0804.3819/ms.tex
1: % v. 04/23/2008 JR 
2: 
3: %\documentclass[12pt,preprint]{aastex}
4: \documentclass[apjl]{emulateapj}
5: %\documentclass[manuscript]{aastex}
6: %\usepackage{psfig}
7: %\usepackage{epsf}
8: 
9: \shorttitle{ACS OBSERVATIONS OF EDIG IN EDGE-ON SPIRALS}
10: \shortauthors{ROSSA ET AL.}
11: 
12: %%%%%%%%%%%%%%%
13: % User defined macros
14: %%%%%%%%%%%%%%%
15: 
16: \newcommand{\etal}{{et al.~}}
17: 
18: \newcommand{\lta}{\lesssim}
19: \newcommand{\gta}{\gtrsim}
20: 
21: \newcommand{\kms}{\>{\rm km}\,{\rm s}^{-1}}
22: \newcommand{\kpc}{\>{\rm kpc}}
23: \newcommand{\Mpc}{\>{\rm Mpc}}
24: \newcommand{\pc}{\>{\rm pc}}
25: \newcommand{\Msun}{\>{\rm M_{\odot}}}
26: \newcommand{\Lsun}{\>{\rm L_{\odot}}}
27: 
28: %%%%%%%%%%%%%%%
29: % Title and authors
30: %%%%%%%%%%%%%%%
31: 
32: \begin{document}
33: 
34: \title{{\it HST} ACS OBSERVATIONS OF STAR FORMATION DRIVEN 
35: OUTFLOWS IN NEARBY EDGE-ON SPIRAL GALAXIES: DEPENDENCE OF HALO MORPHOLOGY 
36: ON STAR FORMATION ACTIVITY\altaffilmark{1}}
37: 
38: %% Use \author, \affil, and the \and command to format
39: %% author and affiliation information.
40: %% Note that \email has replaced the old \authoremail command
41: %% from AASTeX v4.0. You can use \email to mark an email address
42: %% anywhere in the paper, not just in the front matter.
43: %% As in the title, use \\ to force line breaks.
44: 
45: \author{J\"{o}rn Rossa\altaffilmark{2,3},
46: Michael Dahlem\altaffilmark{4},
47: Ralf-J\"urgen Dettmar\altaffilmark{5}, and 
48: Roeland P. van der Marel\altaffilmark{2}}
49: 
50: \altaffiltext{1}{Based on observations made with the NASA/ESA Hubble
51: Space Telescope, obtained at the Space Telescope Science Institute,
52: which is operated by the Association of Universities for Research in
53: Astronomy, Inc., under NASA contract NAS 5-26555. These observations
54: are associated with proposal 10416.}
55: 
56: \altaffiltext{2}{Space Telescope Science Institute, 3700 San Martin
57: Drive, Baltimore, MD 21218; marel@stsci.edu}
58: 
59: \altaffiltext{3}{Present address: Department of Astronomy, University
60: of Florida, 211 Bryant Space Science Center, P.O.Box 112055,
61: Gainesville, FL 32611-2055; jrossa@astro.ufl.edu}
62: 
63: \altaffiltext{4}{CSIRO/ATNF - Paul Wild Observatory, Locked Bag 194, Narrabri 
64: NSW 2390, Australia; Michael.Dahlem@csiro.au}
65: 
66: \altaffiltext{5}{Astronomisches Institut, Ruhr-Universit\"at Bochum, 
67: Universit\"atsstrasse 150/NA7, D-44780 Bochum, Germany; dettmar@astro.rub.de}
68: 
69: %%%%%%%%%%%%%%%
70: % Abstract
71: %%%%%%%%%%%%%%%
72: 
73: \begin{abstract}
74: 
75: We present new high spatial resolution narrowband imaging observations of 
76: extraplanar diffuse ionized gas (eDIG) in four late-type, actively star 
77: forming edge-on spirals, obtained with ACS on-board {\it HST}. Our F658N 
78: (H$\alpha$) observations reveal a multitude of structures on both small and 
79: large scales. Whereas all four galaxies have been studied with ground-based 
80: telescopes before, here the small scale structure of the extended emission 
81: line gas is presented for the first time at a spatial resolution of 0\farcs05, 
82: corresponding to 5.0\,pc at the mean distance to the target galaxies. The eDIG 
83: morphology is {\em very} different for all four targets, probably as a result 
84: of their different levels of star formation activity. We find that the morphology 
85: of the eDIG, in particular the break-up of diffuse emission into filaments in galaxy 
86: halos, shows a strong dependence on the level of star formation activity per unit 
87: area, and eDIG can be arranged into a morphological sequence. The eDIG is 
88: resolved into an intricate morphology of individual filaments for NGC\,4700 and 
89: NGC\,7090. The ionized gas in the latter galaxy, being the one with the lowest SF activity 
90: per unit area in our sample, reveals besides some filaments many prominent bubbles 
91: and superbubbles. NGC\,4634 and NGC\,5775 have instead a more diffuse layer of eDIG.
92: The latter two galaxies have the highest SF rate per unit area in our sample and 
93: the observed morphology suggests that the break-up of the smooth eDIG layer into 
94: individual resolved filaments occurs only above a certain threshold of SF activity 
95: per unit area. We note that this is not a distance effect, as NGC\,4634 and 
96: NGC\,4700 are roughly at the same distance, yet reveal very different eDIG 
97: morphology. Combined with ground-based data for samples that span a larger range 
98: of galaxy mass our results indicate that the gravitational potential also plays an 
99: important role in the eDIG morphology. In low-mass galaxies the gas can be expelled 
100: due to shallow gravitational potentials more easily and couple with strong 
101: star formation driven outflows on a local scale. This is in contrast to the more 
102: massive galaxies, which show smooth eDIG layers, unless they are powered by a 
103: superwind, as in the case of nucleated starburst galaxies. The simplistic chimney 
104: scenario is not confirmed from our observations, and future models need to take into 
105: account the detailed physics of feedback, star formation and magnetic fields. 
106: 
107: \end{abstract}
108: 
109: %%%%%%%%%%%%%%%
110: % Keywords
111: %%%%%%%%%%%%%%%
112: 
113: \keywords{galaxies: evolution --- galaxies: halo --- galaxies: ISM --- galaxies: spiral --- 
114: galaxies: structure}
115: 
116: \clearpage
117: 
118: %%%%%%%%%%%%%%%
119: % Beginning of main text
120: %%%%%%%%%%%%%%%
121: 
122: \section{INTRODUCTION}
123: \label{s:intro}
124: 
125: During the last decade considerable progress in our understanding of the 
126: diffuse ionized gas (DIG) in halos of late-type spiral galaxies has been 
127: achieved. The occurrence of DIG outside of classical \hbox{H\,{\sc ii}} 
128: regions is firmly believed to be correlated with the star formation activity 
129: in the galaxy disk. 
130: 
131: The fractional contribution of DIG to the total H$\alpha$ emission in 
132: galaxies reaches typically values of $\approx$ 25-50\% \citep[e.g.,\,][]{
133: wal94,fer96,leh95}, and DIG is thus an important constituent of the 
134: interstellar medium (ISM). Recent studies with a larger sample of galaxies 
135: from the SINGG survey revealed even slightly higher fractions of $0.59\pm0.19$ 
136: \citep{oey07}. 
