Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 360, 49-56 (2000)

Previous Section Next Section Title Page Table of Contents

3. Results

3.1. Continuum

An 18 cm continuum image of NGC 5793 is shown in Fig. 1. The restoring beam size is [FORMULA] milliarcsecond (mas) in P.A. = [FORMULA]. The rms noise in the image is 0.21 mJy beam -1. The map shows a twisted radio structure split into two parts: a central bright component [C1(C)] with accompanying minor components [C1(NE) and C1(SW)] and a western elongated feature in P.A. [FORMULA] -70o (C2(C)), connected to a knot C2(W). The subcomponents C1(NE) and C1(SW) are almost symmetrically extended along a P.A. of 40o around the central peak of C1(C). The knots C1(NE), C2(C), and C2(W) may comprise a bending jet extending over nearly 18 pc. The peak flux density of 308 mJy beam-1 at C1(C) corresponds to a brightness temperature of [FORMULA] K. Brightness temperatures of other components range from 0.2 - 2.0[FORMULA] K.

[FIGURE] Fig. 1. A cleaned continuum image at 18 cm. The contour levels are -1.9, -1.0 (dashed), 1.0, 1.9, 2.7, 3.7, 5.2, 7.2, 10, 14, 19, 27, 37, 52, 72, and 100 [FORMULA] of the peak surface brightness of 308.4 mJy beam-1. The synthesized beam of 12.3 [FORMULA] 4.4 milliarcsecond (mas) in P.A. = -7.5o is shown in the bottom left-hand corner. The position offsets are indicated within the frame. A dashed line shows the optical P.A. of the galaxy. The equivalent linear scale length in parsec is shown at the bottom.

Fig. 2 shows a 6 cm map with the restoring beam size of [FORMULA] mas in P.A. = -4.7o. The rms noise in the image is 0.043 mJy beam-1. The peak flux density at C1(C) is 91 mJy beam-1 corresponding to a brightness temperature of 2.6[FORMULA] K, the brightness temperature of the other components is of the order of [FORMULA] K. The overall structure is similar to that at 18 cm but there are three important differences. First, component C1(SW) seen at 18 cm is not detected at 6 cm at all. Second, in the western elongation, we can clearly identify a new bright component C2(E) at its eastern edge, lying at a distance of 32 mas (7.4 pc) from the peak C1(C). Finally, the position of the peak of component C2(C) is offset northward by [FORMULA] 5 mas with respect to that at 18 cm.

[FIGURE] Fig. 2. A cleaned 6 cm continuum image. The contour levels are [FORMULA] of the peak surface brightness of 91.1 mJy beam-1. The synthesized beam of 3.7 [FORMULA] 1.4 mas in P.A. = [FORMULA] is shown in the left-hand corner.

The total flux density is 647 [FORMULA] 9 mJy at 18 cm and 191 [FORMULA] 3 mJy at 6 cm. Based on a comparison with the VLA B-configuration data (Gardner & Whiteoak 1986), we estimate that our images recover [FORMULA] 65% of the total flux in the 18 cm map, but only [FORMULA] 30% in the 6 cm map. Spectral indices of each component in Figs. 1 and 2 are measured between the two wavelengths and are summarized in Table 2. The 6 cm flux densities were obtained after convolving with the 18 cm beam.


Table 2. VLBI components in NGC 5793.
a) From an image convolved to the 18 cm beam size
c) 5[FORMULA] upper limit value

3.2. The OH absorption

The phase and gain solutions obtained with one absorption-free IF channel were applied to the IF channel which contains both absorption and continuum emission. The continuum level was determined by averaging all channels. The continuum emission was then subtracted from the spectral line visibility data cubes. Fig. 3 shows the 18 cm VLBA spectra with a spectral resolution smoothed to 5.6 km s-1, toward the component C1(C); the two main OH transitions at 1665 and 1667 MHz are clearly seen in absorption. The 1667 MHz absorption spectrum was also previously detected with the VLA (Gardner & Whiteoak 1986). After continuum subtraction the spectra have been converted to optical depth using the above continuum level. In both transitions, several velocity features are detected, although some with poor signal-to-noise ratio. We fitted a single Gaussian component to the spectra shown in Fig. 3 in order to determine the optical depths, velocity centers, and velocity widths. The quality of the data did not warrant using additional components. The resultant values of opacity, integrated intensity and opacity intensities are listed in Table 3. The absorption velocity center (Gaussian fitted) at 1667 MHz is [FORMULA] = 3454 [FORMULA] 4 km s-1 while the center velocity derived by Gardner & Whiteoak (1986) is [FORMULA] = 3462 km s-1, which is very similar. The velocity centroid of the 1665 MHz absorption is 3796 [FORMULA] 3 km s-1 (3454 km s-1, referenced to the 1665 MHz transition), offset by 342 km s-1. Note that the expected velocity difference of these two transitions is 351 km s-1. The difference is probably due to sub-structure of the spectra. There is also weak evidence for 1665 and 1667 MHz OH absorption against components C1(NE) and C2(W), similar to that against C1(C), while the absorption against C1(SW) and C2(C) is too weak for us to estimate optical depths. However, the optical depths, line shapes, and velocity widths of the OH towards C2(W) do not show any significant difference from those towards C1(C). The absorption profiles show that the 1667 MHz feature is as deep as the 1665 MHz feature, though the thin component around at the systemic velocity in the 1665 MHz profile seems to contain a spurious noise. The hyperfine ratios of the peak OH optical depths and the OH opacity intensity derived by single Gaussian-fitting of the absorption line of each transition is 1.2 [FORMULA] 0.1 and 2.5 [FORMULA] 0.2 (see Table 3), showing significant deviation from the theoretical LTE ratio of 1.8 seen in galactic objects (Elitzur 1992).

