Franx (1994, JF94). The sample Lucey et al. (1991) investigated, is not complete, while the observations of Andreon et al. (1996, 1997) have been performed partially in very bad seeing conditions.

No CCD photometry of similar quality and uniformity as the data of JF94 exists for the galaxies of the outer region of the Coma cluster. Since one special issue of this project is to investigate the structure of galaxies as a function of their environmental density, complementary photometry was needed in the outer regions of the Coma cluster. Therefore we selected the 20 galaxies brighter than b_26.5=15.0 mag and with projected distances larger than 40 arcmin from the cluster center from the (photographic) catalogue of Godwin et al. (1983, GMP83). For these galaxies we obtained Kron-Cousins $R_{\rm c}$ band CCD surface photometry at the 1.3 m telescope of the MDM observatory in March 1995. We used a Loral 2048 $\times$ 2048 chip with a scaling of 0.508'' per pixel and observed the galaxies for 500 s with a PSF of 1.5-2''. Sky subtraction, cosmic ray elimination via a $\kappa - \sigma$ clipping procedure and the removal of foreground stars were performed with MIDAS packages. Following Bender & Möllenhoff (1987, BM87) the surface brightness profile of the galaxies were obtained by fitting the isophotes with ellipses. Furthermore the deviations from ellipses were studied by means of Fourier series expansions (see BM87). The resulting fourth cosine coefficients a₄ indicate boxy (negative values) or disky isophotes (positive values). Based on surface brightness, ellipticity, the a₄ and a₆ profiles as well as visual inspection eight of the investigated objects were classified to be early-type galaxies. The results of the isophotal shape analysis of these 8 galaxies are shown in Fig. 1 and listed in Tables 5 to 12 in Appendix A.

Also based on surface brightness, ellipticity and a₄ profiles as well as visual inspection of the CCD frames we classified the remaining 11 late-type galaxies as follows. S0/Sa: GMP3273; Sa: GMP1900, GMP2544, GMP2987, GMP5006, GMP6523; Sb: GMP3837, GMP6339; SB0: GMP5886; SBa: GMP6265; SBb: GMP5437. The data of these galaxies will be presented and discussed in a separate paper.

Figure 1: The results of the isophote shape analysis of the CCD photometry of the 8 early type galaxies with D>40' and b26.5<13 mag. The plots to the left show (from top to bottom) the R-band surface brightness, the a3/a, a4/a, a6/a coefficients as a function of the 1/4 power of the semi-major axis distance a in arcsec. The surface brightness is shown along the major (upper profile) and the minor axis (lower profile). The plots to the right show (from top to bottom) the position angle, the ellipticity and the coefficients b3/a, b4/a, b6/a

Figure 1: continued

In order to be comparable to the data from JF94 we derived the luminosity weighted mean of a₄

$$\langle a_4 \rangle := \frac{\int_{a_{\rm min}}^{a_{\rm max}... ... I\,{\rm d}a};\,\,\,\,\,\, (a \mbox{ -- the semi-major axis)}$$ (1)

for the 8 early-type galaxies of the observed sample. The integration is done from $a_{\rm min}~=~1.5''$ to the maximum semi-major axis $a_{\rm max}$ at which a₄ could be determined (see Tables 5 to 12). $\langle a_4 \rangle$ as well as the ellipticity at $r_{\rm e}$ of these 8 objects is listed in Table 1 (Nos. 28-35). According to BM87 the typical errors of a₄ and $\epsilon$ are 0.2% and 0.005, respectively.

For 5 of the 8 early-type galaxies in the outer sample HST V band surface photometry is available (Principal Investigator: John Lucey; Proposal ID: 5997). Because our data were taken in non-photometric nights we used these HST images to determine the Vband zero-point and calibrated our R-band frames applying a typical colour for elliptical galaxies $V-R_{\rm KC}=0.6$ mag (Poulain & Nieto 1994, PN94). The scatter in $V-R_{\rm KC}=0.07$ mag and dominates the error in the zero point, while the accuracy due to the HST photometry is about 0.01 mag (Holtzman et al. 1995). For the remaining 3 galaxies (GMP0756, 1176, 1990) we computed the zero-point using the b magnitudes inside the 26.5 mag/arcsec² surface brightness isophote as given by GMP83, adopting their mean b-r colours. Their r band magnitudes are approximately equal to the Johnson R. To convert this Johnson R in Kron-Cousins magnitudes we used the colour equations given by Fernie (1983): $(V-R)_{\rm J} = 0.034 (\pm0.05) + 1.364 (\pm 0.016)$ ( $V-R)_{\rm KC}$. Since V is the same in both systems and $V-R_{\rm KC}=0.6$ mag from PN94, this yields to $R_{\rm KC} = R_{\rm J} +0.25$ ($\pm 0.06$). The accuracy of the zero-point is limited by the error given by GMP83 to be $\delta$ mag = 0.15 mag.
Total magnitudes $m_{\rm T}$, half luminosity radii $r_{\rm e}$and the surface brightness $SB_{\rm e}$ at $r_{\rm e}$ were derived using the algorithm of Saglia et al. (1997a), which optimally fits the (circularly averaged) surface brightness profiles as a sum of seeing convolved r^1/4 and exponential components and corrects for sky subtraction errors. Saglia et al. (1997b) showed that the systematic errors due to the extrapolation method, sky subtraction and seeing convolution for the total magnitudes $\delta m_{\rm T} < 0.15$ mag and for the effective radius is $\delta r_{\rm e} < 10$%. On the other hand the deviations that are introduced by using circular apertures instead of elliptical ones are significantly smaller ( $\delta m_{\rm T} < 0.05$ mag for $\epsilon < 0.5$; see Saglia et al. 1997a). Finally, the values of the total magnitude $m_{\rm T}$ and the surface brightness SB have been corrected for galactic extinction and K - correction has also been applied. We have used the B band galactic extinction determined by Burstein & Heiles (1984). For GMP 0144, GMP 5568 and GMP 5975 A_B = 0.03 mag and can be converted to the R band using A_R=0.58 A_B (Seaton 1979). For the remaining five galaxies the same extinction is assumed. The K-correction was provided by Rocca-Volmerange & Guiderdoni (1988, K_R = 0.028 mag). The surface brightness was also corrected for cosmological (1+z)⁴dimming.
Table 1 shows the derived photometric parameter $R_{\rm c_{\rm T}}$, $\langle SB_{\rm e}\rangle$ and $r_{\rm e}$ for the 8 observed early-type Galaxies in the outer part of the Coma cluster (Nos. 28-35).

Five of the 8 early type galaxies identified here were already previously observed by Lucey et al. (1991, Letal91). For four objects (Nos. 28, 29, 33, 35) the differences from our derived magnitudes are lower than the systematic errors (0.15 mag) described above. Only for galaxy No. 34 (GMP 5568) the difference is larger (0.75 mag), but Letal91 state that their photometric parameters are poorly defined. Their derived effective radii tend to be smaller than the values derived in this paper, since Letal91 fitted the surface brightness profile only with an r^1/4 law, neglecting any exponential component. Nevertheless the mean difference of the quantity $\log r_{\rm e} - 0.3 \langle SB_{\rm e}\rangle$ (see Saglia et al. 1997b) is negligible (0.03) and showing a scatter of only 0.01. Hence there is no systematic zeropoint offset the two dataset.