Astron. Astrophys. 323, 323-336 (1997)

2. Optical data

2.1. The observations

CCD images were obtained at the Capilla Peak Observatory's ¹ 0.6 m f/15.1 telescope. The telescope was equipped with a CCD camera with a RCA chip (Laubscher et al. 1988), and broadband filters which are closely matched (Beckert and Newberry 1989) to the Johnson UBV (Johnson 1955) and the Kron-Cousins VRI system (Cousins 1976, Bessell 1979). The plate scale of the CCD was 0.658 arcseconds per pixel; the CCD read-out noise was about 80e per pixel. All galaxy images were taken with a pre-flash, which primes the CCD with about 80 counts/pixel, to suppress a minor nonlinearity of the CCD at low count rates.

The goal was to image all galaxies in the sample through the B, R and I filters, with three exposures of 600 seconds for each filter at photometric conditions. This observing program was completed with a few exceptions, as can be seen in Table 1. The seeing on these images varies between 1.2 and 1.7 arcseconds. Each night, at least two Landolt (1973, 1983) stars were observed several times for intensity calibration and derivation of atmospheric extinction coefficients and color terms. Between each exposure, the telescope was moved by a small fraction of the CCD's field of view in order to position bad columns and hot pixels of the CCD chip at slightly different places on each image.

Table 1. Overview of the broadband observations at Capilla Peak

All of the galaxies were observed at Capilla Peak with an H filter set, with filters centered on the H emission line (the 'on' filter), and on the nearby continuum (the 'off' filter). The H filters at Capilla Peak have a FWHM of about 75 and are spaced about 80 apart. The small overlap of these filters led to worst case transmissivities of slightly over 40 for some galaxies. Some of the galaxies were also imaged at the Kitt Peak ² 0.9m telescope in its f/7.5 configuration in January 1992, but weather conditions allowed the taking of a few images only. An overview of the H observations is given in Table 2. The exposure times given in the table are for either the 'on' or the 'off' filter, which was the same in all cases. When several exposures were taken, 'off' and 'on' filtered exposures were alternated to minimize variations in the airmass and seeing between 'off' and 'on' filtered images. Standard stars for the observations at Capilla were Cyg OB9 #2 and G191 B2B. These two stars are from the list of spectrophotometric standards by Massey et al. (1988) and were selected for the relative absence of spectral features over the wavelengths covered by the H filters.

Table 2. Overview of the H observations

2.2. Broadband data reductions

Image processing was performed using the standard procedures available in the IRAF package. Details of the photometric reductions can be found in Deeg (1992) for the broadband observations and in Deeg (1993) for the H observations. Astrometric positions of the galaxy centers were obtained on the B band images with the 'Mann'-measuring engine at the NRAO-AOC in Socorro. These positions were used to create overlays with the radio maps with an accuracy of about 1".

Color terms for the filter-camera system at Capilla Peak were determined from separate observing runs with large numbers of Landolt stars in 1989 and 1991 (Beckert, 1991). Extinction coefficients were fitted for each night based on nightly observations of Landolt stars, using the program 'Photom', written by D. Beckert at the University of New Mexico.

Instrumental magnitudes for each galaxy were determined within a series of elliptical apertures. The ellipses were created by isophotal fitting (Jedrzejewski 1987) to the R band images with the 'ellipse' task in the STSDAS package. The sets of elliptical apertures derived from the R band image (which shows the greatest extent of the galaxies) were then used to obtain the instrumental magnitudes of the galaxies in each filter. Transformation to surface brightnesses allows derivation of the diameter , which is the galaxies' major isophotal diameter extending to a surface brightness of 25 arcsec . The B band magnitude within that diameter, , and the color coefficients and were then obtained. From the same sets of elliptical apertures, the "equivalent" effective diameter , as defined in the RC3 (de Vaucouleurs et al., 1991) was derived. The effective aperture is defined as a circular aperture through which passes the flux , which is half of a galaxies' total B band flux. The colors , and magnitudes within these effective apertures were also obtained.

Extinction corrections for the color excess from the Galaxy, , were applied to the measured results, based on the galactic B band extinction . The RC3 gives values for from Burstein and Heiles (1982, 1984). With (Huchra 1977, Whitford 1958) and , the extinction correction is for colors. Similarly, from the average interstellar extinction curve by Savage and Mathis (1979), one obtains for colors: . is 0.3 for all galaxies, except for II Zw 40, where the RC3 indicates a large, and uncertain value of .

