DEEP NEAR-INFRARED SURFACE PHOTOMETRY OF 57 GALAXIES IN THE LOCAL SPHERE OF INFLUENCE

The observational properties of nearby galaxies, such as fluxes, colors, morphologies, and sizes, reflect their underlying physical properties (stellar/baryonic and dark matter content, star-formation rates, formation history, and angular momenta); however, the way these observational and physical properties are exactly related is still poorly understood. By technical necessity, the observational quantities are mainly based on the optical B band (390–480 nm). However, galaxies evolving in low-density environments with little external stimulation for star formation often contain significant quantities of dust (e.g., Driver et al. 2007), which can attenuate and distort their optical light profiles. In contrast, dust attenuation is vastly reduced at near-infrared (NIR) wavelengths, hence the NIR provides a spectral regime where a more accurate, unaltered representation of a galaxy's underlying stellar distribution can be obtained (Gavazzi et al. 1996a). Furthermore, the stellar mass of most galaxies is dominated by the quiescent old stellar component whose energy output peaks at NIR wavelengths. Even in the extreme case of Blue Compact Dwarf (BCD) galaxies, previously thought to be primeval galaxies forming their first stars at the present epoch (Thuan & Izotov 1997), the analysis of their resolved stellar populations has revealed the presence of stars that are at least a few Gyrs of age (e.g., the BCD galaxies: VII Zw 403, Mrk 178, and I Zw 36 as discussed by Schulte-Ladbeck et al. 1998, 2000, 2001, respectively; SBS 1415+437 discussed by Aloisi et al. 2005; I Zw 18 by Aloisi et al. 2007; and CGCG 269-049 by Corbin et al. 2008).

In order to obtain a deeper understanding of the connection between light and matter distribution in galaxies, a representative sample of nearby stellar systems needs to be studied in detail. The Local Sphere of Influence (LSI; D < 10 Mpc) contains large numbers of early- (dE) and late-type (dIrr) dwarf galaxies, which make up about 85% of the local galaxy population (Kraan-Korteweg & Tammann 1979; Schmidt & Boller 1992; Karachentsev et al. 2004). Dwarf galaxies contribute about 4% to the local luminosity density and about 10–15% to the local H i mass density (Karachentsev et al. 2004). Due to their proximity to the Milky Way, LSI galaxies are ideal for a NIR study that requires a significant number of dwarf systems.

Previous NIR surveys include the Two Micron All Sky Survey (2MASS; Skrutskie et al. 2006) and deeper, targeted galaxy surveys (Gavazzi et al. 1996c, 1996b, 2000; Boselli et al. 2000). 2MASS photometry for galaxies suffers from a number of important drawbacks that become more evident as the samples of independently-investigated galaxies become larger. The short integration time of 2MASS observations resulted in most of the low surface brightness (LSB) dwarfs in the LSI remaining undetected, and, if they were detected, 2MASS underestimated the fluxes by as much as 70% (Andreon 2002). The targeted H-band observations of Gavazzi et al. (1996c, 1996b, 2000) and Boselli et al. (2000) were inherently deeper; however, the samples included few LSB dwarfs. This serious limitation demands a deeper and higher-resolution study to investigate those galaxies that were beyond the reach of photometric NIR studies to date.

A reference atlas of images needs to have the necessary spatial resolution to probe the morphological fine structure of these nearby galaxies and contain a significant number of dwarf galaxies that are generally overlooked. An LSI sample has the additional advantage that an increasingly large number of nearby galaxies have accurately known distances. Karachentsev et al. (2006) reported that 214 out of 451 LSI galaxies have distance estimates (with less than 10% uncertainty) by means of the tip magnitude of the red giant branch (TRGB), the Tully–Fisher relation, and the surface brightness fluctuations (SBF) method (e.g., Jerjen et al. 1998, 2001; Karachentsev et al. 2004). The remaining galaxies have rough distance estimates from either the luminosity of their brightest stars, their radial velocities, or their suspected membership to a known galaxy group.

The purpose of this paper is to present a NIR H-band (1.65 μm) atlas of 57 LSI galaxies, probing to flux levels of approximately 4 mag arcsec⁻² (or 40 times) fainter than 2MASS. The majority of the galaxies presented here are much fainter than those in previous targeted surveys. We derive photometric parameters for each object, such as the total magnitude, the effective radius, the effective surface brightness, and Sérsic fitting parameters. By using the best distances that are currently available in the literature, we are able to derive physical parameters such as their luminosities and stellar masses.

The paper is organized as follows. We describe the sample selection in Section 2. In Sections 3 and 4, we discuss the observing strategies, the data reduction, and the photometric calibration of the images. The 11 galaxies in the sample that remained undetected at our faint detection limit or had images that could not be usefully analyzed are discussed in Section 5. In Section 6, the new data are compared to 2MASS photometry and optical (B-band) data, and the luminosity–surface brightness relation is discussed. Interesting properties of individual galaxies are described in Section 7. Finally, the results are summarized in Section 8.

We have compiled a list of 470 galaxies with estimated distances less than 10 Mpc from the Milky Way,⁵ from the catalogs of Schmidt & Boller (1992), Côté et al. (1997), Jerjen et al. (2000) and Karachentsev et al. (2004). Approximately 70% of LSI galaxies are members of seven nearby galaxy groups including the Local Group (LG). Each group contains one or more massive spiral or elliptical galaxies accompanied by a population of dwarf satellites, which tend to be dwarf ellipticals (dE). The southern hemisphere contains 174 LSI galaxies, 113 of which are members of a nearby group. We randomly selected 68 program galaxies with a range of total apparent B-band magnitudes (between m_B = 9 and 18 mag, as well as several with no optical detection to date) and morphologies (Hubble types E3 through to Sc, including many irregular and dwarf galaxies), 19 of which were members of a nearby group. Therefore, our sample contains 80% $\big(\!\!=\!\!\frac{68-19}{174-113}\big)$ of the southern-hemisphere field galaxies and 17% (=19/113) of the group members. The distribution of these LSI galaxies is shown in Figure 1.

Zoom In Zoom Out Reset image size

The selected galaxies further provide a complementary data set to the Local Volume H i Survey⁶ (LVHIS), which is an H i imaging survey of all LSI galaxies south of declination δ = −30° that were detected by the H i Parkes All-Sky Survey (HIPASS; Barnes et al. 2001). The former H i survey is currently being carried out at the Australia Telescope Compact Array (ATCA). The basic properties of our sample galaxies are listed in Table 1, which is organized as follows: Column (1)—galaxy name; Column (2)—morphological type in the Hubble (1936), Sandage (1961), and Sandage & Binggeli (1984) classification schemes; Columns (3) and (4)—equatorial coordinates for the epoch J2000; Column (5)—total B-band magnitude and its source (when the uncertainty associated with this value is not provided, an error of 0.2 mag is adopted); Columns (6) and (7)—distance to the galaxy (from Karachentsev et al. 2004, 2006; Seth et al. 2005; Carrasco et al. 2001) with an indication of the method used: TRGB, SBF, (MEM) group membership, or (H) Hubble flow distance D = v_LG/H₀, where H₀ = 73 km s⁻¹ Mpc⁻¹ is adopted (Wilkinson Microwave Anisotropy Probe, WMAP; Spergel et al. 2007); Column (8)—heliocentric radial velocity, v_☉, from the NASA Extragalactic Database (NED); Column (9)—LG velocity, v_LG, from NED; Column (10)—reddening estimate, E(B − V), from Schlegel et al. (1998; the associated error is 16%); and Columns (11)–(14)—the observing log, which includes the observation date, strategy, total exposure time, and seeing.