137: 
138: In the theoretical picture the gas, most likely driven by collective 
139: supernovae, is expelled into the halos of the galaxies. \citet{nor89} 
140: developed a theoretical model in which the hot gas is transported from the 
141: star forming regions in the disk via {\em chimneys} into the halo, provided 
142: the superbubbles created from SNe fulfill the breakout criterion. Depending 
143: on the strength of the gravitational potential the gas may fall back onto the 
144: galactic disk, which is described as the {\em galactic fountain} scenario 
145: \citep{sha76,bre80,avi00}. Starburst driven mass loss, at least for dwarf 
146: galaxies with shallower potential wells, can be an effective process, as the 
147: gas can escape more easily in these systems \citep{mac99}.
148: 
149: Several ground-based narrowband imaging investigations have been undertaken 
150: during the past fifteen years with a boost of investigations occuring in the 
151: last decade, to study the morphology of DIG in halos of non-starburst 
152: {\em quiescent} spiral galaxies \citep{det90,ran90,ran92,pil94,ran96,
153: hoo99,col00,ros00,ros03b,mil03}. Not all of the studied galaxies revealed 
154: extraplanar DIG (eDIG), but it was shown that extraplanar DIG is a relatively 
155: common feature among actively star forming galaxies, and about 41\,\% of a 
156: survey of 74 galaxies have eDIG detections. The most active ones were 
157: found to possess a bright gaseous halo \citep{ros03a}. Correlations of eDIG 
158: with other constituents of the ISM exist, such as radio-continuum halos 
159: \citep[e.g.,][]{dah94,tul00,dah06}, the hot ionized medium (i.e. halos), 
160: traced by X-ray observations \citep{str04,tul06}, and dust filaments in the 
161: disk-halo interface \citep{how97,how99,how00,ros03b,ros04,tho04}.   
162: Starburst galaxies and low luminosity Seyfert galaxies have been investigated as 
163: well with regard to the diffuse emission \citep{leh95} and \citep{rob07}, 
164: respectively. 
165: 
166: In general, the spiral galaxies with very high SFR per unit area show a pervasive 
167: DIG layer (i.e. a halo) with a typical vertical extent of $|z| \sim$ 
168: 2\,kpc. Individual filaments, occasionally reaching distances of up to 6\,kpc, 
169: are superimposed in a few cases as well \citep{ros03b}. As there is a variety 
170: of morphological features generally discernible even with low spatial 
171: resolution ground-based studies (e.g.,\,filaments, plumes, bubbles, shells), 
172: clearly the need for a more magnified view onto the individual SF regions and 
173: DIG filaments in edge-on spirals is needed to study the small scale 
174: structure and to identify the connection points between the filaments and 
175: their origin within the SF regions in the midplane. Only very few previous 
176: high spatial resolution investigations were targeted to address these 
177: questions \citep[e.g.,][]{cec01,cec02,ros04}. For a more thorough overview on 
178: the DIG topic in general and on phenomena related to these star formation 
179: driven gaseous outflows we refer to the older, but still quite relevant, 
180: review articles by \citet{det92,dah97} and also to the more recent review on 
181: galactic winds by \citet{vei05}.
182: 
183: In this study we present high spatial resolution narrowband H$\alpha$ images 
184: of four actively star forming galaxies, observed with {\it HST}/ACS to 
185: unravel the morphology of the gaseous outflows on both small and large scales as 
186: a function of star formation activity, These data are augmented by broad-band 
187: {\it HST}/ACS images, whose primary purpose was to serve for the continuum 
188: subtraction. However, they are also used to study the distribution of dust 
189: filaments in the disk-halo interface.  
190: 
191: The paper is structured in the following way. In \S~\ref{s:sample} we 
192: describe how the current {\it HST}/ACS sample was selected, and in 
193: \S~\ref{s:obs} we present the observations and observing strategy. In 
194: \S~\ref{s:data} we describe briefly the data reduction techniques and 
195: analysis of the images. In \S~\ref{s:results} the results are presented, and 
196: in \S~\ref{s:discuss} we discuss the results. In \S~\ref{s:sum} we finally 
197: summarize our basic results and state our conclusions. 
198: 
199: \section{THE EDGE-ON GALAXY SAMPLE}
200: \label{s:sample}
201:    
202: The small edge-on galaxy sample presented here is a sub-sample of four 
203: galaxies from our previous H$\alpha$ survey \citep{ros00,ros03a,ros03b}. 
204: The basic properties of these galaxies are listed in Table~\ref{t:obs}. 
205: The galaxies were selected to be observed with {\it HST}/ACS, because they 
206: have very different levels of energy input into their disk ISM (expressed
207: as $L_{\rm{FIR}}/D^2_{25}$ and listed with other properties in Table~\ref{t:prop}). 
208: All four targets have eDIG detections \citep{ros00,ros03a,ros03b}, but ground-based 
209: imagery suggests different halo morphologies that can be further studied at the highest 
210: possible resolution only by means of {\it HST}/ACS H$\alpha$ imagery. Galaxies 
211: with different energy input rates were chosen to study a possible dependence 
212: of the halo morphology on the level of star formation in the underlying 
213: disks. Naively, one might expect galaxies with energy input levels near 
214: the threshold for the creation of halos (see Figure~\ref{f:dddhst}) to 
215: exhibit only individual, possibly filamentary extraplanar structures.
216: On the other hand, galaxies with widespread, high-level star formation 
217: are expected to have a multitude of filaments, if they are associated 
218: with individual sites of star formation in the underlying disks.
219: 
220: We avoided galaxies with prominent dust lanes and those at low Galactic 
221: latitudes so as to have low foreground opacities for studies of the 
222: optical emission line gas.
223: 
224: We note here that two galaxies in the present sub-sample, NGC\,4634
225: and NGC\,5775, have close interaction partners (within two Holmberg
226: radii), while the remaining two, NGC\,4700 and NGC\,7090, don't.
227: 
228: \section{OBSERVATIONS AND OBSERVING STRATEGY}
229: \label{s:obs}
230: 
231: The observations of the four actively star forming spirals were carried 
232: out between June 23 and August 21 2005 with the Advanced Camera for Surveys 
233: (ACS) onboard {\it HST} (GO\#10416; PI: J.~Rossa). All observations were carried out 
234: using the Wide Field Channel (WFC), which yields a $202\arcsec \times 202\arcsec$ 
235: field of view. This is achieved by employing two SITe 2048$\times$4096 pixel CCDs 
236: with a pixel scale of 0\farcs05 per pixel. For more details we refer to the ACS 
237: instrument handbook \citep{gon05}. The narrowband observations using the 
238: F658N filter were acquired in three {\it HST} orbits for each galaxy, and the 
239: broadband images (to be used for continuum subtraction and to study the dust 
240: morphology in the galaxies) were obtained with the F625W filter in one {\it 
241: HST} orbit each. In total, 16 {\it HST} orbits were allocated to obtain deep 
242: continuum subtracted H$\alpha$ images of the four target galaxies. The 
243: observational details, including observation dates, integration times, and 
244: filter characteristics, are listed in Table~\ref{t:exp}. 
245: 
246: The individual exposures were taken in dither mode, to allow for good 
247: cosmic ray removal and to fill in the CCD gaps. We have used a different 
248: dither pattern for the broadband and narrowband observations. For the 
249: narrowband (F658N) observations we have used a 3-point dither pattern with 
250: a 4-point sub-pattern, spread over three {\it HST} orbits for each galaxy. 
251: The broadband (F625W) observations were obtained in a 2-point dither with 
252: a 2-point sub-pattern. The first pattern in each configuration applies 
253: a jump over the CCD gaps, and the secondary pattern is applied to allow 
254: for good cosmic ray and hot pixel removal.     
