Detection of Low Surface Brightness Galaxies in 2MASS

Abstract

The study of low surface brightness galaxies (LSB) has become popular in recent years owing to their extreme properties (e.g., mass to light ratio) which make them attractive as the harbingers of the so called "missing" dark (baryonic) matter in the universe. The Two Micron All Sky Survey (2MASS) is expected to detect 1 million galaxies, a fraction of which will be LSB type galaxies. We have carried out a preliminary study of LSB galaxies using near-infrared data acquired with the 2MASS prototype camera. We report the results from a study of the Coma cluster and a field located near the north galactic pole.

What is a Low Surface Brightness galaxy? Historically low surface brightness (LSB) galaxies are delineated from their more "normal" cousins by their central surface brightness. In general, LSB galaxies have a central surface brightness fainter than 22 or 23 mag per sq. arcsec at B band (cf. Bothun et al. 1991). This loose definition allows a wide of variety of objects classified as such, including spirals, ellipticals, dwarf ellipticals, dwarf irregulars and in the deepest studies, globular clusters (associated with giant galaxies). Because they are (possibly) abundant in number and their mass to light ratio is large, LSB galaxies may provide enough mass to cluster systems (e.g., Coma) to close (or significantly narrow) the "missing" dark matter gap. The true nature of these objects is still not well known, including their stellar population(s), metallicity, star formation efficiency and evolution, and as noted, their overall mass. A number of large-area coverage studies have been carried out to address these issues (to site just a few: Sandage et al. 1985; Schombert et al. 1992; Impey et al. 1996; Bothun et al. 1993; Bernstein et al. 1995; Ulmer et al. 1996; Sprayberry et al. 1996; O'Neil et al. 1997; Impey and Bothun 1997). Adopting the central surface brightness criteria of 22 mag per sq. arcsec at B as the definition of an LSB galaxy, we can estimate the surface brightness in the near-infrared using a color of (B-K) = 3 or 4 (B-I ranges from 2 to 3 for LSBs, O'Neil 1997, private comm). Thus, our definition of an LSB galaxy in the near infrared becomes

18.0 - 19.0 mag per arcsec2 at K
18.5 - 19.5 mag per arcsec2 at H
19.0 - 20.0 mag per arcsec2 at J

J:
1-sigma = 21.2 mag per sq. arcsec
2-sigma = 20.5 mag per sq. arcsec

K:
1-sigma = 19.9 mag per sq. arcsec
2-sigma = 19.2 mag per sq. arcsec

Based on the flux limits of the survey and the typical background noise, 2MASS should be sensitive to LSBs with central surface brightness up to 20 mag per sq. arcsec, or equivalent to a B surface brightness of 23 to 24 mag per sq. arcsec. This corresponds to "bright" LSB galaxies, easily detected in deep photographic and CCD optical limited-coverage surveys now underway and far from the state of the art (e.g., Rlim ~27, O'Neil et al. 1997). Nevertheless, 2MASS should detect a wealth of LSB type galaxies due to the all sky coverage. This paper provides some preliminary information on what we should expect to see in 2MASS with LSB galaxies using data acquired for the Coma cluster (z=0.023), and the galactic pole region SA 57 . Questions we address: How many LSB galaxies do we detect? Can these galaxies be well fit by an exponential (thus suggesting a population type)? How do their colors compare with normal galaxies? Outstanding Issues: Are normal or high surface brightness galaxies (HSB) and LSB galaxies distinct populations? Or, are LSB galaxies simply a "faint" extension of the HSB -- normal -- galaxy population?

The data was acquired with the 2MASS prototype camera (single channel mode), during observing runs at KPNO during spring of 1995. Five square degrees, comprising the SA57 region (NGP) and one square degree, centered on the Coma cluster, were examined in detail. JHK 2MASS scans cover about 1 sq. degree (in 6 degree strips). One scan through Coma contains over 150 galaxies down to J~15. The sensitivity and depth of the 2MASS image product is demonstrated in Figure 1, in which we show some J-band examples of LSB galaxies found in the Coma cluster. The integrated J mag for the set ranges from 14 to 15.

• J-band "postage stamp" images of a representative sample LSB galaxies toward the Coma cluster. The images are about 1 arcmin in width. Integrated mags range from 14 to 15.

The central surface brightness was computed from the center or peak pixel value of the galaxy (for 2MASS, the pixel size is 1 arcsec, but the true resolution is closer to 2 arcsec). Figure 2 shows the central surface brightness vs. integrated mag for LSB galaxies, normal (or HSB) galaxies and for stars. The horizontal dashed lines denote the limit for LSB candidacy, as given by the adopted definition given in section 1. Note that LSB galaxies need only satisfy the limit in at least one band (thus, for example, a galaxy may satisfy the limit in K band, but not in J band).