NIR H-band images were obtained for the 68 program galaxies during five observing runs between 2004 October and 2006 September using the Infrared Imager and Spectrograph 2 (IRIS2; Tinney et al. 2004) at the 3.9 m Anglo-Australian Telescope (AAT). Table 1 lists the observing log of these observations. Atmospheric conditions were clear, if not always photometric, and the seeing ranged from 10 to 29 with a mean of 13. The IRIS2 detector is a 1024 × 1024 Rockwell HgCdTe Astronomical Wide Area Infrared Imager-1 (HAWAII) array with a pixel scale of 045 pixel⁻¹, resulting in an instantaneous field of view (FOV) of 77 × 77.

Two different observing strategies were employed, depending on the anticipated angular extent of the target compared to the IRIS2 FOV.

• 1.  
  Jitter Self-Flat (JSF). The majority of our sample has an optical diameter less than 2' and, given that the sky is typically 7–8 mag arcsec⁻² brighter in the H band than in the B band, we anticipate that these objects barely fill 10–20% of the array in the infrared. These targets were observed in a 3 × 3 grid pattern with a spacing of 90'', resulting in a 47 × 47 overlap region, common to all pointings, that encompasses not only the target galaxy, but also a substantial amount of the background sky. A maximum of ∼30 s was spent on any one pointing, consisting of multiple 5–10 s integrations (depending on the sky brightness at the time, while aiming to keep the combined object + sky counts well within the linear regime), which were then averaged before being stored. This nine-point jitter pattern was repeated up to six times, leading to a total on-source exposure time of just under half an hour per galaxy. This method was also employed on larger, but well-resolved, targets such as the Argo Dwarf Irregular galaxy.
• 2.  
  Chop Sky Jitter (CSJ). In accordance with the recommendations of Vaduvescu & McCall (2004), objects filling ≳40% of the array FOV require matching observations of adjacent blank sky to track changes in the background level and illumination pattern. Five jittered observations (10'' offsets) of the target galaxy were bracketed and interleaved with six jittered observations of the (relatively blank) sky, 10' north or south. At each object or sky jitter position, 3 × 10 s or 6 × 5 s integrations were averaged. This pattern was repeated between five and 12 times, for a total on-source exposure time of up to half an hour per galaxy.

The data reduction was carried out by using the ORAC-DR⁷ pipeline within the starlink package. Observations made with the JSF method employed the JITTER_SELF_FLAT recipe, while those with the CSJ method used the CHOP_SKY_JITTER recipe. Preprocessing of all raw frames included the subtraction of a matching dark frame, linearity and interquadrant crosstalk correction, and bad pixel masking.

Considerable care is taken to ensure accurate flat-fielding over the entire field of the array. For JSF observations, we create an interim flat field by taking the median at each pixel of the nine normalized object frames, then divide each of the nine images by this interim flat field. Extended sources within these flat-fielded object frames are automatically detected and masked, and an improved flat field is created from the masked versions of the nine normalized object frames. We apply a correction for astrometric distortion internal to IRIS2 by resampling the proper flat-fielded images; then spatial additive offsets between images are computed using point sources common to all images. The nine images are mosaiced together by applying offsets in intensity to the registered images in order to produce the most consistent sky value possible in the overlap regions. Next, we construct a new flat field and mosaic for each set of nine jittered frames; then all the mosaics are registered and coadded to form a master mosaic. Occasionally, significant variations in the level and/or structure of the background sky (on a temporal or spatial scale smaller than that sampled by the array within the ∼5 minute period of the nine jittered frames) resulted in a noticeable residual structure in the ensuing mosaic, forcing us to exclude that mosaic from the master mosaic. The total on-source exposure time, after discarding such data, is shown in Table 1.

For CSJ observations, the six sky frames are first offset in intensity to a common modal value, then a flat field is formed from the median value at each pixel. All six sky frames and five object frames are flat-fielded, then the modal pixel values of the two sky frames bracketing each object frame are averaged and subtracted from that object frame. We then perform image registration and mosaicing on each set of five sky-subtracted object frames, as was similarly done for the JSF observations. These mosaics are registered and coadded to form a master mosaic (with the exception of any mosaic showing a residual sky structure as described above), yielding the on-source exposure times as shown in Table 1.

Of the 68 galaxies observed, 11 remained undetected or could not be usefully analyzed. This was because either our H-band surface brightness limit of μ_lim ≈ 25 mag arcsec⁻², or 20 L_☉ arcsec⁻² at a distance of 1 Mpc (adopting M_H,☉ = 3.35 mag; Colina et al. 1996), was not low enough, or the galaxy light was heavily contaminated by Galactic foreground stars (see Table 2). In both situations, the data were not processed further. For instance, the companion galaxies NGC 2784 DW1 and KK 98-73 were observed parallel to NGC 2784. While NGC 2784 and KK 98-73 can be seen, NGC 2784 DW1, located between NGC 2784 and KK 98-73, is barely visible (see Figure 5, second row, middle panel, where KK 98-73 is visible to the bottom left of the image). The images of the other 57 galaxies are shown in Figures 2–6.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 2. Galaxies Not Analyzed Further

Name (1)	R.A. (J2000.0) (2)	Decl. (J2000.0) (3)	Reason (4)	mH (mag) (5)	MH,0 (mag) (6)	$\log _{10} ({\mathcal{M}}_{\ast })$ (7)
KK 2000-03	02:24:44.58	−73:30:49.2	Marginal detection	...	...	...
KK 2000-04	03:12:46.14	−66:16:12.5	No galaxy detected	>11.8	> − 16.5	<7.0
KK 2000-06	03:14:26.14	−66:23:27.9	No galaxy detected	>11.8	> − 16.5	<7.9
ESO 490-G017	06:37:57.09	−26:00:03.1	Foreground stars	...	...	...
HIZSS 003	07:00:29.3	−04:12:30	Foreground stars	...	...	...
ESO 558-PN011	07:06:56.80	−22:02:26.0	Foreground stars	...	...	...
AM 0717-571	07:18:37.90	−57:24:46.5	No galaxy detected	>11.8	> − 18.7	<8.8
NGC 2784 DW1	09:12:18.5	−24:12:41	No galaxy detected	>11.8	> − 18.3	<8.7
SJK98 J1616-55	16:16:49.0	−55:44:57	Galactic plane	...	...	...
HIZOAJ 1616-55	16:18:46	−55:37:30	Galactic plane	...	...	...
ESO 594-G004	19:29:58.97	−17:40:41.3	Foreground stars	...	...	...