255: 
256: \section{DATA REDUCTION AND ANALYSIS}
257: \label{s:data}
258: 
259: \subsection{Data reduction}
260: \label{ss:datared}
261: 
262: The {\it HST}/ACS data were calibrated using the {\it CALACS} and 
263: {\it Multidrizzle} packages. The former removes the various instrumental 
264: signatures, including bias correction and flat-fielding. The latter is 
265: used to correct for the geometric distortion of the ACS camera and also 
266: performs cosmic ray rejection and finally combines the images, obtained in 
267: the specific dither pattern. For further details we refer to the ACS data 
268: handbook \citep{pav05} and to \citet{koe05} for a description of multidrizzle. 
269: 
270: There are a few detector artifacts visible at very low intensities. Among them 
271: are some faint stripes along CCD columns visible across the images. These are 
272: charge transfer trails of hot pixels, cosmic rays, and other sources. However, 
273: these CTE losses are well defined features, which are located at a constant position, 
274: and whose pattern is repeated with a constant distance and position angle. On very 
275: faint levels, when viewing the entire frame in low resolution on a computer screen, 
276: effects of cross talk are discovered. They have their origin in the four CCD 
277: quadrants corresponding to the four amplifiers of the detector array. The features 
278: though look somewhat different from the usual cross talk features seen in 
279: \citet{gia04}. The features in our images are not inverted in intensity and 
280: thus can easily be mistaken for faint diffuse emission. However, the distance 
281: and position angle of this feature in the continuum subtracted H$\alpha$ image of 
282: NGC\,4634 mimics the shape of the galaxy at a very faint level, and its endpoint 
283: coincides with the one end of the galaxy, at a position that has an offset and 
284: inverted position. As an almost identical feature was detected in the image of 
285: NGC\,5775 as well, which looked very reminiscent, we conclude, that this is 
286: indeed related to the detector, and not a real feature of the diffuse 
287: emission. At higher contrast the structure resolves and is barely visible at 
288: all.    
289:  
290: \subsection{Calibration and Analysis of the Broadband Imagery}
291: \label{ss:broad}
292: 
293: The broadband (F625W) images of our target galaxies were obtained for two 
294: purposes. The primary aim was to subtract the continuum from the on-band 
295: (F658N) images, which contain some continuum flux. Secondly, they allow a 
296: detailed comparison of extraplanar dust features with the DIG morphology, 
297: which are best seen in unsharp-masked versions of the broadband images. 
298: These two physically distinct ISM phases \citep[see\,][]{how00} can be 
299: studied in detail to search for possible correlations between the two 
300: extraplanar ISM constituents on small scales. To enhance the contrast between 
301: the dust structures and the background light from the galaxy more easily, we 
302: have constructed unsharp-masked versions of our broadband F625W (SDSS $r$) 
303: images. These were obtained by dividing the F625W images by a smoothed 
304: version of the same images. The latter ones were created by using a 
305: Gaussian filtering technique with a kernel size adopted to account for 
306: the difference in distance among the sample galaxies.   
307: 
308: \subsection{Calibration and Analysis of the Narrowband Imagery}
309: \label{ss:narrow}
310: 
311: In order to calibrate the narrowband images the broadband and narrowband 
312: images were first aligned by measuring the positions of stars in the field of 
313: view. After alignment the shifted broadband images were checked by blinking 
314: them on a computer screen against the narrowband images. In order to obtain a 
315: continuum-free H$\alpha$ image, we scaled and subtracted each combined 
316: broadband F625W image from the respective narrowband F658N image. This was 
317: done in a similar fashion as it was applied in our previous studies 
318: \citep[e.g.,][]{ros00,ros04}. The use of individual foreground stars instead 
319: proved difficult, as there were only limited suitable stars in the given 
320: aperture visible. More importantly, they also showed too much color variation 
321: so that the scaling factor could not be determined very accurately. Therefore, 
322: we chose selected reference regions within the galaxy that were not affected 
323: by emission. For this purpose we closely compared the F625W and F658N images 
324: to determine suitable regions within the disk, and also relied on previous 
325: ground-based images.  
326: 
327: Several of the continuum subtracted images show black-white point source pairs
328: which are due to slight imperfections in alignment and differences in the PSF. 
329: However, they do not affect our discussion and understanding of the extended 
330: emission. Due to the given FWHM of the used narrowband filter, it should be noted 
331: that emission from the adjacent [\ion{N}{2}] doublet is included in the F658N 
332: filter passband. For simplicity, however, we refer for the remainder of the 
333: paper to an H$\alpha$ image, whenever we speak of a continuum subtracted 
334: F658N filter image. 
335: 
336: 
337: \section{RESULTS}
338: \label{s:results}
339: 
340: In this section we describe in detail the morphology of the four edge-on 
341: spirals as revealed from both the broadband and narrowband images. 
342: Specifically, we describe the DIG morphology visible in the emission line 
343: images. We also comment on the dust features visible in the broadband 
344: images and on a newly detected dwarf irregular galaxy in the halo of NGC\,4634.
345: 
346: The layout of the images is as follows. In \S~\ref{ss:broadimage} we present the 
347: three-color images (Figure~\ref{f:colorfigures}), which show both the stellar and the gas 
348: distribution for all of the galaxies. In \S~\ref{ss:narrowimage} we present the emission 
349: line images (Figures~\ref{f:n4634sixpanel}, \ref{f:n4700sixpanel}, \ref{f:n5775sixpanel} 
350: and \ref{f:n7090sixpanel}). Each of the four figures consists of six images. The overview 
351: images are shown in the upper left corner, and the individual enlargements in the remaining 
352: five subpanels. For NGC\,4700 we show an enlargement of the three-color image 
353: (Figure~\ref{f:n4700acs}), revealing the central region. The unsharp-masked images (tracing 
354: the dust distribution), which are discussed in \S~\ref{ss:dust}, are shown in a separate 
355: figure (Figure~\ref{f:dust4}), as well as the dIrr galaxy near NGC\,4634 
356: (Figure~\ref{f:dIrr}), presented in \S~\ref{ss:dIrr}.
357: 
358: \subsection{ACS broadband imagery}
359: \label{ss:broadimage}
360: 
361: We have combined the broadband and narrowband images to better display the 
362: interplay of gas and stars in the disk-halo region. Our three-color 
363: images (F625W = blue; F658N = green; continuum subtracted F658N = red) of 
364: NGC\,4634, NGC\,4700, NGC\,5775 and NGC\,7090 are presented in Figure~\ref{f:colorfigures}. 
365: They provide us with an overview of the disk morphology, in particular the stellar 
366: distributions (i.e. star densities) within each galaxy. This information can later be used 
367: to investigate potential dependencies of halo properties on the {\it local} level of disk 
368: activity.
369: 
370: \paragraph{NGC\,4634}
371: Star forming regions (and regions of increased stellar densities) are quite 
372: wide-spread throughout the disk of NGC\,4634. The broadband image shows a 
373: number of \ion{H}{2} regions shining through the galaxy's dust lane. Around 
374: the high surface-brightness inner disk fainter ``emission'' is found, probably 
375: from stars in outer spiral arms that may be slightly warped. No significant 
376: stellar overdensities are found in these outer regions.
377: 
378: \paragraph{NGC\,4700}
379: The stellar disk of NGC\,4700 looks more irregular than that of
380: NGC\,4634. The dust morphology is patchy, rather than forming a
381: continuous band along the disk plane. The most prominent star concentrations 
382: are found in the central area. The northeastern peak of the stellar 
383: distribution has a higher stellar density, than the southwestern peak. 