[FIGURE] Fig. 3. The OH absorption spectra of 1667 (left) and 1665 MHz (right) with a spectral resolution of 31.3 kHz, or 5.6 km s-1. The absorption intensities are labelled as flux density and optical depth. The relative LSR velocities are referred to the rest frequency of each OH transition at 1665.402 and 1667.359 MHz, or [FORMULA] = 3445 and 3454 km s-1. The peak velocities ([FORMULA] = 3190, 3449, 3519, and 3677 km s-1) and velocity ranges of the H2O maser emission are indicated in the 1667 MHz frame by downward-pointing arrows and thick lines. The systemic velocity ([FORMULA]) of 3442 km s-1 is denoted by an upward-pointing arrow.


Table 3. Parameters of the OH absorption feature.
a) Gardner & Whiteoak 1986; No published data for 1665 MHz
b) Values derived from Gaussian fitted curves applied to original spectra in Fig. 3
c) Converted to Radio LSR definition; V(radio heliocentric) - 8.0 km s-1

Fig. 4 shows an OH absorption intensity image, integrated over the velocity span of 44.8 km s-1 at 1667 MHz. The integrated absorption flux density is -158 mJy km s-1 beam-1. Significant OH absorption is detected mainly around the position of C1(C) as seen in Fig. 4. The absorption is apparently concentrated mostly around the position of C1(C) indicating that it traces the background continuum emission there. In Fig. 5, channel maps of the 1667 MHz OH absorption integrated every 5.6 km s-1 are shown; nine channel maps covering a velocity range -23 to 23 km s-1 with respect to the systemic velocity are displayed. The absorption shows a velocity gradient from north to south, starting approximately at [FORMULA] = 3430 km s-1 at 5 mas north of C1(C) and moving to [FORMULA] = 3459 km s-1 at 3 mas south of C1(C). In Fig. 6 a velocity contour map weighted with the OH absorption intensity (the first moment map) is shown. A velocity shift of about 40 km s-1 is seen over a distance of about 20 mas or 4.6 pc in the velocity range of 3425 [FORMULA] 3465 km s-1 along P.A.[FORMULA] 140o, resulting in a velocity gradient of 8.7 km s-1 pc-1. The range of this velocity gradient almost symmetrically spans the systemic velocity of NGC 5793, [FORMULA]= 3442 km s-1. On the other hand, a velocity gradient in the 1665 MHz transition does not clearly have the trend as that seen in the 1667 MHz. This is mainly due to the low signal-to-noise ratio of the 1665 MHz line data; the spiky feature seen only in the 1665 MHz profile makes it difficult to deduce a reliable result.

[FIGURE] Fig. 4. An integrated intensity map of 1667 MHz OH absorption, integrated over the velocity width of 44.8 km s-1. The integrated velocity range is [FORMULA] = 3422-3467 km s-1. Dashed contour levels are 98, 80, 60, 40, 20, and 10% of the peak intensity of -158 mJy km s-1 beam-1. The synthesized beam is shown in the left-hand corner. Significant absorption intensity is seen only around C1(C).

[FIGURE] Fig. 5. Velocity channel maps of the 1667 MHz OH absorption, each integrated over 5.6 km s-1 intervals. The contour levels are at -25, -20, -15, -10, -5 ([FORMULA]) (dashed), 5, and 10 (solid) mJy beam-1. Each map is labeled in the upper right hand corner with the velocity ([FORMULA]) at the center of each channel. The cross in each panel denotes the position of the phase center. The synthesized beam is shown in the upper-left corner panel.

[FIGURE] Fig. 6. The first moment map shows a distinct velocity field for the OH absorption at 1667 MHz covering the central region C1. Velocity contours are presented at 5 km s-1 intervals, and velocities are labeled in units of km s-1 ([FORMULA]). The 18 cm synthesized beam is plotted in the bottom left corner. The velocity gradient is clearly seen nearly along the major axis of the galaxy ranging [FORMULA] = 3430-3455 km s-1.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 2000

Online publication: July 27, 2000