Download table as:  ASCIITypeset image

Instrumental magnitudes for 50–100 field stars were measured on each image by employing the standard Interactive Reduction and Analysis Facility (IRAF) point-spread function (PSF) fitting routines. Cross-correlating the stellar positions with the 2MASS Point Source Catalog provided H-band magnitudes and allowed the photometric calibration of each field (see Figure 7). The stars that deviate from the 45° line were usually either extremely red or blue, where the transformation between 2MASS and IRIS2 H bands (Ryder 2007) broke down. The 1σ uncertainty in the zero point was calculated to be between 0.01 and 0.04 mag, depending on the number of stars used for the calibration.

Zoom In Zoom Out Reset image size

To ensure accurate galaxy surface photometry down to the faintest possible isophotes, we cleaned the images of foreground stars by using procedures written within the IRAF package. Thereby, stars in the field around a galaxy were carefully replaced with nearby patches of plain sky. If superposed on the galaxy, the galaxy light under the star was restored by replacing the contaminated area with its mirror image with respect to the galaxy center. The galaxy center was defined as the center of the luminosity-weighted light distribution. The star-removal process was visually monitored to identify small-scale structures and asymmetries, and to ensure accurate removal of foreground stars whilst not removing sources associated with the galaxy itself. The effectiveness of the cleaning procedure is illustrated in Figure 8, where pre- and post-cleaning images are shown for the two galaxies ESO 468-G020 and IC 1959.

Zoom In Zoom Out Reset image size

Both the fields of IC 5152 (Figure 4) and UGCA 438 (Figure 6) have a bright foreground star, which obscures a large portion of the galaxy. To obtain rough photometric parameters for these objects, the contribution of the bright stars to the total flux needed to be removed. Because such a large portion of both galaxies is obscured, it was necessary to remove the affected quarter of the image and replace it with the opposite quarter, which was rotated by 180° (i.e., in the case of IC 5152, the top-right quarter was replaced with the rotated bottom-left quarter of the image). This method failed in the case of the dIrr galaxy AM 0521-343, where an even brighter star (CD-34 2225, m_H = 7.1) near the faint dwarf galaxy prevented proper cleaning (see Figure 2).

Simulated circular aperture photometry of the star-subtracted H-band images produced a growth curve as a function of the geometric mean radius $\sqrt{ab}$ (where a and b are the galaxy's major and minor axes, respectively). The asymptotic intensity corresponds to the total apparent magnitude, m_H, that can be recovered down to the background noise level of the image. The largest source of uncertainty is the sky level. By systematically varying the sky brightness, we determined which growth curve converges best to a plateau as far as possible from the center of the galaxy. We measured the half-light geometric mean radius, r_eff, at half the asymptotic intensity and calculated the mean surface brightness within that radius: 〈μ_H〉_eff. The overall uncertainty for the total magnitude, m_H, is between 0.05 and 0.30 mag; for the mean effective surface brightness, 〈μ_H〉_eff, it is less than 0.2 mag arcsec⁻²; and for the half-light radius, r_eff, the uncertainty is of the order of 5%. The image of AM 0521-343 (Figure 2) contains a bright foreground star, and thus the sky brightness plus the contribution from the stellar halo was estimated at the galaxy's position, and simulated aperture photometry was performed out to the radius of asymptotic intensity (defined on the side of the galaxy opposite to the contaminating star).

We determined the radial surface brightness profile of a sample galaxy by differentiating the growth curves with respect to radius. Depending on the total integration time of the image, the profiles could be reconstructed down to a surface brightness limit between 24.5 mag arcsec⁻² < μ_lim < 26 mag arcsec⁻². They are shown, with a linear radius scale, in Figures 9–11. The error bars are calculated as the rms scatter of the intensity along each isophote. We calculated the position angle and ellipticity for each isophote, with the galaxy center fixed, with IRAF's ELLIPSE package as a function of radius. The ellipticity and position angle were, in general, settled in the outer regions of the galaxies; however, in some cases, they varied significantly in the inner regions.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 3 lists the measured properties of the 57 sample galaxies: Column (1)—galaxy name; Column (2)—observed integrated apparent magnitude, m_H,obs; Column (3)—effective radius, r_eff; Column (4)—mean effective surface brightness, 〈μ_H〉_eff; Column (5)—ellipticity; and Column (6)—position angle of the major axis measured in degrees from North through East (P.A.=90°). The ellipticity and position angle listed are those of the outermost fitted isophote.

The surface brightness profile for each galaxy was fitted with the Sérsic function, μ(r) = μ₀ + 1.086(r/α)ⁿ (or I(r) = I₀exp(−(r/α)ⁿ)), using IRAF's NFIT1D procedure. Extrapolation of the surface brightness profile to infinity with the help of the Sérsic function allows us to make an accurate estimate of the amount of flux that remained undetected in the sky background noise. By using the Sérsic parameter, n, the scale length, α, and the central surface brightness, μ₀, we calculated the magnitude of the galaxy between the maximum radius and the limit r = ∞:

$$\begin{eqnarray} \Delta m &=& {-}2.5\log _{10}(I_{\rm missing}/I_{\rm tot})\nonumber \\ &=&{-}2.5\log _{10}(\Gamma [2/n,(r_{\rm max}/\alpha)^n]/\Gamma [2/n]), \end{eqnarray} \tag{ 1 }$$

where

$$\begin{eqnarray} I_{\rm missing} &=& \int _{r_{\rm max}}^{\infty } 2\pi I_0 \exp [-(r/\alpha)^n ] r\,dr \nonumber \\ &=&\frac{2\pi I_0 \alpha ^2}{n}\Gamma [2/n,(r_{\rm max}/\alpha)^n] \end{eqnarray} \tag{ 2 }$$

and

$$\begin{eqnarray} I_{\rm tot} &=& \int _{0}^{\infty } 2\pi I_0 \exp [-(r/\alpha)^n ] r\,dr \nonumber \\ &=&\frac{2\pi I_0 \alpha ^2}{n}\Gamma [2/n]. \end{eqnarray} \tag{ 3 }$$

Thereby, Γ(a, x) is the upper incomplete gamma function and r_max was taken to be the radius at which the growth curve reached within 5% of the asymptotic intensity. The missing flux introduces a systematic error to the total magnitude m_H,obs. This correction Δm was typically less than 0.2 mag (see Table 4), except in the few cases, e.g. NGC 3115, where the galaxy's angular size extended beyond the IRIS2 FOV. We note that these galaxies have been observed in the CSJ mode to estimate the sky level from dedicated, blank sky observations. AM 0521-343 was not corrected for missing flux as the bright foreground star prevented the fitting of its surface brightness profile.