384: Both regions seem to be more or less equidistant from the center of the 
385: galaxy. There is one particular region located within the disk that has quite 
386: prominent extinction by dust. It is also interesting to note that the patchy 
387: dust lane is slightly skewed with respect to the stellar disk plane. 
388: 
389: \paragraph{NGC\,5775}
390: NGC\,5775 has a well-defined disk with evidence of widespread, 
391: high-level star formation, as indicated by copious regions of 
392: increased stellar density. Its dust lane is similar to that of
393: NGC\,4634 and more pronounced than that of NGC\,4700. The dust lane is 
394: most pronounced in the outer disk regions (i.e. at large galactocentric 
395: radii). However, NGC\,5775 has a relatively compact nuclear region, that 
396: has the bulk of the stellar densities. 
397: 
398: \paragraph{NGC\,7090}
399: NGC\,7090's disk is similar to that of NGC\,4700, with a slightly more 
400: prominent dust lane and an apparently wider, more amorphous overall 
401: appearance. The dust lane is very filamentary. Most parts of the dust lane 
402: are located in the regions above the midplane. Not much dust is present in the 
403: southern part of the disk (below the midplane). The distribution of its stellar 
404: component is asymmetric. The maximum is located to the southeast of the center, but 
405: there are several peaks of stellar density distributions along the disk.  
406: 
407: 
408: \subsection{ACS emission line images}
409: \label{ss:narrowimage}
410: 
411: As mentioned above, these images, called ``H$\alpha$'' images, 
412: include unknown contributions from the adjacent [\ion{N}{2}] lines and 
413: thus trace the warm interstellar emission line gas in general. Both the 
414: disk and the extraplanar gas can be studied in all four target 
415: galaxies. Images showing the full extent in the ACS field-of-view are shown for
416: each galaxy in the upper left panel of Figures~\ref{f:n4634sixpanel}, 
417: \ref{f:n4700sixpanel}, \ref{f:n5775sixpanel} and \ref{f:n7090sixpanel}. 
418: 
419: In addition, close-up presentations of individual areas in the four
420: target galaxies are presented below in the five remaining sub-panels of each 
421: figure. We show five individual fields for each galaxy. Comments on the observed 
422: morphology are summarized in Table~\ref{t:morph}. We also include some information 
423: on previous ground-based studies and related disk-halo investigations from other 
424: wavelength regimes that have a direct connection to our present study. 
425: 
426: \paragraph{NGC\,4634}
427: 
428: This Virgo cluster member edge-on spiral was previously imaged in the DIG 
429: context in H$\alpha$ using ground-based data \citep{ros00}. A diffuse halo 
430: was detected with an extent of 1.2\,kpc above/below the galactic plane. 
431: 
432: Extraplanar emission was also confirmed spectroscopically \citep{tul00a,ott02}. 
433: In addition to revealing eDIG, extraplanar dust was also detected at high 
434: galactic latitudes of about 1\,kpc \citep{how99,ros03b}. 
435: 
436: As already discussed in \citet{ros03a}, the presence of an H$\alpha$ halo 
437: is almost always accompanied by a hot gaseous (X-ray) halo, as high SF rates 
438: and high $S_{60}/S_{100}$ are strong indicators of halo emission. Not 
439: surprisingly, a hot tenuous and far more extended X-ray gaseous halo was found 
440: in NGC\,4634 in the soft energy band, making use of XMM-Newton observations 
441: \citep{tul06}. A unique feature among the studied sample of nine edge-on 
442: spirals is the absence of hard X-ray emission in NGC\,4634. 
443: 
444: Our ACS H$\alpha$ image of NGC\,4634 (see Figure~\ref{f:n4634sixpanel}) shows widely 
445: distributed star formation in its disk, as already suggested by the broadband 
446: image introduced above. Individual \ion{H}{2} regions are seen through gaps in 
447: the dust lane. But there are no prominent H$\alpha$ bubbles visible.
448: 
449: In the halo extended H$\alpha$ emission is found that is, despite
450: the superb spatial resolution of {\it HST}/ACS, unresolved (see 
451: Figure~\ref{f:n4634sixpanel}). No filaments, loops or similar structures 
452: are discernible, apart from a few minor exceptions at large galactocentric radii. 
453: 
454: \paragraph{NGC\,4700}
455: 
456: This southern edge-on spiral was previously imaged by means of ground-based 
457: H$\alpha$ observations \citep{ros03b}. A bright halo with superimposed filaments 
458: reaching into the halo out to $|z| \approx 2-3$\,kpc was detected. The filaments 
459: appear to emanate from the brightest SF regions, anchored within the disk. 
460: 
461: Also, a radio-continuum halo was detected \citep{dah01}, revealing similar 
462: shapes and maxima at the optical SF counterparts. The extent of the 
463: cosmic-ray halo is much larger than the optical halo. 
464: 
465: Spectra of the galaxy itself have been published by \citet{kew01}, which 
466: show besides prominent H$\alpha$ and [\ion{N}{2}] emission relatively 
467: strong [\ion{O}{3}] emission lines. However, \citet{kew01} do not assign 
468: a classification for this galaxy, which they list as IRAS\,$12465-1108$. 
469: An imaging HST survey of \citet{mal98} lists NGC\,4700 as an \ion{H}{2} galaxy.
470: Spectral classifications in the NASA Extragalactic Database (NED) range 
471: from \ion{H}{2} region to Sy\,2 type, so there is not a clear-cut classification 
472: available for this galaxy as of yet.
473: 
474: NGC\,4700 (Figure~\ref{f:n4700sixpanel}) displays an enormous wealth of H$\alpha$ 
475: emitters in its disk plane with a large number of bright star forming regions 
476: and many bubbles interspersed. Contrary to the rather irregular appearance
477: of its stellar disk in the broadband image (cf. \S~\ref{ss:broadimage}), 
478: the H$\alpha$ emission distribution looks much more confined to what 
479: looks like a thin disk. The ACS H$\alpha$ image (specifically the central 
480: region displayed in Figure~\ref{f:n4700acs}) is by far the most spectacular one 
481: in our sample. Given the distance of NGC\,4700, which is very similar to 
482: NGC\,5775, this represents no bias in terms of resolution (i.e. proximity). 
483: The H$\alpha$ image reveals a very complex morphology. There are numerous faint 
484: filaments protruding into the halo region. 
485: 
486: The circumnuclear region does not host strong SF activity, and it seems as if 
487: the bulk of the emission has been ejected into two lobe regions. These are, 
488: despite the higher densities, located within the galactic plane (roughly 
489: aligned with the disk plane) and not perpendicular to it. Nonetheless, considerable 
490: structure (e.g.,\,filaments) is also revealed perpendicular to the disk north 
491: and south of the nuclear region. Currently we do not have any kinematic 
492: information of the lobe regions and the galaxy in general, so we can only 
493: speculate about the possible origin of the H$\alpha$ lobes. But it seems most 
494: likely as if these were created by a powerful mass and energy ejection from the 
495: nuclear region, which may host an AGN. 
496: 
497: On top of the disk emission an equally astounding wealth of extraplanar
498: H$\alpha$ emission is found, with a large number of long, thin filaments
499: emanating vertically from the central disk plane. Several of these filaments
500: can be traced back to individual knots of H$\alpha$ emission in the central 
501: part of the disk, suggesting that the gas is energized from within the disk 
502: plane. The majority of the long filaments are emanating from and around 
503: regions close to the nucleus. However, we note that from optical observations 
504: (both R-band and H$\alpha$) no clear indication of a nucleus can be assessed. 