• Central surface brightness vs. the integrated flux. LSB galaxies are denoted by the filled squares, HSB galaxies by small triangles and stars by small points. The horizontal dashed lines represent the defined limit for LSBs -- they need only satisfy the limit in at least one band. The vertical dashed lines represent the sensitivity (completeness) limit the 2MASS survey.

There are a number of LSBs that are situated well above the HSB population, but they tend to be very faint (integrated flux) subject to significant distortion from background noise. For the brighter LSB galaxies, the central surface brightness is not significantly different than HSB galaxies with comparable integrated flux (see the J band in particular).

Do LSB galaxies form a distinct population in Figure 2? It seems more likely that the Coma+Sa57 LSBs are simply an extension of the HSB or normal galaxy surface brightness curve. Moreover, the colors of LSB galaxies (see Figure 3) range from 0.5 to 1.5 in J-K, similar to HSB galaxies -- particularly if you take into account the scatter in the color measurement.

• Color vs. brightness for Coma galaxies. LSBs are denoted by filled square symbols (and thicker errorbars) and HSBs by small triangles.

An exponential function was fit to the mean surface brightness profile (per band) for each LSB galaxy. For the brighter galaxies in which the elliptical parameters are well determined, the radial profile was constructed from elliptical annuli. For most of the remaining LSB galaxies, the annuli were circular.

The exponential was fit over a range of 2 to 10 arc seconds -- comprising most (if not) all of the area that is not lost in the background noise for a typical LSB galaxy. The inner 2 arc seconds was avoided to minimize the effect of the PSF. Only points with >2*sigma where used in the fit, where sigma represents the uncertainty in the mean surface brightness measurement (poisson statistics). Figure 4 shows a typical LSB galaxy and its radial profile.

• SA57 LSB galaxy . As seen in J (upper left panel), H (upper right), K (lower left) and optical DSS (lower right). The integrated J mag is ~14.9. Radial surface brightness. The dashed lines represent the best fit exponential. The derived scale length is ~5 arcsec.

The derived scale length for Coma galaxies, Figure 5, ranged from a low of ~1 arcsec (typical of the fainter LSBs -- indicative of our lack of sensitivity), corresponding to about 500 pc (assuming H0 = 75 km/s/Mpc and DM = 34.9 for Coma) to 10 arcsec for bright LSBs with a very flat profile, corresponding to ~4.6 kpc for Coma galaxies. No redshift data is available for the SA57 galaxies so the distance is unknown. Based on their scale lengths, however, the brighter SA57 galaxies are probably closer than the Coma cluster.

• Derived radial profile scale lengths. Low surface brightness galaxies (large squares) and high surface brightness galaxies (denoted by small triangles).

A simple exponential appears to fit the profiles very well in most cases. The results suggest that we are either measuring the "disks" of the LSB galaxies, as opposed to a "spherical" or "bulge" component in which we would expect to have a profile more similar to r**1/4 law or at least a departure from a simple exponential , or it may suggest a selection effect, 2MASS is not as sensitive to dwarf ellipticals as it is to faint spirals. In any case, this result is consistent with that seen in the optical for COMA LSB galaxies (see Bernstein et al. 1995). A larger sample is required to better address this issue.

In this study, we examined small fields centered on the Coma and SA57 regions (near the north galactic pole). Nearly 20% of all galaxies detected in 2MASS have a central surface brightness faint enough to be classified as type LSB, but most are in the faintest integrated flux bins (near the sensitivity limit of the study) and thus their true number density is sure to be incomplete.

Base on the Coma and SA57 preliminary results, the central surface brightness for LSBs appear to be a faint extension of the normal galaxy surface brightness curve. In addition, the J-K color for the LSB galaxies is undistinguishable form that of HSB galaxies. There is no compelling evidence here to suggest that the LSB galaxies belong to a different galaxy population. We caution, however, that the LSB sample is far too small to draw any definitive conclusions at this time.

Simple exponential functions are well fit to most radial profiles, suggesting that we are either measuring the "disks" of the LSB galaxies, as opposed to a "spherical" or "bulge" component which we would expect to have a profiles departing from simple exponentials, or it reflects a kind of selection effect -- LSBs at the bright limit (of the definition threshold) are well fit by exponentials, regardless of their morphology (i.e., we are not sensitive to the bulge/disk differences). Nevertheless, these results are consistent with that seen in the optical for Coma LSB galaxies.

2MASS (and DENIS) should easily detect low surface brightness galaxies in the local group and at least as far as the Coma cluster (z = 0.02). The sample, however, will be far from complete given the sensitivity limits of both all sky surveys.