Table 4. Galaxy Parameters: Derived

Name (1)	reff (kpc) (2)	μ0 (mag arcsec−2) (3)	n (4)	α (arcsec) (5)	Δm (mag) (6)	Reff (arcsec) (7)	Reff (kpc) (8)	MH,0 (mag) (9)	log${}_{10}({\mathcal{M}}_{\ast }) \,(\log{}_{10}({\mathcal{M}}_{\odot }))$ (10)
SC 18	0.11 ± 0.01	22.11 ± 0.09	2.08 ± 0.3	20.15 ± 1.0	0.18	16.1 ± 1.3	0.14 ± 0.01	−11.5 ± 0.2	5.9 ± 0.2
ESO 349-G031	0.46 ± 0.03	21.57 ± 0.4	1.21 ± 0.2	26.80 ± 4.0	0.05	33.7 ± 6.6	0.52 ± 0.09	−14.6 ± 0.2	7.2 ± 0.2
ESO 294-G010	0.18 ± 0.02	20.11 ± 0.4	1.10 ± 0.5	18.20 ± 5.8	0.14	26.1 ± 13.1	0.24 ± 0.09	−14.2 ± 0.2	7.0 ± 0.2
ESO 473-G024	1.0 ± 0.1	21.63 ± 0.3	1.17 ± 0.3	23.41 ± 6.5	0.28	30.8 ± 11.0	1.2 ± 0.3	−16.1 ± 0.3	7.8 ± 0.2
SC 24	0.09 ± 0.02	21.58 ± 0.4	0.98 ± 0.2	12.97 ± 4.6	0.18	22.4 ± 9.6	0.12 ± 0.04	−10.7 ± 0.3	5.6 ± 0.2
IC 1574	0.72 ± 0.04	20.54 ± 0.08	1.50 ± 0.1	42.91 ± 2.3	0.01	42.9 ± 3.3	1.02 ± 0.06	−16.6 ± 0.2	8.0 ± 0.2
ESO 540-G030	0.50 ± 0.03	21.70 ± 0.4	1.17 ± 0.4	26.77 ± 9.3	0.01	35.2 ± 16.3	0.6 ± 0.2	−14.7 ± 0.2	7.2 ± 0.2
UGCA 15	0.52 ± 0.03	21.51 ± 0.03	1.49 ± 0.07	54.51 ± 1.2	0.05	54.9 ± 2.1	0.89 ± 0.03	−15.0 ± 0.2	7.3 ± 0.2
ESO 540-G032	0.48 ± 0.03	20.89 ± 3.1	0.90 ± 1.7	17.73 ± 19.2	0.15	36.0 ± 84.9	0.6 ± 1.1	−14.7 ± 0.2	7.2 ± 0.2
AM 0106-382	1.0 ± 0.1	19.91 ± 0.4	0.56 ± 0.1	3.33 ± 2.3	0.13	27.2 ± 21.0	1.1 ± 0.8	−16.4 ± 0.2	7.9 ± 0.2
NGC 0625	0.72 ± 0.03	17.20 ± 0.04	0.82 ± 0.1	23.38 ± 0.8	0.05	58.0 ± 10.5	1.1 ± 0.1	−19.2 ± 0.2	9.0 ± 0.2
SC 42	0.04 ± 0.01	18.21 ± 0.6	0.43 ± 0.05	0.75 ± 0.5	0.04	22.5 ± 16.8	0.10 ± 0.03	−10.6 ± 0.2	5.6 ± 0.2
ESO 245-G005	1.07 ± 0.08	20.95 ± 0.04	1.46 ± 0.09	61.74 ± 2.3	0.04	63.3 ± 3.6	1.36 ± 0.07	−17.2 ± 0.2	8.2 ± 0.2
ESO 115-G021	0.76 ± 0.05	18.81 ± 0.1	0.95 ± 0.08	27.52 ± 3.6	0.09	50.3 ± 7.9	1.1 ± 0.1	−17.7 ± 0.2	8.4 ± 0.2
ESO 154-G023	1.39 ± 0.08	19.85 ± 0.03	1.20 ± 0.03	85.02 ± 2.5	0.19	108.1 ± 3.9	3.00 ± 0.06	−18.6 ± 0.2	8.8 ± 0.2
NGC 1313	2.0 ± 0.3	16.41 ± 0.2	0.46 ± 0.03	12.69 ± 3.6	0.86	260.6 ± 83.2	5.2 ± 0.7	−22.3 ± 0.4	10.3 ± 0.2
NGC 1311	0.61 ± 0.04	17.70 ± 0.06	0.93 ± 0.03	19.30 ± 1.2	0.05	36.7 ± 2.6	0.97 ± 0.04	−18.5 ± 0.2	8.7 ± 0.2
AM 0319-662	0.43 ± 0.02	22.19 ± 0.4	1.77 ± 0.8	29.05 ± 6.8	0.06	25.6 ± 8.7	0.5 ± 0.1	−14.2 ± 0.2	7.0 ± 0.2
IC 1959	0.63 ± 0.02	18.46 ± 0.08	1.42 ± 0.06	32.04 ± 1.6	0.04	33.7 ± 1.9	1.04 ± 0.04	−18.2 ± 0.2	8.6 ± 0.2
AM 0333-611	1.35 ± 0.08	20.69 ± 1.5	0.73 ± 0.3	8.58 ± 6.9	0.31	28.6 ± 28.4	1.8 ± 1.4	−17.3 ± 0.2	8.3 ± 0.2
IC 2038	1.3 ± 0.1	17.62 ± 0.2	0.70 ± 0.05	6.32 ± 1.2	0.03	23.7 ± 5.0	1.9 ± 0.3	−19.3 ± 0.2	9.1 ± 0.2
IC 2039	1.13 ± 0.05	15.04 ± 0.6	0.37 ± 0.03	0.25 ± 0.2	0.03	20.1 ± 12.9	1.6 ± 0.7	−19.7 ± 0.2	9.2 ± 0.2
NGC 1705	0.37 ± 0.04	15.33 ± 0.4	0.44 ± 0.04	0.90 ± 0.4	0.15	23.7 ± 11.6	0.6 ± 0.2	−18.5 ± 0.2	8.8 ± 0.2
NGC 1744	1.8 ± 0.1	17.86 ± 0.1	0.69 ± 0.05	21.97 ± 3.6	0.13	86.1 ± 16.7	3.3 ± 0.4	−20.3 ± 0.2	9.5 ± 0.2
AM 0521-343	0.36 ± 0.01	...	...	...	...	...	...	−15.9 ± 0.5	7.7 ± 0.3
KKS 2000-55	1.59 ± 0.04	18.05 ± 0.4	0.63 ± 0.09	11.36 ± 4.5	0.15	59.6 ± 27.1	2.9 ± 0.7	−20.2 ± 0.2	9.4 ± 0.2
ESO 364-G029	1.30 ± 0.07	21.11 ± 0.05	1.78 ± 0.1	60.81 ± 2.2	0.02	53.3 ± 2.7	1.94 ± 0.07	−17.5 ± 0.2	8.3 ± 0.2