505: 
506: Instead two giant SF complexes (almost equidistant from the center) are 
507: located within the disk plane, which resemble in appearance the lobe 
508: regions of a radio jet (on smaller scales, though). However, no jet is visible, 
509: nor is there any clear indication of a distinct nucleus. These shell-like structures 
510: are located within the disk, and show a radially expanding structure within the 
511: disk, where the pressure gradient is much higher as opposed to moving perpendicular 
512: to it (such as the winds in nucleated starbursts). The very edges of the disk show 
513: substantial amounts of SF. Many superbubbles and shells are detected, and there does 
514: not appear to be an abrupt cut-off in star formation within the disk. The expanding 
515: shells seem indicative of an AGN (Seyfert type), however no clear classification has 
516: been established. 
517:  
518: \paragraph{NGC\,5775}
519: 
520: NGC\,5775 is a well studied edge-on spiral galaxy in the disk-halo context, 
521: which was imaged extensively by ground-based observations in H$\alpha$ 
522: \citep{leh95,col00,tul00,ros03b}. Extraplanar emission was imaged out 
523: to $|z|\lta$\,5\,kpc above the galactic plane. It is the galaxy with the 
524: highest $L_{\rm{FIR}}/D^2_{25}$ ratio in our current sample and it is a 
525: starburst-type galaxy, although it is morphologically quite different from 
526: that of nucleated starbursts. A gaseous halo with individual H$\alpha$ spurs or 
527: filaments, associated with the X-shaped structure of the large scale magnetic 
528: field, was discovered \citep{tul00}. 
529: 
530: EDIG has been detected spectroscopically even out to $|z|\sim$\,9\,kpc 
531: \citep{ran00,tul00}. In addition, extended radio continuum emission has been 
532: detected as well \citep{hum91,dur98,tul00}. A tenuous X-ray halo, spatially 
533: coexistent (and more extended) than the radio continuum and H$\alpha$ halo, 
534: was detected with Chandra \citep{str04} and with XMM-Newton \citep{tul06}. 
535: NGC\,5775 appears to be interacting with its face-on companion NGC\,5774, which 
536: was first indicated by \ion{H}{1} measurements \citep{irw94}.
537: 
538: In NGC\,5775 quite a large number of prominent \ion{H}{2} regions is found 
539: in its disk plane, together with a number of bubbles 
540: (see Figure~\ref{f:n5775sixpanel}). The disk is surrounded by extended H$\alpha$ 
541: emission that is not resolved into individual structures. The extraplanar 
542: H$\alpha$ emission is strongest and widest in the central area, where the level 
543: of star formation in the disk is also the highest, but appears to be truly 
544: diffuse. The high spatial resolution of our {\it HST}/ACS images does not allow 
545: for the detection of the very faint surface emission of the filaments. Unfortunately, 
546: the background artifacts discussed in \S~\ref{ss:datared} did not allow to bin the data.
547: 
548: \paragraph{NGC\,7090}
549: 
550: This southern edge-on spiral is the galaxy with the smallest 
551: $L_{\rm{FIR}}/D^2_{25}$ in our current sample, and it was imaged in 
552: H$\alpha$ before, using ground-based data \citep{ros03b}. These images already 
553: revealed an inhomogeneous layer of extended DIG, which seems to correlate spatially 
554: with the star formation regions within the disk. In addition to this H$\alpha$ study, 
555: an extended radio continuum halo was detected as well \citep{dah01}, based on ATCA 
556: observations with different configurations at frequencies of 1.43\,GHz and 2.45\,GHz. 
557: The measured scale-heights for these observations range from 1.19 to 2.16\,kpc. 
558: 
559: The overview image of our ACS H$\alpha$ emission line image of NGC\,7090 (see 
560: top left panel of Figure~\ref{f:n7090sixpanel}) confirms the asymmetry of the 
561: star formation activity in its disk very clearly. To the east of the nuclear 
562: region there is a big apparent gap in the H$\alpha$ emission distribution. In 
563: general, the eastern half of the galaxy disk is considerably less active than the 
564: western one. Part of this can be attributed to dust, which is prominently visible 
565: in the broadband image.
566: 
567: The image also reveals that there are not just \ion{H}{2} regions 
568: (as traced by diffuse, high surface-brightness H$\alpha$ emission) 
569: in the disk, but lots of individual, resolved bubbles in addition. 
570: These are not only present in the more active western half, but also 
571: to the east, where a few individual bubbles are found out to large 
572: galactocentric radii.
573: 
574: Images of individual regions in H$\alpha$ reveal the presence of a multitude 
575: of conspicuous, {\it resolved} individual structures, including partial 
576: loops and filaments perpendicular to the disk plane. A few of the extended 
577: filaments are detected in the southern part of the galaxy. These connect 
578: visibly to the bright star forming regions in the disk (see middle right panel 
579: of Figure~\ref{f:n7090sixpanel}). 
580: 
581: It is obvious that the H$\alpha$ emission distribution is apparently
582: widest above regions with the most active star formation in the disk.
583: There are a multitude of bubbles and supershells detected (see 
584: top right and middle left panels of Figure~\ref{f:n7090sixpanel}).
585: 
586: \subsection{Dust features in the ACS broadband images}
587: \label{ss:dust}
588: 
589: Filtering the low spatial frequencies out of the F625W broadband images, 
590: one can study the smaller-scale absorption features caused by dust in front 
591: of background stellar emission.
592: 
593: Apart from the gaseous constituents of the ISM in galaxies, dust is also 
594: present on both small and large scales. To understand the morphology of
595: the extraplanar ISM in our target galaxies, it is interesting to compare 
596: the structure of extraplanar dust to that of the gaseous emission filaments. 
597: For this purpose, similarly to the previously studied cases, we have created 
598: unsharp-masked versions of the broadband F625W (SDSS $r$) images to enhance 
599: the contrast between those absorbing high spatial frequency dust structures 
600: and the smoother distribution of the background light from the galaxy's
601: stellar populations.  
602: 
603: Despite the fact that our target galaxies were selected to have a less dusty plane 
604: than galaxies such as NGC\,891 \citep{how97,ros04}, there are still sufficient 
605: numbers of small-scale dust filaments for such a comparison. Due to the fact 
606: that we only obtained broadband images in one waveband (the basic reason being 
607: that these are only used for the continuum subtraction), we cannot quantify 
608: the physical properties of the dust. We therefore only qualitatively describe the 
609: extraplanar dust content in these galaxies with respect to the eDIG morphology.  
610: 
611: The images are shown in Figure~\ref{f:dust4}. They reveal that the 
612: distribution of the dust filaments is also quite different among the four 
613: target galaxies. NGC\,4634 and NGC\,5775 show the most prominent features, 
614: whereas NGC\,4700 and NGC\,7090 have the least extraplanar dust features. 
615: For NGC\,4700, in particular, but also to some extent for NGC\,7090, the presence 
616: of many resolved stars in the disk make an analysis relatively difficult, as the 
617: smoothing process creates some artifacts around the stars, which overlap prominently 
618: in the crowded field. Therefore, a direct one-to-one comparison of dust filaments 
619: with emission line features (e.g., filaments) is only possible for NGC\,4634 and NGC\,5775.
620: 
621: However, as presented in \S~\ref{ss:narrowimage}, there are a few large-scale 
622: H$\alpha$ filaments discernible at high galactic latitudes 
623: in NGC\,4634 and NGC\,5775. The distribution of dust structures in NGC\,4634 
624: show both alignment parallel and perpendicular to the disk, whereas 
625: generally at higher $|z|$ the dust (sometimes very filamentary) is aligned 
626: perpendicular to the disk. But it should be noted that this is only true 
627: for the western side of the galaxy. The distribution of dust features on 
628: the other side (east of the center) is somewhat more confined to the disk and 
629: oriented more parallel than perpendicular, but mostly a random mix of 
630: orientations. It should also be noted that the dust features are not 
631: correlated on a one-to-one basis with the H$\alpha$ filaments. This was 
632: already stated earlier \citep{how99,ros03b}, based on groundbased observations. 