The internal B band extinction of the galaxies, , is listed in the RC3 for II Zw 70, Mkn 297, Mkn 527 and III Zw 102. The internal color excesses and were derived from using the relations: and , which are similar to those derived in the RC2 (de Vaucouleurs, de Vaucouleurs, and Corwin, 1976). itself was calculated from Eqs. (63) and (64) in the RC3 and depends on the ratio of major to minor diameters, , and on the morphological 'T' classification, both taken from the RC3. Our photometric results were not used for the determination of , as the ellipse fitting algorithm could not find the ellipticity reliably for isophotes around the fainter outer parts of the galaxies. The internal color excesses were calculated for those galaxies for which was listed in the RC3. They are rather negligible with with the exception of II Zw 70, where has an uncertain value of 0.18, due to an uncertain 'T' classification. Internal extinction corrections were not included in the calculation of the corrected color index , but are listed separately in Sect. 2.3. No redshift corrections were made to the color indices. The maximum error introduced by this is less than 0.07 mag for the colors, with the exception of Haro 15, where it might be up to 0.15 mag, as this galaxy is more distant than the others. Within the R band lies the H emission at , and estimates were made on the fraction of the R band flux which is due to H emission. The H flux was found to contribute less than 10%, with the exception of II Zw 40, where H contributes about 40% to its nucleus' R band flux. Hence, 'H corrected' colors of II Zw 40's nucleus would be bluer by about 0.55 mag.

2.3. Broadband results

The morphology of the galaxies is discussed in Sect. 4. Here, the quantitative results of the photometry are given; they are listed in Table 3a. Uncertain values are indicated by colons. The parameters in Table 3a are:

: apparent major isophotal diameter measured at a surface brightness level of arcsec^-2, and the effective diameter, as introduced in the previous section.
: B band magnitude measured within an elliptical aperture with .
: B band surface brightness in magnitudes per square arcsec within the effective aperture .
: Galactic extinction in B band magnitudes, from the RC3.
() : color excess due to extinction in the Milky Way.
() : color excess internal to the sample galaxy.
: color index, corrected for galactic extinction, not corrected for internal differential extinction or redshift, within an aperture of about size, centered at the maximum of the R band surface brightness ('center'). Several of the galaxies do not have a well defined nucleus; and the color indices at the 'center' indicate a somewhat arbitrary region close to the geometric center of the galaxy; these cases are indicated by parentheses.
: ditto, within the effective aperture .
: ditto, within an elliptic aperture with major diameter .
colors, within the same apertures as the values, corrected for galactic extinction only.

Table 3a. Photometry from Capilla Peak observations

For comparison, Table 3b shows photometric results taken from the literature. Unless indicated, they are all taken from the RC3. The parameters in Table 3b are as follows:

: As in Table 3a.
: Total (asymptotic) B band magnitude
: Total corrected magnitude. For II Zw 40 it was calculated using the prescriptions in the RC3, Sect. 3.6.d
: As in Table 3a.
: Galactic extinction in B band magnitudes
: The ratio between the major isophotal diameters , , measured or reduced to a B -band surface brightness level of 25 mag arcsec^-2.
: color index, corrected for galactic extinction, internal extinction, and redshift, within the effective aperture . The RC3 lists only the uncorrected index ; was derived using the difference between the corrected and uncorrected total color index which are both given in the RC3, hence: . The validity of this procedure is indicated in the explanatory Section of the RC3.
: color index, corrected for galactic extinction, internal extinction and redshift, within effective aperture
: ditto, total (asymptotic) color index
: color index, not corrected, within effective aperture

Table 3b. Photometric results from the literature

The parameters and are listed in both tables and are given to allow a direct comparison between the new photometric results and values in the literature. The RC3 was chosen as the preferred reference, as it is the largest and most homogeneous compilation of extragalactic photometric data. The RC3 quotes the B band brightness as the total (asymptotic) magnitude , which can only be obtained by extrapolation from magnitudes at detectable surface brightnesses. Discounting our results for II Zw 40 (its is fairly meaningless due to its shape) and II Zw 70 (data taken during poor sky conditions), the offset between , measured from the Capilla Peak data, and , from the RC3 is:

The small scatter in the offset indicates that the photometry taken at Capilla Peak is reliable and was performed in a consistent fashion; the offset itself is mostly a result of the intrinsic difference between and . This difference was included in the derivation of the effective magnitude (which defines ) from by using: . Whereas the measured values for all lie within the error limits given by the RC3, the same cannot be said for the results based on the effective aperture, . The relation given in the RC3 for is valid for normal spiral galaxies only, and the effects of the irregular shapes of the sample galaxies are more pronounced at the smaller diameter than at . No systematic offset can be found between and . Using the five galaxies for which is given in the RC3, it is: arcsec. It should be noted, that in the RC3 has been derived using standard growth curves from measurements at various apertures. For irregular galaxies, the growth curves are certainly not well suited to standardization, and the formal errors quoted in the RC3 might be too small. The same argument extends of course to parameters derived from , particularly the surface brightness . The color indices within the effective apertures are only weakly affected by the uncertainty of , as the color indices do not significantly vary over aperture changes of about 2 arcsecs. The values of the measured colors are not easily verifiable, as independent colors could be found only for two galaxies in Thuan & Martin (1981). No references to colors involving I band measurements could be found in the literature. In the absence of a comparison, the uncertainty of the measured and color indices can only be crudely estimated to be about . As an example of the scatter between published color indices, the compilation by Brosch & Loinger (1991) of colors of III Zw 102 from various sources in the literature is enlightening - differences of 0.5 mag at comparable apertures seem to be common.

2.4. H data reduction

The relative wide band of the H filters of about resulted in substantial continuum emission in the H images. To remove the continuum emission, a set of 'off' images centered on the nearby continuum was taken. The 'off' images were scaled so that foreground stars in corresponding H and 'off' images are of the same brightness. The 'off' image was then subtracted from the H image, leaving only the H line emission.

The images taken at Capilla Peak were calibrated with images of the spectrophotometric standard stars Cyg OB9 #2 and G191 B2B (Massey et al., 1988), taken through the H filter set at various airmasses. Details of the calibration procedure are given in Deeg (1993). The observations taken at Kitt Peak, which were under marginally photometric conditions, were roughly calibrated by observations of I Zw 18 and NGC 2363, which have known H fluxes of erg s^-1 cm^-2 for I Zw 18 (Davidson et al. 1989, for a square aperture of centered on the nucleus) and erg s cm^-2 for NGC 2363 (Kennicutt et al. 1980; largest H II region of NGC 2363). The instrumental magnitudes of the reference objects were obtained by using, as closely as possible, the same apertures as given in the literature.

2.5. Results of the H observations

Here, only the results of the photometry are derived. Table 4 lists the observed H [N II ] fluxes, , and the aperture in which they were measured. Considering the marginal weather conditions during the Kitt Peak observing run and consequent poor internal consistency, the H fluxes from the Capilla Peak and the Kitt Peak observations were weighted with a ratio of 4:1. Uncertainties in the flux values are estimated to be about 20%, except for II Zw 70, where the uncertainty is 50%, as none of its observations was under photometric conditions.

Table 4. H photometry

To derive the H luminosities intrinsic to the emission line regions, the observed H [N II ] fluxes are subject to corrections for blending with [N II ] line emission, for galactic extinction, and for internal extinction of the emitting H II regions. Corrections for [N II ] line blending were taken from [N II ]/H flux ratios reported in the literature, given in Table 4. If several literature sources quoted different values, averages were taken. The correction for galactic extinction of the emitting H line, , is based on the differential galactic extinction, , which has been introduced in Sect. 2.2. Based on Caplan and Deharveng (1986) and using the galactic interstellar extinction curve of Savage and Mathis (1979), one obtains for in units of magnitude: . The H II regions' internal extinction, , is derived from the measurements of the Balmer line ratio (Caplan & Deharveng 1986; Brocklehurst 1971; Osterbrock 1989). The values of the line ratio, I(H /H ), are included in Table 4. This derivation assumes uniform interstellar extinction within the H II region; cases with large corrections for internal extinction should therefore only be considered rough estimates, and derived intrinsic H fluxes may be valid within a factor of 2 to 3 only. The completely corrected H flux is then given by:

Equation (2) was calculated with and without considering the internal extinction . Disregarding should represent a lower limit to the H flux densities and is given as in Table 4; intrinsic fluxes resulting from the complete correction are given as . The table also gives estimates of the sizes of the H emitting regions, , which were taken from the images presented in Sect. 4.

© European Southern Observatory (ESO) 1997

Online publication: June 5, 1998

helpdesk.link@springer.de