ESO 121-G020	0.52 ± 0.04	20.82 ± 0.8	0.95 ± 0.4	12.22 ± 5.9	0.19	22.3 ± 14.0	0.7 ± 0.3	−15.3 ± 0.2	7.4 ± 0.2
ESO 308-G022	0.92 ± 0.07	20.87 ± 0.3	0.71 ± 0.2	10.76 ± 5.2	0.53	38.8 ± 22.8	1.4 ± 0.5	−16.7 ± 0.2	8.0 ± 0.2
KKS 2000-09	0.26 ± 0.03	14.12 ± 4.6	0.42 ± 0.2	0.28 ± 1.8	0.06	9.7 ± 63.4	0.3 ± 1.7	−17.9 ± 0.2	8.5 ± 0.2
Argo	0.89 ± 0.04	21.87 ± 0.1	1.32 ± 0.2	48.77 ± 4.6	0.27	55.4 ± 7.7	1.3 ± 0.1	−16.1 ± 0.2	7.8 ± 0.2
ESO 059-G001	0.76 ± 0.04	20.23 ± 0.2	1.33 ± 0.2	36.68 ± 6.2	0.1	41.3 ± 8.3	0.9 ± 0.2	−17.2 ± 0.2	8.2 ± 0.2
AM 0737-691	2.5 ± 0.3	20.67 ± 0.2	1.07 ± 0.2	21.41 ± 3.9	0.1	32.0 ± 7.1	2.5 ± 0.6	−19.0 ± 0.3	9.0 ± 0.2
KK 2000-25	0.03 ± 0.01	18.13 ± 0.08	1.12 ± 0.03	9.41 ± 0.5	0.03	13.1 ± 0.7	0.03 ± 0.01	−11.9 ± 0.2	6.1 ± 0.2
ESO 006-G001	0.65 ± 0.04	19.20 ± 0.5	1.07 ± 0.2	16.11 ± 5.2	0.08	24.1 ± 8.9	0.8 ± 0.2	−17.8 ± 0.2	8.5 ± 0.2
UGCA 148	0.90 ± 0.03	19.85 ± 0.07	1.37 ± 0.09	22.15 ± 1.3	0.05	24.2 ± 1.9	1.15 ± 0.07	−18.0 ± 0.2	8.5 ± 0.2
NGC 2784	1.3 ± 0.1	15.26 ± 0.4	0.82 ± 0.1	26.85 ± 9.0	0.04	66.6 ± 24.4	3.2 ± 0.5	−24.0 ± 0.2	10.9 ± 0.2
KK 98-73	0.62 ± 0.04	19.48 ± 0.3	0.95 ± 0.1	10.74 ± 2.2	0.36	19.6 ± 4.8	0.9 ± 0.1	−17.4 ± 0.2	8.3 ± 0.2
UGCA 153	1.0 ± 0.1	20.93 ± 0.2	1.15 ± 0.1	33.31 ± 5.2	0.11	44.8 ± 8.5	1.5 ± 0.2	−16.7 ± 0.3	8.0 ± 0.2
NGC 2835	3.5 ± 0.4	16.50 ± 0.04	0.52 ± 0.01	14.27 ± 0.8	0.54	160.0 ± 10.7	6.4 ± 0.2	−23.1 ± 0.3	10.6 ± 0.2
UGCA 162	0.77 ± 0.05	20.22 ± 0.02	1.35 ± 0.02	47.70 ± 0.7	0.32	52.9 ± 1.0	2.00 ± 0.02	−17.3 ± 0.2	8.3 ± 0.2
ESO 565-G003	0.60 ± 0.03	19.97 ± 0.5	1.01 ± 0.2	12.69 ± 4.1	0.05	20.8 ± 7.6	0.8 ± 0.2	−16.7 ± 0.2	8.0 ± 0.2
NGC 2915	0.39 ± 0.02	15.97 ± 0.2	0.63 ± 0.03	5.21 ± 0.8	0.03	27.4 ± 4.8	0.50 ± 0.07	−18.6 ± 0.2	8.8 ± 0.2
NGC 3115	1.3 ± 0.1	10.20 ± 0.4	0.31 ± 0.02	0.18 ± 0.1	0.14	62.1 ± 43.4	3.0 ± 0.9	−24.4 ± 0.2	11.1 ± 0.2
IC 4662	0.39 ± 0.02	17.27 ± 0.2	0.94 ± 0.05	22.68 ± 2.9	0.03	42.3 ± 5.8	0.50 ± 0.06	−18.3 ± 0.2	8.7 ± 0.2
ESO 461-G036	0.56 ± 0.08	21.16 ± 0.2	1.80 ± 0.4	21.07 ± 2.8	0.14	18.3 ± 3.4	0.7 ± 0.1	−15.9 ± 0.3	7.7 ± 0.2
DDO 210	0.21 ± 0.02	22.07 ± 0.08	1.52 ± 0.2	64.49 ± 4.7	0.27	63.8 ± 6.4	0.29 ± 0.03	−12.9 ± 0.2	6.5 ± 0.2
IC 5052	1.61 ± 0.06	18.07 ± 0.02	1.27 ± 0.02	88.82 ± 1.4	0.28	105.4 ± 2.1	3.08 ± 0.04	−20.3 ± 0.2	9.5 ± 0.2
IC 5152	0.54 ± 0.03	17.65 ± 0.08	0.96 ± 0.04	35.44 ± 2.7	0.01	63.5 ± 5.5	0.64 ± 0.05	−18.3 ± 0.2	8.7 ± 0.2
ESO 468-G020	0.41 ± 0.02	21.13 ± 0.07	1.28 ± 0.1	24.93 ± 1.9	0.13	29.3 ± 3.0	0.55 ± 0.04	−14.9 ± 0.2	7.3 ± 0.2
UGCA 438	0.39 ± 0.03	20.49 ± 0.2	1.31 ± 0.08	38.34 ± 3.4	0.03	43.9 ± 4.4	0.47 ± 0.04	−15.6 ± 0.2	7.6 ± 0.2
IC 5332	3.19 ± 0.08	16.92 ± 0.2	0.52 ± 0.04	8.51 ± 2.4	0.21	95.4 ± 30.5	4.4 ± 1.0	−22.0 ± 0.2	10.1 ± 0.2
ESO 347-G017	1.00 ± 0.07	18.19 ± 0.4	0.74 ± 0.08	9.02 ± 2.9	0.22	29.0 ± 10.3	1.3 ± 0.4	−18.3 ± 0.2	8.7 ± 0.2
NGC 7713	2.0 ± 0.1	16.98 ± 0.2	0.83 ± 0.06	24.46 ± 3.8	0.02	59.0 ± 10.6	2.7 ± 0.4	−21.4 ± 0.2	9.9 ± 0.2
UGCA 442	0.67 ± 0.04	19.94 ± 0.03	1.42 ± 0.04	57.38 ± 1.5	0.44	60.4 ± 2.0	1.25 ± 0.03	−17.3 ± 0.2	8.3 ± 0.2
ESO 348-G009	1.28 ± 0.07	21.26 ± 0.05	1.31 ± 0.07	44.50 ± 2.0	0.15	51.0 ± 3.1	2.14 ± 0.08	−17.2 ± 0.2	8.2 ± 0.2
NGC 7793	1.7 ± 0.1	16.47 ± 0.08	0.62 ± 0.02	28.72 ± 3.0	0.46	159.4 ± 18.7	3.0 ± 0.2	−21.9 ± 0.2	10.1 ± 0.2