633: 
634: 
635: \subsection{Dwarf irregular galaxy in the halo of NGC\,4634}
636: \label{ss:dIrr}
637: 
638: With our high spatial resolution {\it HST}/ACS observations, we are able to 
639: re-investigate the extraplanar blob, detected by us in the halo of NGC\,4634, 
640: which was named Patch\,1 \citep{ros00}. We had concluded from our previous 
641: ground-based observations that this patch may either be an extraplanar SF 
642: region or a dwarf galaxy in the halo of NGC\,4634, given that the FWHM of the 
643: used filter excluded the possibility of a foreground or background object. 
644: 
645: Our deep {\it HST}/ACS broadband and narrowband observations now clearly indicate 
646: that this patch is a galaxy. The narrowband F658N observations reveal a complex system 
647: composed of several knots associated with gas, reminiscent of a dwarf 
648: irregular (dIrr) galaxy (see Figure~\ref{f:dIrr}). The broadband image reveals 
649: several individual star associations that are resolved in this dIrr galaxy, 
650: and the major morphology looks reminiscent of a satellite merger galaxy, 
651: showing hints of tidal streams owing to the gravitational attraction from 
652: NGC\,4634. Given the peculiar morphology, this dIrr galaxy is possibly already 
653: in the early process of producing a minor merger with its parent galaxy 
654: (NGC\,4634). It should be noted that the broadband image shows a very thick 
655: stellar disk of NGC\,4634, unlike any of the other studied edge-on galaxies. 
656: This is consistent with scenarios in which disk thickening is the result of
657: a previous interaction, possibly previous minor mergers \citep[e.g.,][]{yoa06,wys06}. 
658: 
659: Augmented longslit spectra (R.-J. Dettmar et al. 2008, in preparation) seem 
660: to strengthen this interpretation of Patch\,1 being a dIrr galaxy, which is 
661: deficient in [\ion{N}{2}] and [\ion{S}{2}], based on emission line 
662: diagnostics. Henceforth, we label this galaxy as $\rm{J}124239.58$+141751.83, 
663: which refers to the coordinates of the brightest H$\alpha$ emission region 
664: in its central part. We will report on the physical properties and stellar 
665: populations of this interesting dIrr galaxy in detail in a future paper.   
666: 
667: \subsection{Ancillary data}
668: 
669: The {\it HST}/ACS data presented above were obtained as part of a
670: multi-wavelength observing campaign. For the galaxies presented here
671: we have other, independent measurements of various phases of the ISM
672: in their halos published by us previously, including X-ray data 
673: \citep[][Ehle et al.~2008, in preparation]{tul06} and radio-continuum 
674: data \citep{tul00,dah01}. 
675: %
676: A correlation study of the ACS data with other tracers of extraplanar 
677: gas in this sample galaxies will be presented separately, as some of the  
678: observations are still pending. 
679: 
680: 
681: \section{DISCUSSION}
682: \label{s:discuss}
683: 
684: \subsection{EDIG morphology as a function of SF activity}
685: \label{ss:morphvssfprop}
686: 
687: Over roughly the past decade and a half evidence has been mounting that
688: extended gaseous halos around late-type spiral galaxies are formed as a 
689: consequence of the star formation activity in the underlying disks. A high 
690: IRAS $S_{60}/S_{100}$ integral flux ratio $\geq 0.4$, which is a measure 
691: of a high mean dust temperature in galaxies, was identified as a very reliable 
692: indicator of massive star formation \citep{hec90}.
693: 
694: In such a scenario, emission features in gaseous halos are expected to be 
695: closely tied to the energy input rate, which is required to exceed a threshold 
696: to drive outflows perpendicular to the disk plane in the first place. This 
697: was first addressed by \citet{dah95} for radio data and later derived from 
698: optical narrowband imagery \citep{ran96,ros03a}. 
699: %
700: More recently, \citet{dah06} have shown that there is a correlation of the 
701: existence of galaxy halos (detected in radio continuum and H$\alpha$ 
702: observations) not only with the global energy injection rate, but also with 
703: the {\it local} energy input level in a galaxy and, in addition, with its mass 
704: density. We will investigate in the following whether the eDIG properties of 
705: the galaxies in the present sample can be explained in this context.
706: 
707: The two galaxies with filamentary eDIG structures are NGC\,4700 and NGC\,7090.
708: As tabulated in Table~\ref{t:prop}, these are the two galaxies at the lower end 
709: of the energy input range ($L_{\rm FIR}/D_{25}^2$). On the other hand, NGC\,4634 
710: and NGC\,5775 display only unresolved extraplanar H$\alpha$ emission. These two 
711: galaxies have higher energy input rates from massive star formation in their disks. 
712: Accordingly, one might expect them to have a larger number of individual sites of 
713: star formation and thus also a larger amount of associated halo emission.
714: 
715: One might naively expect this difference to be an observational artifact,
716: based on the assumption that physically larger and more massive galaxies
717: just have more of everything. More star formation regions lead to more
718: filaments emanating from the disks, potentially leading to a quasi-diffuse
719: eDIG structure caused by the superposition of many individual filaments.
720: %
721: However, there are no indications of filaments being detected in the more
722: massive galaxies of the present sample (NGC\,4634 and NGC\,5775) and also
723: other galaxies studied previously, in areas where the spatial density 
724: of star-forming regions decreases, e.g., near the radial edges of their disks. 
725: The difference in eDIG morphologies is equally unlikely to be caused by a
726: distance bias, because NGC\,4700 (with a filamentary eDIG) and NGC\,5775
727: (with a diffuse eDIG) are at roughly equal distances (see Table~\ref{t:obs}). 
728: Again, this makes it unlikely that the diffuse eDIG is the result of an
729: unresolved superposition of many filaments.
730: 
731: A first indication at what might cause this difference can be gleaned by
732: studying various properties of the sample galaxies. NGC\,4634 and NGC\,5775 
733: are the two galaxies with diffuse eDIG, one of them is also the most massive 
734: galaxy in the present sample (see Table~\ref{t:prop}). From the width of the 
735: \ion{H}{1} emission line we derive an estimated total mass for NGC\,5775 of about 
736: $1.91\times10^{11}\,{\rm M}_\odot$. Interestingly though, in contrast, the 
737: mass of NGC\,4634 is almost a factor of two below that of NGC\,7090. So this 
738: complicates the interpretation that the mass alone is a sensitive indicator 
739: for the morphology of eDIG. 
740:  
741: The general trend and the earlier work by \citet{dah06} indicate that the mass 
742: (or mass density) of a galaxy is an important control parameter for the 
743: morphology of the eDIG. Low mass galaxies such as the dwarf NGC\,55 
744: \citep[e.g.,][]{tul03}, NGC\,2188 \citep{dom97} and the blue compact 
745: dwarf galaxy IZw18 \citep[e.g.,][]{hun95} show filamentary structure, whereas 
746: the more massive galaxies (with widespread star formation, not the powerful 
747: nucleated starbursts; e.g., NGC\,891 \citep{ros04}) tend to have relatively 
748: smooth layers of eDIG. An explanation for this could be the fact that the gas 
749: can overcome the gravitational potential wells of the galaxies more easily. 
