4 Fabry-Perot images and velocity fields

In Figs. 1 to 14 we present the Fabry-Perot images of the observed galaxies. In most cases a continuum image of each galaxy is given in the upper-left panel, and a monochromatic (H$\alpha$) image in the mid-left panel.

  

Figure 1: ESO 350-IG38. Upper left: Continuum isointensity image. Mid-left: Monochromatic (H$\alpha$) isointensity map. The (relative) step between two consecutive levels is the same in both images. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the H$\alpha$ image (thin solid lines). The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) derived from the velocity field and based on $S = 45\hbox{$^\circ$}$, ${\rm incl} = 40\hbox{$^\circ$}$ and ${\rm PA} = 320\hbox{$^\circ$}$. The receding side is indicated by $\bigoplus$ and the approaching side by $\bigodot$.The weighted average of the two sides is indicated by $\bullet$ with the lengths of the errorbars equal to the $\pm 1 \sigma$ dispersion of the data points within the sector. Velocity points which are based on only one or two data points are enclosed by brackets (). The curve (solid line) is a smoothed splinefit to the average RC. The cloud of small dots represent those velocity points within $\pm 30 \hbox{$^\circ$}$ from the major axis. Arrows ($\uparrow$) on the abscissa, and the crosses ($\times$) on the fitted RC, mark the location of 0.4 R25, 0.8 R25, 1.0 R25, 1.2 R25. The scale on the upper abscissa is the radius in kpc (on the major axis), based on the radial velocity and H0 = 75 km s-1/Mpc

  

Figure 2: ESO 350-IG38 continued. Upper left: Line profiles in the centre of the non-decomposed velocity field plus isovelocity contours of the total, non-decomposed, velocity field. Each box corresponds to one pixel. The Y-axis scale is normalised to the peak intensity in each pixel. Mid-left: The line profiles in the centre of the two fitted components after decomposition of the velocity field. Each box corresponds to one pixel and the intensity scale is normalised to the pixel with the brightest H$\alpha$ flux. Upper right: Isovelocity contours of the first component of the ionised gas superimposed on the H$\alpha$ emission from the first component. Bottom: Rotation curve (RC) based on the decomposed velocity field. The usual symbols corresponds to the first component, and are based on $S = 45\hbox{$^\circ$}$, ${\rm incl} = 40\hbox{$^\circ$}$ and ${\rm PA} = 320\hbox{$^\circ$}$. The secondary component, which is counter rotating, is shown as boxes with crosses in, and represent the weighted mean of both sides. It is based on $S = 85\hbox{$^\circ$}$, ${\rm incl} = 40\hbox{$^\circ$}$ and ${\rm PA} = 140\hbox{$^\circ$}$; N.B. that this RC is based on 9 pixels only. For further explanations see the caption of Fig. 1 and Sect. 5

  

Figure 3: ESO 480-IG12: Upper left: Continuum isointensity image. Mid-left: First monochromatic (H$\alpha$) component plotted with the same step as in the continuum image. The isovelocity contours of the first component are overlaid (thick solid lines). Upper right: Isovelocity contours of the first component of the ionised gas (thick lines) superimposed on the two Gaussian components of the two velocity fields (thin lines) plotted within squares of 2 pixels. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) based on $S = 60 \hbox{$^\circ$}$, ${\rm incl} = 52\hbox{$^\circ$}$ and ${\rm PA} = 234\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 4: ESO 480-IG12 continued. Upper left: Monochromatic (H$\alpha$) isointensity image of the total (non-decomposed) velocity field. Bottom left: Second monochromatic component plotted with the same step as in the continuum of the first component images (Fig. 3). Right: Isovelocity contours of the second component of the ionised gas (thick lines) superimposed on the non-decomposed line profiles (thin lines) plotted within squares of $2 \times 2$ pixels. The intensity scale (Y-axis) in each box with line profiles is normalised to the peak intensity in that box. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left