Download table as:  ASCIITypeset image

The corrected total apparent magnitude, m_H = m_H,obs − Δm, was converted into a luminosity by using the standard equations

$$\begin{eqnarray} &&M_{H,0}= m_{H}-5\log _{10} D - 25 - A_{H} \end{eqnarray} \tag{ 4 }$$

and

$$\begin{eqnarray} &&L=10^{0.4(M_{H,\odot }-M_{H,0})}, \end{eqnarray} \tag{ 5 }$$

where M_H,☉ = 3.35 mag is the H-band luminosity of the Sun (Colina et al. 1996) and A_H = 0.576 · E(B − V) is the Galactic extinction (Schlegel et al. 1998). The accuracy of the reddening-corrected absolute magnitude, M_H,0, is dominated by the accuracy of the distance.

The effective radius, r_eff, is the aperture radius that encloses half of the total light of a galaxy. This quantity is systematically underestimated unless corrected for the amount of undetected flux. The corrected effective radius, R_eff, is implicitly defined by the equation

$$\begin{equation} \frac{I_{\rm tot}}{2} \,=\,\frac{2\pi I_0 \alpha ^2}{n}\cdot \Gamma \left[\frac{2}{n}, \left(\frac{R_{\rm eff}}{\alpha }\right)^{n}\right], \end{equation} \tag{ 6 }$$

where I₀, α, and n are the Sérsic parameters for a particular galaxy. The solution for all n ∈ [0.25, 2] can be approximated by the equation

$$\begin{equation} R_{\rm eff} \approx \alpha \left(\frac{2}{n} -0.33211\right)^{1/n}. \end{equation} \tag{ 7 }$$

Within the quoted uncertainties, there is no deviation between the approximate analytical solution (Equation 7) and the numerical solution of Equation (6).

The H-band luminosity of each galaxy was converted into a stellar mass using a mass-to-light ratio of ϒ^H_* = 1.0 ± 0.4. This conversion factor is discussed in detail in Section 4.2. The derived parameters are listed in Table 4 for the 57 LSI galaxies: Column (1)—galaxy name; Column (2)—effective radius, r_eff, in kpc; Columns (3)–(5)—the Sérsic parameters μ₀, n, and α, respectively; Column (6)—the missing flux, Δm; Columns (7) and (8)—the corrected effective radius, R_eff, in arcsec and kpc, respectively; Column (9)—absolute H magnitude, M_H,0; and Column (10)—the total stellar mass, $\log _{10}{\mathcal{M}}_{\ast }$.

4.1. Galactic Extinction Correction

4.2. The H-Band Mass-to-Light Ratio

Observations for 11 galaxies were not included in our photometric study because either the galaxy remained invisible in the final mosaics, despite our faint H-band surface brightness limit of 24–26 mag arcsec⁻², or the galaxy was detected, but foreground stars interfered with the analysis. In this section, we discuss the four galaxies labeled as "no galaxy detected" in Table 2: AM 0717-571, KK 2000-04, KK 2000-06, and NGC 2784 DW1. While they remain candidates for galaxies with a pure young stellar component, we show it is unlikely that this is the case. We also include KK 2000-03, which had a marginal detection, and the galaxy pair HIZOAJ1616-55 and SJK98 J1616-55, which were not detected, but the images had serious foreground contamination.

A lower bound of the total apparent magnitude for these galaxies can be calculated. For this purpose, we consider a hypothetical galaxy with a constant star density equivalent to the survey's mean surface brightness limit of 〈μ_0,lim〉 = 25 mag arcsec⁻² out to a cutoff radius, r_cut (at which point the stellar density drops to zero). We set the cutoff radius, r_cut, to be 250 arcsec, which is equivalent to the size of the largest galaxies in our sample. This yields the brightest apparent magnitude that an undetected galaxy could possibly have:

$$\begin{equation} m_{\rm tot}> \mu _{0,{\rm lim}} -2.5\log _{10} \big(\pi r_{\rm cut}^2\big)=11.8\;{\rm mag}. \end{equation} \tag{ 8 }$$

This lower bound is applicable to AM 0717-571, KK 2000-04, KK 2000-06, and NGC 2784 DW1, but not KK 2000-03, which has a foreground star located directly in front of the galaxy (see discussion below), nor HIZOAJ 1616-55 and SJK98 J1616-55, which have serious foreground contamination. A lower bound on the absolute magnitude and an upper bound on the stellar mass is calculated by using distances from the literature (see Table 2).

5.1. AM 0717-571

5.2. HIZOAJ 1616-55 and SJK98 J1616-55

5.3. KK 2000-03

5.4. KK 2000-04 and KK 2000-06

5.5. NGC 2784 DW1

6.1. LSI Survey Versus 2MASS Photometry

It is instructive to see how galaxies can change their appearance when going from 2MASS to the deeper LSI observations. For example, our image of the barred Sc galaxy NGC 2835 (Figure 12) reveals the rich NIR morphology and the extent of this spiral galaxy for the first time. The almost face-on view presents a well-ordered four- or five-arm spiral pattern outlined by the star-dominated Population II disk, which closely traces the gas-dominated Population I disk morphology observed in the B band (Sandage & Bedke 1994). The 2MASS image for the irregular Sculptor group galaxy ESO 473-G024 qualitatively shows the limitation of that survey in the study of dwarf galaxies. With a central H-band surface brightness of ≈20.5 mag arcsec⁻², ESO 473-G024 remains effectively undetected in 2MASS (Figure 12). Our image uncovers a smooth, dE-like morphology with little evidence of irregularity. This stands in stark contrast to the B-band image that is dominated by a number of prominent H ii regions and dust features.

Zoom In Zoom Out Reset image size

It has been previously pointed out by Andreon (2002) that the short integration time of 2MASS failed to detect most of the lower surface brightness (dwarf) galaxies and that, if they were detected, fluxes were underestimated by as much as 70%. To investigate this issue further, in Figure 13 we plot the difference between our total extrapolated apparent magnitudes (m_H,obs − Δm; Table 3, Column 2 and Table 4, Column 6) and the total magnitudes from the 2MASS All-Sky Extended Source Catalog for the 21 galaxies that we have in common, as a function of the mean effective surface brightness (〈μ_H〉_eff; Table 3, Column 4). LSI galaxies with a surface brightness fainter than μ_H = 18 mag arcsec⁻² are affected at different levels and have missing flux in the range between 0.2 and 2.5 mag. Even in the cases of the luminous galaxies NGC 2784 and NGC 3115, our analysis finds that their 2MASS H-band magnitudes are 0.5 mag too faint.

Zoom In Zoom Out Reset image size

To demonstrate that the disparity between the 2MASS and our H-band magnitudes is not caused by differences in the measuring procedure, we analyzed 2MASS images by using our method. The photometric parameters listed in the 2MASS Extended Source Catalog, as well as the surface brightness profiles, were reproduced within the quoted uncertainties. It should be noted that the 2MASS Large Galaxy Atlas (Jarrett et al. 2003) has employed two different methods for recovering the flux below the sky background noise, and that we have made the comparison to the magnitudes that were obtained by extrapolating the surface brightness profile. The 2MASS magnitudes, which were obtained by using Kron (1980) apertures, are systematically fainter (see Figure 9 of Jarrett et al. 2003), and therefore the difference between our total apparent magnitudes and the 2MASS magnitudes obtained by using Kron apertures is also larger.