750: The range of masses spanned by our sample is rather small (see Table~\ref{t:prop}), 
751: so it may not be surprising that any trends with mass are less clear in our sample 
752: than those observed in samples that contains both dwarf and giant galaxies.
753: 
754: Of our targets only NGC\,7090 and, to a lesser degree, NGC\,4700, contain bubbles, shells 
755: or supershells. In NGC\,7090 a large number of such features, of various sizes, are found.
756: Also, the SF regions in this galaxy are loosely scattered within the disk, and the major 
757: outflows are confined to only one side above the disk (see descriptions of the eDIG 
758: morphology in the individual regions within each galaxy listed in Table~\ref{t:morph}). 
759: 
760: Among our studied sample, the resolved small-scale DIG halo morphology is showing 
761: a clear dependence on the star formation rate per unit area $(L_{\rm{FIR}}/D^2_{25})$. 
762: That is, galaxies with $L_{\rm{FIR}}/D^2_{25} \geq 10$ (in units of 
763: $\rm{10^{40}\,erg\,s^{-1}\,kpc^{-2}}$), show a smooth, diffuse DIG halo, as opposed 
764: to the galaxies with values $\leq 6$, which show primarily filamentary DIG. The studied 
765: galaxies thus can be arranged into a {\it sequence} of eDIG diffuseness as a function 
766: of $L_{\rm{FIR}}/D^2_{25}$. More generally speaking, there is a morphological DIG sequence 
767: among SF galaxies observed evolving from galaxies which reveal mostly bubbles and 
768: supershells to galaxies which show filamentary emission and to cases which have smooth 
769: eDIG layers (i.e. halos). Where there are bubbles (i.e. enclosed structures) the breakout 
770: condition cannot have been met yet.
771: 
772: \subsection{The role of the intergalactic environment}
773: \label{ss:environ}
774: 
775: Although it has been established that eDIG originates from the SF regions 
776: within the disk, the galactic environment in which the individual galaxies 
777: reside, is of course of importance, too. It may, e.g., contribute to the 
778: extent of the emission that is detected. 
779: 
780: Of the four galaxies studied here, two (NGC\,4634 and NGC\,5775) have nearby
781: companion galaxies (i.e. they form galaxy pairs). NGC\,4700 and NGC\,7090, in 
782: contrast, are rather isolated spirals. 
783: 
784: Our current sample is not large enough to draw any statistical conclusions.
785: But since one galaxy from each category has a filamentary and the other a 
786: diffuse eDIG morphology, the presence of a nearby companion does not appear 
787: to dominate the eDIG properties in these cases. However, significant 
788: exceptions may be expected in dense group environments, where galaxies 
789: experience effects such as ram pressure stripping \citep[e.g.,][]{ken99}. 
790: 
791: 
792: \subsection{Clues from observations at other wavebands}
793: \label{ss:multiwave}
794: 
795: It is important to ask whether or not there seems to be a more global connection 
796: between the various ISM constituents that are responsible for driving a kiloparsec 
797: outflow or wind. Obviously, there is a connection linking the various ISM phases 
798: as described for modest galaxy samples \citep[e.g.,][]{tul06,dah06}, but the 
799: details are far from being well understood. 
800: 
801: From the sample of \citet{tul06}, it was concluded that the presence of a diffuse 
802: ionized gas halo was always accompanied by a grand scale X-ray halo, which 
803: generally revealed that the hot tenuous gas is far more extended than the DIG. 
804: In those cases also a cosmic ray halo, traced by radio-continuum observations, was 
805: discovered. Linking the cosmic-ray halo to the DIG halo, showed clear evidence 
806: that the mass surface density was a good proxy for the presence of a gaseous halo. 
807: The mass of the stellar disk ($M_{\rm K}$), derived from K-band observations, is divided 
808: by the area of the optical disk ($A_{25} = \pi r_{25}^2$), which yields the mass surface 
809: density. 
810: 
811: For NGC\,4700 a value of $M_{\rm K}/A_{25} = 17.4\,{\rm M_{\sun}\,pc}^{-2}$ was 
812: determined \citep{dah06}. Despite having higher values than the more quiescent 
813: galaxies, this is a rather modest number compared to local starburst galaxies 
814: which can have mass surface densities of up to eight times as high.
815: 
816: However, does the mass density critically control whether a galaxy has a 
817: kiloparsec outflow or rather a more filament-driven DIG morphology? Our present 
818: sample is certainly too small to answer this question unambiguously, as two galaxies 
819: each fall into the category of either having a filamentary DIG morphology, or showing 
820: a more smooth DIG layer. 
821: 
822: \subsection{Observations confront Theory: Where have all the Chimneys gone?}
823: \label{ss:obsvstheo}
824: 
825: One of the greatest challenges any theoretical concept poses is how 
826: accurately it can reflect or predict the observational details. In terms of 
827: the chimney model \citep{nor89}, past observations from different wavebands, 
828: both in galactic and extragalactic studies, have failed to reveal many of 
829: these structures that were predicted. More specifically, the number of 
830: predicted chimneys is of the order of several hundred in a galaxy of modest 
831: SF activity. Yet, only very few chimneys have been detected, if any at all. 
832: So the prediction falls short by at least two orders of magnitudes. Previous 
833: reports include a Galactic chimney in the Perseus arm \citep{nor96} and 
834: the Galactic chimney GSH\,$277+00+36$ \citep{mcg03} as well as evidence for 
835: a chimney breakout in the Galactic supershell GSH\,$242-03+37$ \citep{mcg06}. 
836: Extragalactic chimneys or structures identified as such so far have been 
837: observed only in the starburst galaxy M\,82 \citep{wil99,gar01}. But we have 
838: to carefully consider whether this is due to an observational bias or attributed 
839: to certain physical contraints. 
840: 
841: Previously, NGC\,891 was imaged in H$\alpha$ \citep{ros04}, where no evidence 
842: was found for the chimney mode. However, the situation in NGC\,891 is 
843: complicated by the fact that it has a very prominent dust layer, which 
844: possibly could have obscured any chimney structures (i.e. chimney walls) 
845: in the lower disk-halo interface ($|z| \leq 500$\,pc). Therefore, those 
846: structures could have gone undetected even if they existed. Our currently 
847: studied galaxies by and large have a much less dusty galactic midplane, so 
848: the effects of dust obscuration should be considerably mitigated.  
849: 
850: However, from our imaging studies it is obvious that there are no clear 
851: indications for the chimney mode in these actively star forming galaxies. 
852: In NGC\,4634 and NGC\,5775 there are no such structures observed in H$\alpha$. 
853: Are there indications of other related phenomena? Evidently, there are 
854: several features (e.g.,\, curved filaments) visible whose morphology show an almost 
855: closed loop, indicative of the galactic fountain scenario. However, without 
856: detailed knowledge of the kinematics (e.g.,\,mapping of the velocity field) 
857: we cannot say what parts of the looped filament are still moving upward and 
858: what parts possibly are already descending. Furthermore, multiple episodes of SF 
859: driven outflows may have altered the geometry of the outflows along the line of 
860: sight. Hence the chimney structures may look like sheets, or picked up more 
861: complex structures generated by overlapping filaments. Nonetheless, it seems 
862: obvious that these features are associated with an SF driven outflow. 