  

Figure 5: ESO 338-IG04: Upper left: Continuum isointensity image. Note the bright field star superimposed on the western part of the galaxy. Mid-left: Monochromatic (H$\alpha$) isointensity map. The step between two consecutive levels is the same for the continuum and monochromatic images. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the monochromatic image (thin solid lines). The step between two consecutive contours is 10 times larger than the step used in the mid-left image. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) based on $S = 45\hbox{$^\circ$}$, ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 60\hbox{$^\circ$}$. The arrows ($\uparrow$) on the abscissa mark the location of 0.4 R25, 0.8 R25, 1.0 R25 and 1.2 R25. Due to the disagreement between the approaching and receding sides, no average RC has been computed. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 6: ESO 338-IG04 continued: Upper: Rotation curve (RC) of main component using the "masked'' velocity field, i.e. with points north-east of the centre and in the eastern tail (up to where the isophotes of the bright central starburst region starts in Fig. 5) excluded, and limiting the half sector to $S = 30\hbox{$^\circ$}$. As before ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 60\hbox{$^\circ$}$. Bottom: Rotation curve of the "perpendicular" component, based on $S=20\hbox{$^\circ$}$, ${\rm incl}=62\hbox{$^\circ$}$ and ${\rm PA} = 336\hbox{$^\circ$}$. Its centre lies 5$\hbox{$^{\prime\prime}$}$east of the point where the H$\alpha$ flux has its peak. For further explanation of what is shown in the RCs see the caption of Fig. 1 and Sect. 5

  

Figure 7: ESO 338-IG04 continued: Based on FP1 data, with lower velocity resolution: Upper left: Continuum isointensity image. Note the bright field star superimposed on the western part of the galaxy. Mid-left: Map of the velocity dispersion (FWHM) in units of km s-1. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the monochromatic H$\alpha$ image (thin solid lines). The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) based on the FP1 velocity field and $S = 55\hbox{$^\circ$}$, ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 60\hbox{$^\circ$}$. The arrows ($\uparrow$) on the abscissa mark the location of 0.4 R25, 0.8 R25, 1.0 R25 and 1.2 R25. Due to the disagreement between the approaching and receding sides, no average RC has been computed. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 8: ESO 338-IG04 B: Upper left: Continuum isointensity image. Mid-left: Monochromatic H$\alpha$ isointensity map. The step between two consecutive levels is the same for the continuum and monochromatic images. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the monochromatic image (thin solid lines). The step between two consecutive contours is 10 times lower than the step used in the mid-left image. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) based on $S = 40\hbox{$^\circ$}$, ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 230\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 9: ESO 185-IG13: Upper left: Total monochromatic H$\alpha$ emission (no decomposition of first and secondary components), with the total velocity field overlaid. Mid-left: Profiles within square boxes of $2 \times 2$ pixels of the decomposed first and second monochromatic components. Upper right: Isovelocity contours of the ionised gas of the first component (thick solid lines) superimposed on the first monochromatic H$\alpha$ image (thin solid lines). The step between two consecutive contours is 4 times larger here than those used for the continuum contours. The spatial scale in each image is shown in the lower right corner. North is up, east is left. Bottom: Rotation curve (RC) of the primary (1st) component based on $S = 40\hbox{$^\circ$}$, ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 25\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 10: ESO 185-IG13 continued: Upper left: Total continuum isointensity image. Mid-left: Line profiles of the non-decomposed data plotted within squares of $2 \times 2$ pixels. The intensity scale (Y-axis) in each box with line profiles is normalised to the peak intensity in that box. Upper right: Isovelocity contours of the ionised gas of the second component (thick solid lines) superimposed on the H$\alpha$ image of the secondary component (thin solid lines). The step between two consecutive contours is 4 times larger here than those used for the continuum contours but is the same as the step chosen for the first component. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) of the secondary component based on $S = 40\hbox{$^\circ$}$, ${\rm incl} = 50\hbox{$^\circ$}$ and ${\rm PA} = 225\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 11: ESO 400-G43: Upper left: Continuum isointensity image. Mid-left: Monochromatic H$\alpha$ isointensity map. The step between two consecutive levels is the same for the continuum and monochromatic images. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the monochromatic image (thin solid lines). The step between two consecutive contours is 6.7 times larger than the step used in mid-left image. Bottom: Rotation curve (RC) based on $S = 55\hbox{$^\circ$}$, ${\rm incl}=55\hbox{$^\circ$}$ and ${\rm PA} = 225\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 12: ESO 400-G43 B: Upper left: Continuum isointensity image. Mid-left: Monochromatic H$\alpha$ isointensity map. The step between two consecutive levels is 5 times larger for the monochromatic image than for the continuum map. Upper right: Isovelocity contours of the ionised gas (thick solid lines) superimposed on the monochromatic image (thin solid lines). The step between two consecutive contours is 4 times larger than the step used in mid-left image. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) based on $S = 45\hbox{$^\circ$}$, ${\rm incl}=60\hbox{$^\circ$}$ and ${\rm PA} = 305\hbox{$^\circ$}$. The fitted curve (solid line) has been corrected for absorption in the centre due to the high inclination of this galaxy, therefore the curve lies slightly above the measured velocity points at radii smaller than $6\hbox{$^{\prime\prime}$}$ For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 13: Tololo 0341-407: Upper left: Continuum isointensity image. Mid-left: Total H$\alpha$ emission (addition of both components). Upper right: Isovelocity contours of the ionised gas (thick solid lines) of the first component of the eastern galaxy, and the isovelocity contours of the second component (thin lines) of the western galaxy superimposed on the monochromatic H$\alpha$ image of the first component (where the eastern galaxy dominates). The step between two consecutive contours is the same as those used for the continuum contours. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) of the eastern galaxy first component, based on $S = 60 \hbox{$^\circ$}$, ${\rm incl}=15\hbox{$^\circ$}$ and ${\rm PA}=295\hbox{$^\circ$}$. For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