The Extended Source Catalog is contaminated by a small (1–5%) number of artifacts, which can significantly affect the photometry of real sources (Jarrett et al. 2000). Each of our galaxy images was visually inspected for artifacts, which were effectively removed (unless documented otherwise). This is obviously an impractical approach for the much larger 2MASS data set, and Jarrett et al. (2000) noted that the pipeline is not 100% effective. It is conceivable that this accounts for a fraction of the discrepancy.

In summary, our finding is in good agreement with that of Andreon (2002) and, again, emphasizes that the 2MASS magnitudes are significantly fainter than those obtained from deeper NIR imaging. As the mean effective surface brightness correlates with the luminosity of a galaxy (see Section 6.3), serious selection biases must be expected, for instance, for the 2MASS-based H-band galaxy luminosity function at magnitudes fainter than M_H = −20 mag.

6.2. Optical–NIR Magnitude Transformation

The B − H color of each galaxy is an indicator of the ratio of the Population II to Population I stars, as modulated by the effects of dust. A comparison of the absolute B- and H-band magnitudes, corrected for Galactic extinction, is shown in Figure 14 for our sample galaxies. We also plot the Virgo cluster data from the GOLD Mine database (Gavazzi et al. 2003). The data of the Virgo galaxies were extinction-corrected by using A_B = 0.13 mag and A_H = 0.01 mag. We have adopted the mean cluster distance of 15.8 Mpc (Jerjen et al. 2004), based on surface brightness fluctuation measurements of early-type galaxies. In addition to the Virgo cluster data, we have also included the data for 30 bright spiral galaxies from Kassin et al. (2006), which were corrected for extinction (using Schlegel et al. 1998). By including the two additional data sets, we are able to investigate the B − H color for late-type galaxies over a range of 15 mag. Figure 14 shows that there is a tight correlation between the B- and H-band luminosities of a galaxy. This linear relation is

$$\begin{equation} M_{H,0}=(1.14\pm 0.02)M_{B,0}-(0.74\pm 0.32), \end{equation} \tag{ 9 }$$

which is a least-squares fit to the Virgo cluster data. While the more luminous galaxies in our sample obey the relation closely, the residual plot suggests that the scatter marginally increases in the dwarf regime and possibly has a slight upwards trend to redder colors (average residual ≈1 mag). The most deviant galaxy in our sample, KK 2000-25, appears underluminous in the B band by 3 mag. While we cannot exclude the possibility that this galaxy has had an unusual star-formation history, we need to point out two things. First, KK 2000-25 is located almost in the Galactic plane (b = 1.28°) and thus has a large B-band extinction uncertainty. Second, the B-band magnitude for KK 2000-25 is visually estimated from a photographic film (Huchtmeier et al. 2001). Consequently, the deviation from the line of best fit could be entirely due to a large uncertainty in the B-band magnitude. By excluding KK 2000-25, there is no correlation between the residual in the B − H plot and the galaxy distance, Galactic latitude, or mean effective surface brightness (see Figure 15).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

It is worth noting that the least-squares fit deviates from a line of unity slope. The gradient of 1.14 ± 0.02 implies that dwarf galaxies are, in general, bluer than the more luminous galaxies. It is well known that galaxy color correlates with luminosity (e.g., Tully et al. 1998; Hogg et al. 2002; Blanton et al. 2001, 2003).

Figure 14 gives a useful indication of the stellar population of galaxies. Galaxies that lie well below the line are bluer than most galaxies, which suggests that they have a relatively young stellar population. Conversely, galaxies that lie well above the line are redder than expected, indicating a larger old stellar population. The tight correlation (correlation coefficient = 0.97) between the B- and H-band luminosities comes somewhat as a surprise. A B-band light profile of a galaxy can be significantly attenuated and distorted by dust. Moreover, short-lived giant O and B stars contribute to the B-band emission, and, hence, the profile can be distorted by transient star-formation events. The stellar mass of most galaxies is dominated by the older, low-luminosity stellar population, whose energy output peaks at NIR wavelengths (Gavazzi et al. 1996a). Hence, it has been argued that the NIR is the optimal wavelength regime for investigations of structural properties (Driver 2004). The tight correlation between the B and H bands, however, suggests that the advantages of the H band may not be as significant as previously thought, at least for late-type giant galaxies. A detailed comparison of the observed scatter with the predictions from population synthesis models (e.g., Bruzual & Charlot 1993, 2003; Maraston 1998; Li & Han 2008) is beyond the scope of this paper because of the wide range of stellar compositions and star-formation histories represented by our galaxy sample.

The B − H color for each galaxy can be compared with the morphological type (Figure 16). There we include the combined samples of Kassin et al. (2006) and the Virgo cluster data in the white boxes and the new LSI data are shown by the black boxes. The combined Virgo cluster and Kassin et al. sample is ten times the size of our sample, but is dominated by giant, luminous galaxies. Our sample, in contrast, contains four times as many irregular dwarf galaxies. Therefore, when interpreting the color versus morphology plot, it must be noted that our sample dominates the morphological bin of irregular galaxies and the literature data dominate the larger galaxies.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

The comparison of morphology to the B − H color shows that the early-type galaxies are redder than the late-type, irregular, and dwarf galaxies. A similar study by Jarrett et al. (2003) for galaxies in the 2MASS Large Galaxy Atlas also showed this trend.

6.3. Luminosity–Surface Brightness Relation

In Figure 17, we plot the mean effective surface brightness of our sample galaxies as a function of absolute magnitude. In addition to our H-band data, we include 560 late-type Virgo cluster galaxies (obtained from the GOLD Mine database; Gavazzi et al. 2003). The mean effective surface brightness for Virgo cluster galaxies was calculated as

$$\begin{equation} \langle \mu _{H}\rangle _{\rm eff}=M_H + 2.5\log _{10}\!\big(\pi r_{\rm eff}^2\big) \end{equation} \tag{ 10 }$$

and the data were corrected for extinction by using A_H = 0.01 mag. The morphologies of Virgo cluster galaxies included in the sample range from S0 to Sd, Irr, and BCD (listed as types 1 to 18 in the GOLD Mine database).

