We present a detailed study of rotational asymmetry in galaxies for both morphological and physical diagnostic purposes. An unambiguous method for computing asymmetry is developed, robust for both distant and nearby galaxies. By degrading real galaxy images, we test the reliability of this asymmetry measure over a range of observational conditions, e.g. spatial resolution and signal-to-noise (S/N). Compared to previous methods, this new algorithm avoids the ambiguity associated with choosing a center by using a minimization method, and successfully corrects for variations in S/N. There is, however, a strong relationship between the rotational asymmetry and physical resolution (distance at fixed spatial resolution); objects become more symmetric when less well-resolved.We further investigate asymmetry as a function of galactic radius and rotation. We find the asymmetry index has a strong radial dependence that differs vastly between Hubble types. As a result, a meaningful asymmetry index must be specified within a well-defined radius representative of the physical galaxy scale. We enumerate several viable alternatives, which excludes the use of isophotes. Asymmetry as a function of angle (A φ ) is also a useful indicator of ellipticity and higher-order azimuthal structure. In general, we show the power of asymmetry as a morphological parameter lies in the strong correlation with (B − V ) color for galaxies undergoing normal star formation, spanning all Hubble types from ellipticals to irregular galaxies. Interacting galaxies do not fall on this asymmetry-color "fiducial sequence," as these galaxies are too asymmetric for their color. We propose to use this fact to distinguish between 'normal' galaxies and galaxies undergoing an interaction or merger at high redshift.1994; Jangren et al. 1999). A different method -applicable for spirals -has been suggested by Elmegreen & Elmegreen (1982): measures of spiral arm morphology, particularly their patchiness, can be used for classification. Related attempts to classify galaxies have included the use of principle component analysis of photometric structures (Whitmore 1984;Watanabe et al. 1985;Han 1995). These systems revealed correlations of physical and morphological features of galaxies, but have not been generally adopted for practical use, and the basic Hubble (1926) system still lives on.A key element missing from recent work listed above is the connection made by Morgan between image structure and stellar content (i.e. between light concentration, or central surface-brightness, and spectral type). Ironically, in parallel to the above efforts to quantify image structure, there has been considerable effort to develop quantitative methods of spectral classification based on broad-band colors (Bershady 1995) and spectra (Connolly et al. 1995, Folkes et al. 1996, Bromley et al. 1998, Ronen et al. 1999. What is needed, then, is to go full circle to where Morgan left off, by tying together the spectral types with the quantitative classification based image structure. Here,...

In this paper we define an observationally robust, multi-parameter space for the classification of nearby and distant galaxies. The parameters include luminosity, color, and the image-structure parameters: size, image concentration, asymmetry, and surface brightness. Based on an initial calibration of this parameter space using the "normal" Hubble-types surveyed by Frei et al. (1996), we find that only a subset of the parameters provide useful classification boundaries for this sample. Interestingly, this subset does not include distance-dependent scale parameters, such as size or luminosity. The essential ingredient is the combination of a spectral index (e.g., color) with parameters of image structure and scale: concentration, asymmetry, and surface-brightness. We refer to the image structure parameters (concentration and asymmetry) as indices of "form." We define a preliminary classification based on spectral index, form, and surface-brightness (a scale) that successfully separates normal galaxies into three classes. We intentionally identify these classes with the familiar labels of Early, Intermediate, and Late. This classification, or others based on the above four parameters can be used reliably to define comparable samples over a broad range in redshift. The size and luminosity distribution of such samples will not be biased by this selection process except through astrophysical correlations between spectral index, form, and surface-brightness.

We explore the galaxian luminosity-metallicity (L-Z) relationship in both the optical and the near-IR using a combination of optical photometric and spectroscopic observations from the KPNO International Spectroscopic Survey (KISS) and near-infrared photometry from the Two-micron All Sky Survey (2MASS). We supplement the 2MASS data with our own NIR photometry for a small number of lower-luminosity ELGs that are under-represented in the 2MASS database. Our B-band L-Z relationship includes 765 star-forming KISS galaxies with coarse abundance estimates from our follow-up spectra, while the correlation with KISS and 2MASS yields a total of 420 galaxies in our J-band L-Z relationship. We explore the effect that changing the correlation between the strong-line abundance diagnostic R 23 and metallicity has on the derived L-Z relation. We find that the slope of the L-Z relationship decreases as the wavelength of the luminosity bandpass increases. We interpret this as being, at least in part, an effect of internal absorption in the host galaxy. Furthermore, the dispersion in the L-Z relation decreases for the NIR bands, suggesting that variations in internal absorption contribute significantly to the observed scatter. We propose that our NIR L-Z relations are more fundamental than the B-band relation, since they are largely free of absorption effects and the NIR luminosities are more directly related to the stellar mass of the galaxy than are the optical luminosities.

No abstract

We use the KPNO International Spectroscopic Survey (KISS) for emission-line galaxies to identify and describe a sample of local analogues to the luminous compact blue galaxies (LCBGs) that are observed to be abundant at intermediate and high redshift. The sample is selected using criteria believed effective at isolating true examples of LCBGs: SB_e(B-band) < 21.0 mag/arcsec^2, M(B) < -18.5 (for H_o = 75 km/s/Mpc), and B-V < 0.6. Additionally, all LCBG candidates presented are selected to have star-formation as their dominant form of activity. We examine the properties of our LCBGs and compare them to those of other KISS star-forming galaxies of the same absolute magnitude range. We find that the KISS LCBGs lie on the extreme end of a fairly continuous distribution of ``normal'' star-forming galaxies in the plane of surface brightness versus color. This result differs from the results of previous studies that show LCBGs at higher-z to be more separate from the ``normal'' (usually non-active) galaxies they are compared against. On average, LCBGs have a higher tendency to emit detectable flux in the radio continuum, have higher H-alpha luminosities by a factor of 1.6, indicating strong star-formation activity, and have slightly lower than expected metal abundances based on the luminosity-metallicity relation for KISS galaxies. We calculate the volume density of our low-z (z<0.045) sample to be 5.4 x 10^-4 h_75^3 Mpc^-3, approximately 4 times lower than the volume density of the LCBGs at 0.4 < z < 0.7 and ~10 times lower than the volume density of the population at 0.7 < z < 1.0.Comment: 18 pages, 11 figures. Accepted for publication in the 20 December 2004 Ap