  

Figure 14: Tololo 0341-407 continued: Upper left: Individual profiles. Two Gaussian profiles are plotted in each pixel, corresponding to the two monochromatic components and the two velocity fields. Mid-left: Total H$\alpha$ emission (addition of both components) with total (non-decomposed) velocity field overlaid. Upper right: Isovelocity contours of the ionised gas of the first component of the western galaxy (thick solid lines) and isovelocity contours of the second component of the eastern galaxy (thin lines); superimposed on the monochromatic H$\alpha$ image of the second component (where the western galaxy dominates). The step between two consecutive contours is the same as that used for the continuum and first monochromatic contours. The spatial scale in each image is shown in the lower right corner. North is up, east is to the left. Bottom: Rotation curve (RC) of the first component of the western galaxy, based on $S=50\hbox{$^\circ$}$, ${\rm incl} = 50\hbox{$^\circ$}$ and ${\rm PA}= 313\hbox{$^\circ$}$.For further explanation of what is shown in the RC see the caption of Fig. 1 and Sect. 5

The contour levels are usually the same in the continuum and monochromatic images, except that the continuum is the integrated intensity over the used narrow band filter while the monochromatic images present the total H$\alpha$ flux. The threshold of all the images was chosen to be at 3$\sigma$ above the noise in the background. All images have linearly spaced contour levels. The continuum images shown have been smoothed with a Gaussian filter (FWHM of 3$\hbox{$^{\prime\prime}$}$). The H$\alpha$ images are the full resolution maps, but in drawing the contours the data was slightly smoothed with a median filter. In general, in the right panel we present the isovelocity contours of the ionised gas (thick solid lines) superimposed on the H$\alpha$ image (thin solid lines). The number of contour levels were here reduced to make the plots more readable. The heliocentric radial isovelocity lines have been drawn after a smoothing the original profiles with a Gaussian filter (FWHM $1.8\hbox{$^{\prime\prime}$}\times 1.8\hbox{$^{\prime\prime}$}$). The isovelocity contours are labelled in units of km s^-1. In each plot: north is up and east is left, and the scale is given in the lower right corner. In general the relative velocities are certain to within a fraction of the sampling step, but there might be systematic offsets of a few times the FSR, since the narrow band filters used are four to eight FSRs broad, giving a maximum systematic velocity error of a few times 100 km s^-1. However, for all galaxies there are independent velocity determinations available (see Table 2), which give us confidence in the determined systemic velocities (Table 3).

