Surface brightness profiles of dwarf galaxies in the NGC 5044 Group

S.A. Cellone: Dwarf galaxies in the NGC 5044 Group 409orS e = S 0 + 5 log(reα) [ Γ(2− 2.5 logN )N]. (5)It can be shown that reα depends only on N, and then S e doesnot depend on α. These calculated parameters are also shown inTable 4. The magnitudes and colours in Table 3, instead, wereobtained from the observed profiles with no model assumption,to allow for those galaxies whose profiles couldn’t be fitted withthe model.4. Seeing effectsThe effects of seeing on the photometric parametersof elliptical galaxies have been studied using differenttechniques (e. g.: Franx et al. 1989; Peletier et al. 1990;Saglia et al. 1993b). These studies have shown that the effects ofseeing on surface brightness and ellipticity profiles may extendup to several PSF radii.The observations presented in this paper were obtained underrather mediocre seeing conditions and, in addition, the telescopehad small tracking errors. This resulted in a non-circularPSF, with Gaussian fits to its minor and major axis givingFWHM Y ≃ 2.0 ′′ and FWHM X ≃ 2.8 ′′ , respectively, withsmall variations between different frames. However, the actualshape of the PSF was not Gaussian, having notably larger wings.The adopted approach, then, was to evaluate the effects of seeingon this particular set of observations using an empirical PSFobtained from the same data using IRAF–DAOPHOT routines.A set of artificial galaxies following a perfect Sérsic law andspanning a broad range in the relevant parameters was generated,and then convolved with the empirical PSF. The surfacebrightness profiles of these convolved artificial galaxies werethen obtained and fitted with a Sérsic law, in the same way aswas previously done with the observed galaxies. The effectsof seeing were evaluated by comparing the “observed” and the“original” parameters for the convolved artificial galaxies.A total of 55 artificial galaxies were generated with fixedcentral surface brightness, and spanning the following rangesin scale length and shape parameter: 0.001 ′′ ≤ α ≤ 10 ′′ , and0.25 ≤ N ≤ 1.50, respectively. Most of them were round, butseveral non-circular (ɛ =0.5) and a few “nucleated” artificialgalaxies (i. e., with a point source 10% the luminosity of thegalaxy added at its centre, before the convolution) were alsogenerated, in order to check for any differences compared tocircular, non-nucleated galaxies. No noise or constant sky levelwere added. Although a complete study of seeing and instrumentaleffects would require the inclusion of these factors, alongwith the generation of a larger sample of artificial galaxies convolvedwith different PSFs, the present approach is sufficient toevaluate the global effects of seeing on the measured photometricparameters of this particular sample of galaxies.Fig. 2 shows the differences between each measured parameterand the original one [∆S 0 , ∆ log(α), and ∆ log(N), respectively]against the measured effective radius r e . It is evident thatfor sufficiently large galaxies the original parameters are recovabcFig. 2a–c. Measured (after convolution with PSF) minus original parametersagainst measured effective radius for 55 artificial galaxies:a central surface brightness, b logarithm of scale length, and c logarithmof shape parameter. Squares: round, non-nucleated galaxies;triangles: ɛ =0.5, non-nucleated galaxies; crosses: round, nucleatedgalaxies. Dashed lines are least-squares fits for log(r e) ≤ 0.5 ′′ .ered with small errors, but for galaxies with r e 5 ′′ the parametersare systematically underestimated because of seeing andsampling effects. (Note that, at least in principle, it seems possibleto recover the original parameters for the smaller galaxies;however, the scatter is large, specially for N, and trying to apply