1 Introduction

 

 

Table 1: Summary of terms.

Lumps:	bright residual features seen in galaxy images after a median filtered version is subtracted from
	the original image; synonymous expressions: bright spots or knots; physical correspondents: star-
	forming complexes encompassing H$\;$II regions, young star clusters, and stellar associations;
Lumpiness index $\chi$:	fractional flux or ratio of the flux due to the lumps within the residual image and the total galaxy
	light of the original image.
Cluster dimension D:	correlation dimension for a discrete set of lump centers in a plane, i.e. within a radius r around
	a typical lump there are $n \propto r^D$ other lumps; no weighting for lump size or luminosity. If consi-
	dered as an indirect measure for a three dimensional medium's fractal dimension, D may be
	related to the volume filling factor of the empty regions, called porosity.
Concentration index CI:	concentration index as the ratio of lump centers in an inner circle and lump centers in an outer
	annulus, normalized according to some prescription; no weighting for lump size or luminosity.

  

Table 2: Galaxy and bright-lump data for the 72 irregular dwarf galaxies of our sample.

$^\dagger$

References to the published papers of our series on Structure and stellar content of dwarf galaxies: I - Bremnes et al. (1998); III - Bremnes et al. (1999); IV - Bremnes et al. (2000); VI - Barazza et al. (2001); VII - Parodi et al. (2002).

Dwarf irregular galaxies exhibit peculiar morphologies that are dominated by the flashy but seemingly irregular presence of star-forming regions. Unlike with more massive disk systems that modulate star formation by spiral density waves, dwarf irregulars in the field - or non-interacting irregulars in general - constitute ideal testbeds for the study of genuine processes regulating local and global star formation and consequently of galactic evolution. A review addressing several key questions concerning large-scale star formation in irregular galaxies is given by Hunter (1997), and an evaluation among simple models for the onset of star formation in irregulars is provided by Hunter et al. (1998).

Important clues as to hidden constraints shaping the heterogeneous appearance of dwarf irregular galaxies may emerge from detailed investigations concerning the spatial distribution of star-forming regions (e.g., Feitzinger & Braunsfurth 1984). In recent years several studies on the distribution of H$\;$II regions and of young, compact star clusters in late-type spirals and in dwarf irregular galaxies have appeared (Telles et al. 1997; Brosch et al. 1998; Elmegreen & Salzer 1999; Heller et al. 2000; Roye & Hunter 2000; Billett et al. 2002). Applying measures like concentration, asymmetry, and fractional-luminosity indices these authors found the star forming regions to be distributed rather randomly, with some tendency to central concentrations particularly for star-bursting systems.

In this paper we extend these previous studies by analysing the distribution of bright spots or lumps in the B-band images of a sample of 72 late-type ("irregular'') dwarf galaxies. The general equivalence of bright lumps in H$\alpha$ and in broad band blue images as tracers of star formation complexes can be appreciated by comparing galaxy images filtered at the two corresponding wavelengths (e.g., Elmegreen & Salzer 1999; Sparke & Gallagher 2000, pp. 139 and 229). With our homogeneous and relatively large sample at hand we aim at comparing three morphological indices with each other (as applied to lumps within irregulars), none of which has been previously reported in the form presented here or within our context. Indices often serve as the quantitative counterparts to qualitative physical concepts. In particular, lump spreading within a galaxy may be described by concentration indices of different apertures; lump clustering may be represented by the correlation dimension for the two-dimensional lump distribution; finally, the lumpiness (or flocculency) of a galaxy may be measured by means of some fractional light index. Table 1 gives a summary of terms and indices that will be more carefully introduced in the subsequent sections.

Comparing relations among morphological indices, we may deepen our insights into the various mechanisms responsible for the morphology of irregulars. The interstellar matter of dwarf irregulars with different global properties may be different (e.g., metallicity, mean gas density, turbulence, gravitational potential), implying differences in the conditions for the formation of stars. Thus aspects concerning the abundance and distribution of star-forming regions within the galaxies, like clustering and star formation rates, may turn out to vary correspondingly. We address this issue by means of the clustering parameter. Another goal of the paper is to contribute to the discussion concerning the influence of shear due to differential rotation on star formation in gas-rich late-type galaxies (Roye & Hunter 2000; Elmegreen et al. 2002). Based on cellular automata simulations we introduce a possible criterium to be checked for in radial lump number distributions; we show that a few of our galaxies indeed meet the criterium, but more research is needed for conclusive results.

The paper is organised as follows. Section 2 gives an overview of the galaxy sample used and presents a table with basic galaxy parameters as well as the parameter values deduced in the subsequent sections. Section 3 describes the adopted lump detection method and introduces a first index, the lumpiness index. Section 4 presents the radial number and number density distributions for the bright lumps of all the galaxies. A concentration index that is normalized according to the galaxies being exponential-disk systems is introduced and applied to the bright lump distribution. We then extensively discuss the possible reason for a peculiarity seen in the lump number distribution, namely the occurence and relative locations of major and minor peaks. In Sect. 5 we determine the cluster or correlation dimension for the two-dimensional lump patterns, relate it to galaxy absolute magnitude, and deduce a model mean porosity that is linked to the current star formation rate. In Sect. 6 we check for relations among the three indices introduced in the previous sections. We end with a discussion and the conclusions in Sect. 7. Image processing was performed throughout within the IRAF package.

Figure 1: ESO 473-G024. Left: B-band image, taken at the 1.5-m Danish Telescope at La Silla, Chile. Right: residual image after subtracting from the original image its median filtered version.