4.1 Multicomponent decomposition

For some of the target galaxies we see obvious double line profiles (or profiles with very broad and flat peaks) indicating that gas with two different velocities are present. When this was seen in more than a single isolated pixel but in a region of at least 3 by 3 pixels, and with sufficient S/N to rule out a noise peak, we attempted to decompose the velocity field into two components. This was done after subtracting the continuum and employing a small spectral smoothing with a Gaussian filter (FWHM of 15 km s^-1), in order to avoid artifacts due to noise. For this purpose a classical least-squares algorithm was used to fit two Gaussian profiles simultaneously to the full profile in the individual pixels. To reduce the number of free parameters in the fit, we imposed the constraint that the two Gaussians should have the same FWHM (as measured where they are well separated). The decomposition was performed individually on single pixels, but subject to a condition of continuity, i.e. in determining which of the fitted Gaussian components that belong to the same dynamical large scale component, the velocity of either component was not allowed to make sudden jumps from pixel to pixel. Thus the velocity of one dynamical component in a certain pixel had to be consistent with its neighbouring pixels.

For several galaxies, we attempted a decomposition, but the results are only presented when we obtained a regular and continuous large scale secondary component. The failure to decompose the data cube into two components does not guarantee that secondary dynamical components are not present, only that we cannot detect them. Moreover, we have no guarantee that the components are not separated by the value determined here plus the free spectral range of the interferometer. Nevertheless, the separation between the two components are often so small than they should not be apparent if we had used a lower resolution Fabry-Perot. In those cases where we have extracted two dynamical components, individual H$\alpha$ profiles are superimposed on the velocity field of the galaxy. To make it more readable, one profile is usually drawn in a 2 by 2 pixels square box. The abscissa of each small spectrum covers one FSR of 117 km s^-1, while the Y-amplitude is normalised to the peak value of the monochromatic flux in the field of view. In some cases where we see double or broad lines, we also present the line profiles of the non-decomposed velocity field with the Y-amplitude normalised to the maximum intensity in the box, which makes faint parts more visible.

For most galaxies, individual pixels or small regions can be found where the line profiles are double or asymmetric. This is an indication of small/intermediate scale gas motions, e.g. expanding bubbles or stellar winds.

It cannot be excluded that several sources, overlaid on top of each other, have monochromatic emission within the narrow band filter (within a few times 100 km s^-1 of the target galaxy). Then we could misinterpret a background/foreground galaxy for a secondary dynamical component in the galaxy studied. This is however very unlikely in view of the appearance of the data. That the velocity structure within the target galaxies (main component) could be a misinterpretation and rather be due to two separate galaxies a few 100 km s^-1 apart is extremely unlikely since we then would expect to see double components in the profiles unless there was a perfect match in velocity (plus a multiple of the FSR), orientation and rotational velocity and direction. Of course there might be more than two velocity components present, but we do not attempt a decomposition into more components, since third components do not display any meaningful structure over a scale larger than one or a few pixels. Thus third components (and second components in those cases where we do not decompose the velocity field) may reflect local small scale gas motions. A problem in attempting a multicomponent decomposition is that in principle any Gaussian can be decomposed into an infinite number of Gaussian components, and thus the result would be unreliable and model dependent.