A new fitting concept for the robust determination of Sérsic model parameters

Abstract

Context.The Sérsic law (SL) offers a versatile, widely used functional form for the structural characterization of galaxies near and far. Whereas fitting this three-parameter function to galaxies with a genuineSLluminosity distribution (e.g., several local early-type galaxies–ETGs) yields a robust determination of the Sérsic exponentηand effective surface brightnessμeff, this is not necessarily the case for galaxies whose surface brightness profiles (SBPs) appreciably deviate, either in their centers or over an extended radius interval, from theSL(e.g., ETGs with a “depleted” core and nucleated dwarf ellipticals, or most late-type galaxies-LTGs). In this general case of “imperfect”SLprofiles, the best-fitting solution may significantly depend on the radius (or surface brightness) interval fit, the photometric passbands considered and the specifics of the fitting procedure (photometric uncertainties of SBP data points or image pixels, and corrections for point spread function (PSF) convolution effects). Such uncertainties may then affect, in a non-easily predictable manner, automated structural studies of large heterogeneous galaxy samples and introduce a scatter, if not a bias, in galaxy scaling relations and their evolution across redshift (z).Aims.Our goal is to devise a fitting concept that permits a robust determination of the equivalentSLmodel for the general case of galaxies with imperfectSLprofiles.Methods.The distinctive feature of the concept proposed here (iFIT) is that the fit is not constrained through standardχ2minimization between an observed SBP and theSLmodel of it, but instead through the search for the best match between the observationally determined and theoretically expected radial variation of the mean surface brightness and light growth curve. This approach ensures quick convergence to a unique solution for both perfect and imperfect Sérsic profiles, even shallow and resolution-degraded SBPs. iFITallows for correction of PSF convolution effects, offering the user the option of choosing between a Moffat, Gaussian, or user-supplied PSF. iFIT, which is a standalone FORTRAN code, can be applied to any SBP that is provided in ASCII format and it has the capability of convenient graphical storage of its output. The iFITdistribution package is supplemented with an auxiliary SBP derivation tool in python.Results.iFIThas been extensively tested on synthetic data with a Sérsic index 0.3 ≤ η ≤ 4.2 and an effective radius 1 ≤ Reff (″)≤20. Applied to non PSF-convolved data, iFITcan infer the Sérsic exponentηwith an absolute error of ≤ 0.2 even for shallow SBPs. As for PSF-degraded data, iFITcan recover the inputSLmodel parameters with a satisfactorily accuracy almost over the entire considered parameter space as long as FWHM(PSF) ≤ Reff. This study also includes examples of applications of iFITto ETGs and local low-mass starburst galaxies. These tests confirm that iFITshows little sensitivity on PSF corrections and SBP limiting surface brightness, and that subtraction of the best-fittingSLmodel in two different bands generally yields a good match to the observed radial color profile.Conclusions.It is pointed out that the publicly available iFIToffers an efficient tool for the non-supervised structural characterization of large galaxy samples, as those expected to become available with Euclid and LSST.