As previously discussed in de Jong & Lacey (2000), the relationship between the two photometric parameters provides an important link to the underlying physical parameters of a galaxy, namely its total mass ${\cal M}_{\rm tot}$ and total angular momentum. The total angular momentum of a galaxy, expressed as the dimensionless spin parameter $\lambda =J|E|^{1/2}{\cal M}_{\rm tot}^{-5/2}G^{-1}$ (Peebles 1969), is related to the scale length of its disk (Fall & Efstathiou 1980; Dalcanton et al. 1997; Mao et al. 1998). It has been shown by de Jong & Lacey (2000) that λ can be transformed into observable quantities. The authors presented a model of a singular isothermal sphere with $E\propto {\cal M}_{\rm tot}V_c^2$ from the virial theorem and a perfect exponential disk with angular momentum $J_{\rm disk}\propto {\cal M}_{\rm disk}r_{\rm eff}V_c$ (assuming V_disk = V_c). They showed that if $J_{\rm disk}/{\cal M}_{\rm disk}\propto J/{\cal M}_{\rm tot}$ and ${\cal M}_{\rm disk}\propto {\cal M}_{\rm tot}$, then $\lambda \propto r_{\rm eff}V_c^2\big/{\cal M}_{\rm disk}$. Furthermore, by using the relation ${\cal M}_{\rm tot}\propto V_c^3$, which is predicted for dark matter halos, de Jong & Lacey showed that λ ∝ r_effL^ϒ/3, where ϒ is the mass-to-light ratio for the disk. This can be transformed into an expression between the surface brightness and total magnitude by invoking

$$\begin{equation} \Sigma _{\rm eff}=L\big(2\pi r_{\rm eff}^2\big)^{-1}, \end{equation} \tag{ 11 }$$

which results in

$$\begin{equation} \lambda \propto \Sigma _{\rm eff}^{-1/2} L^{-\Upsilon /3 + 1/2}. \end{equation} \tag{ 12 }$$

By using the two identities −2.5log₁₀Σ_eff = 〈μ_H〉_eff + c and −2.5log₁₀L = M + c' (c and c' being constants), we finally arrive at the theoretical luminosity–surface brightness relation:

$$\begin{equation} \langle \mu \rangle _{\rm eff} = (1-2\Upsilon /3) M +5\log _{10}\lambda +C. \end{equation} \tag{ 13 }$$

Here, we can see that the gradient of the luminosity–surface brightness relation is a function of the galaxy mass-to-light ratio and that the dispersion in the empirical relation reflects the distribution of the spin parameter.

The empirical relation between the two observational quantities in Figure 17 for the late-type galaxies of the Virgo cluster sample is analytically best described by the linear equation

$$\begin{equation} \langle \mu _{H}\rangle _{\rm eff}=a\cdot M_{H,0} +b, \end{equation} \tag{ 14 }$$

where a is 0.47 ± 0.08, 0.44 ± 0.06, and 0.48 ± 0.05 for our LSI sample, the Virgo cluster sample, and the combined LSI + Virgo samples, respectively, and b is 29.0 ±  1.4, 28.0 ±  1.3, and 28.8 ±  1.1. The uncertainty in a and b is the formal uncertainty in the linear fit plus the uncertainty due to the robustness of the sample obtained by using bootstrap resampling. In Figure 17, the empirical relation between the mean effective surface brightness and the absolute magnitude is plotted for the combined LSI + Virgo cluster samples.

A comparison of the empirical result (see Equation 14) with the theoretical prediction (Equation 13) allows us to estimate ϒ^H_*: a = (1 − 2ϒ/3) ⇒ ϒ^H_* = 0.80 ±  0.12, 0.84 ±  0.09, and 0.78 ±  0.08 for our LSI sample, the Virgo cluster sample, and the combined data sets, respectively. These results are in excellent agreement with the value of 1.0 ±  0.4 adopted in section 4.2.

7.1. KKS 2000-09

7.2. HIZSS003

7.3. KK 2000-25

We have presented the deepest H-band images available to date for 57 galaxies in the LSI (D < 10 Mpc), obtained by using the NIR camera IRIS2 at the 3.9 m AAT. Of the 68 targets, 11 either remained undetected or could not be usefully analyzed due to contamination by foreground stars. The surface brightness limit reaches down to μ_lim < 26 mag arcsec⁻², 4 mag arcsec⁻² fainter than 2MASS.

The images, cleaned of Galactic foreground contamination, reveal the morphology and extent of many of the galaxies for the first time. For 56 galaxies, we derive radial luminosity profiles, ellipticities, and position angles, together with global parameters, such as total magnitude, mean effective surface brightness, half-light radius, Sérsic parameters, and stellar mass.

No genuine young galaxies have been found in this survey. Some sample galaxies were previously identified on B-band photographic plates but remained undetected in the NIR. In each case, there is a plausible alternative explanation for the non-detection—(1) AM 0717-571: DSS B_J-band morphology resembles that of a Galactic nebula, but true nature still remains unclear; (2) HIZOAJ 1616-55 and SJK98 J1616-55: possibly one or two high velocity clouds; (3) KK 2000-03: superimposed star hampers analysis; however, the marginal detection in the H band suggests an unusual blue galaxy; (4) KK 2000-04: originally assumed to be a companion of NGC 1313, however, possibly a photographic plate flaw; (5) KK 2000-06: originally assumed to be a companion of NGC 1313, but more likely a background galaxy at ≈2250 km s⁻¹; (6) NGC 2784 DW1: intrinsic, extreme low surface brightness dwarf satellite of NGC 2784.

We also detected a double nucleus in KKS 2000-09 and proposed to reclassify this system as a peculiar galaxy. KKS 2000-25 was shown to have distinct spiral arms in the H band and thus should be classified as "Sb." Morphology and angular size strongly suggest that this is a background galaxy beyond 10 Mpc.

We found compelling evidence that the short integration time of 2MASS resulted in a serious underestimation of a galaxy's luminosity. The magnitudes of galaxies, with an H-band surface brightnesses fainter than 18 mag arcsec⁻², obtained in our study are up to 2.5 mag brighter than those obtained by 2MASS. As the mean effective surface brightness correlates with the luminosity of a galaxy, we expect a serious selection biases for a 2MASS-based a H-band galaxy luminosity function fainter than M_H = −20 mag.

There is a tight correlation (correlation coefficient = 0.97) between the B- and H-band magnitudes of a galaxy, and this correlation has been demonstrated over a range of 15 mag. The linear transformation between the B and H bands has a small scatter (0.3 mag) for bright galaxies. In the dwarf regime, there is a marginal increase in scatter and possibly a slight trend for galaxies to be redder (by approximately 1 mag) than indicated by the transformation found for bright galaxies.

The galaxy luminosity–mean effective surface brightness relation has been analyzed to derive a semiempirical stellar mass-to-light ratio of ϒ^H_* = 0.78 ±  0.08 in the H band.

All raw and reduced H-band images of the 57 program galaxies in this NIR survey will be made publicly available and can be obtained via e-mail request.

We thank the referee for the useful comments. The authors acknowledge financial support from the Australian Research Council Discovery Project Grant DP0451426. This paper is based on data obtained with the AAT. The study made use of data products from the 2MASS, which is a joint project of the University of Massachusetts and the Infrared Processing and Analysis Center at the California Institute of Technology, funded by the National Aeronautics and Space Administration and the National Science Foundation. Support for IRIS2 data reduction within ORAC-DR is provided by the Joint Astronomy Centre. This research has made use of the GOLD Mine Database. This research has made use of the NED, which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. This research has made use of NASA's Astrophysics Data System.