Astron. Astrophys. 358, 812-818 (2000)

1. Introduction

There is now great observational evidence that disk-halo interactions in galaxies as well as the structure of the interstellar medium (ISM) is closely related to star formation processes in the disks of spiral galaxies (see the review by Dahlem 1997). Big shells develop around the brightest star forming regions, induced by the energy input of supernovae and the strong stellar winds produced by high-mass stars. These shells can grow enough to be able to break the disk, allowing large amounts of gas to blow out from the disk along these big chimneys (Norman & Ikeuchi 1989). The very hot gas going out through these chimneys cools as it rises until it eventually recombines and condenses to form clouds of neutral gas that fall back to the plane (Shapiro & Field 1976; Bregman 1980). This fountain model then provides an explanation for the origin of high velocity clouds (HVC's) that have been observed in our galaxy and a in few external galaxies (with the galaxy studied in this paper being one of those few (Schulman et al. 1996, hereafter S96)), although alternative explanations have been proposed as well (Blitz et al. 1999). Good reviews on this topic can be found in Wakker & van Woerden (1997) and van der Hulst (1996, 1997). On the other side, the expanding shells can induce new star formation (sequential star formation (SSF)) at their edges. This effect has already been observed in some galaxies (see for example Thilker et al. 1998, hereafter T98) and it is also observed to take place in the case of NGC 5668 in our observations.

According to the chimney model, the structure of the ISM, and in particular whether the chimney phenomenon takes place or not, is controlled by the amount of star formation. The study of the properties of these phenomena in a sample of nearby galaxies can greatly help to understand the structure of the ISM and the nature of disk-halo interactions. Observations of the neutral gas and narrow band imaging of the ionized gas have already been extensively used to study these phenomena (see the review by Dahlem 1997). Scanning long-slit spectroscopy has also been used to study these phenomena (Saito et al. 1992; Tomita et al. 1993, 1994). In this work we make a first attempt to use optical Fabry-Perot spectroscopy in a nearly face-on spiral galaxy to directly study these vertical motions. Therefore we have chosen the spiral galaxy NGC 5668 which is already known to have HVC's and an important rate of star formation. These facts make it a perfect candidate for us to detect important vertical motions in its disk.

In Sect. 2 we describe the general properties of the galaxy NGC 5668, with particular emphasis on the observation of the HVC's. Sect. 3 deals with the observations and data reduction (including calculation of intensity, velocity and velocity dispersion maps). Sect. 4 is devoted to the calculation of the rotation model for the galaxy from the observed velocity field. Sect. 5 describes how the residual velocity field is used to look for systematic deviations of circular rotation and how comparison of the geometry of the residual velocity field with that of the intensity and velocity dispersion in some regions can be used to detect real shells and/or chimneys. Finally, in Sect. 6 we report the shell candidates found in NGC 5668 and some of their properties.

© European Southern Observatory (ESO) 2000

Online publication: June 20, 2000
helpdesk.link@springer.de