863: 
864: Recent models by \citet{tas06} simulated the star formation in galactic disks 
865: involving feedback by SNe. The density, temperature, and pressure maps (see 
866: their Figure~6) show copious amounts of warm and hot gas being driven into the 
867: halo regions, extending out to a few kiloparsecs. In particular, those simulations 
868: look very reminiscent to the morphology seen in our H$\alpha$ image of 
869: NGC\,4700 on a global scale. More recent models of a galactic wind as seen in 
870: starburst galaxies \citep{coo08} replicate very detailed structures including the 
871: filamentary morphology seen as for instance in M\,82. While considerable progress 
872: has been achieved in modeling even small scale structures in a magnetized ISM 
873: \citep[e.g.,][]{avi05}, more refined simulations are necessary in order to 
874: compare them to the high spatial resolution multi-structured eDIG observations, 
875: as those presented in this study.
876: 
877: 
878: \section{SUMMARY}
879: \label{s:sum}
880: 
881: We have observed four edge-on spiral galaxies with different levels of star
882: formation in H$\alpha$ with the {\it HST}/ACS to study the morphology of 
883: their extraplanar DIG (eDIG). 
884: %
885: The distribution of the eDIG is very different among the sample galaxies. 
886: While the eDIG is very filamentary in NGC\,4700 and in NGC\,7090, the 
887: distribution of the eDIG in NGC\,4634 and NGC\,5775 is remarkably smooth. 
888: %
889: The latter two galaxies have the highest SF rate per unit area in our sample 
890: and NGC\,5775 is also the most massive galaxy in the sample. 
891: 
892: Based on our current and earlier results, we conclude that the observed
893: differences are probably a consequence of their different levels of SF activity 
894: {\it and} their total masses and/or mass densities. 
895: %
896: Whereas the SF rate per unit area controls the presence (e.g., the radial
897: extent), shape and amount of eDIG in a galaxy's halo, the total mass or mass density
898: seems to be at least in part responsible for the actual shape of the eDIG. This is 
899: corroborated by a comparison of the morphology of the lower-mass galaxies to those of 
900: SF dwarf galaxies such as NGC\,55 or the blue compact dwarf IZw18. The more 
901: massive galaxies with active SF seem to expell their gaseous outflows more 
902: globally and diffusively.    
903: 
904: The studied galaxies can be arranged into a {\it sequence} of eDIG diffuseness 
905: as a function of star formation activity per unit area. More generally speaking, 
906: there is a morphological DIG sequence among SF galaxies, which evolves from 
907: galaxies which reveal mostly bubbles and supershells to galaxies which show 
908: filamentary emission to cases of intense SF which reveal smooth eDIG layers 
909: (i.e. halos).
910: 
911: %%%%%%%%%%%%%%%
912: % Acknowledgments
913: %%%%%%%%%%%%%%%
914: 
915: \acknowledgments Support for proposal 10416 was provided by NASA through a 
916: grant from the Space Telescope Science Institute, which is operated by the 
917: Association of Universities for Research in Astronomy, Inc., under NASA 
918: contract NAS 5-26555. This project has been supported at Ruhr-University Bochum 
919: by DLR grant 50 OR 0503. This research has made use of the NASA/IPAC Extragalactic 
920: Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute 
921: of Technology, under contract with the National Aeronautics and Space Administration. 
922: This research has made use of NASA's Astrophysics Data System Bibliographic Services. 
923: We also made use of the Lyon Extragalactic database (LEDA).
924: 
925: \clearpage
926: 
927: %%%%%%%%%%%%%%%
928: % Reference List
929: %%%%%%%%%%%%%%%
930: 
931: \begin{thebibliography}{}
932: 
933: \bibitem[Bregman(1980)]{bre80}
934: Bregman, J. N. 1980, \apj, 236, 577
935: 
936: \bibitem[Cecil et al.(2001)]{cec01} 
937: Cecil, G., Bland-Hawthorn, J., Veilleux, S., \& Filippenko, A. V. 2001, 
938: \apj, 555, 338
939: 
940: \bibitem[Cecil et al.(2002)]{cec02} 
941: Cecil, G., Bland-Hawthorn, J., \& Veilleux, S. 2002, \apj, 576, 745
942: 
943: \bibitem[Collins et al.(2000)]{col00} 
944: Collins, J. A., Rand, R. J., Duric, N., \& Walterbos, R. A. M. 2000, \apj, 
945: 536, 645
946: 
947: \bibitem[Cooper et al.(2008)]{coo08}
948: Cooper, J. L., Bicknell, G. V., Sutherland, R. S., \& Bland-Hawthorn, J. 2008, 
949: \apj, 674, 157 
950: 
951: \bibitem[Dahlem(1997)]{dah97} 
952: Dahlem, M. 1997, \pasp, 109, 1298
953: 
954: \bibitem[Dahlem et al.(1994)]{dah94} 
955: Dahlem, M., Dettmar, R.-J., \& Hummel, E. 1994, \aap, 290, 384
956: 
957: \bibitem[Dahlem et al.(1995)]{dah95} 
958: Dahlem, M., Lisenfeld, U., \& Golla, G. 1995, \apj, 444, 119
959: 
960: \bibitem[Dahlem et al.(2006)]{dah06}
961: Dahlem, M., Lisenfeld, U., \& Rossa, J. 2006, \aap, 457, 121
962: 
963: \bibitem[Dahlem et al.(2001)]{dah01}
964: Dahlem, M., Lazendic, J. S., Haynes, R. F., Ehle, M., \& Lisenfeld, U. 2001, 
965: \aap, 374, 42
966: 
967: \bibitem[de Avillez(2000)]{avi00} 
968: de Avillez, M. A. 2000, \mnras, 315, 479  
969: %Disc-halo interaction - I. Three-dimensional evolution of the Galactic disc
970: 
971: \bibitem[de Avillez \& Breitschwerdt(2005)]{avi05} 
972: de Avillez, M. A., \& Breitschwerdt, D. 2005, \aap, 436, 585
973: %%Global dynamical evolution of the ISM in star forming galaxies. 
974: %%I. High resolution 3D simulations: Effect of the magnetic field
975: 
976: \bibitem[Dettmar(1990)]{det90} 
977: Dettmar, R.-J. 1990, \aap, 232, L15
978: 
979: \bibitem[Dettmar(1992)]{det92} 
980: Dettmar, R.-J. 1992, Fund. Cosm. Phys., 15, 143
981: 
982: \bibitem[Domg\" orgen \& Dettmar(1997)]{dom97}
983: Domg\" orgen, H. \& Dettmar, R.-J. 1997, \aap, 322, 391
984: 
985: \bibitem[Duric et al.(1998)]{dur98} 
986: Duric, N., Irwin, J. A., \& Bloemen, H. 1998, \aap, 331, 428
987: 
988: \bibitem[Ferguson et al.(1996)]{fer96}
989: Ferguson, A. M. N., Wyse, R. F. G., Gallagher, J. S., \& Hunter, D. A. 
990: 1996, \aj, 111, 2265
991: 
992: \bibitem[Garc\'{\i}a-Burillo et al.(2001)]{gar01}
993: Garc\'{\i}a-Burillo, S., Martin-Pintado, J., Fuente, A., \& Neri, R. 2001, 
994: \apj, 563, L27 
995: 
996: \bibitem[Giavalisco(2004)]{gia04}
997: Giavalisco, M. 2004, Instrument Science Report ACS 2004-12, 
998: (Baltimore: STScI) 
999: 
1000: \bibitem[Gonzaga et al.(2005)]{gon05}
1001: Gonzaga, S., et al. 2005, ACS Instrument Handbook, Version 6.0, 
1002: (Baltimore: STScI)
1003: 
1004: \bibitem[Heckman et al.(1990)]{hec90}
1005: Heckman, T. M., Armus, L., \& Miley, G. K. 1990, \apjs, 74, 833
1006: 
1007: \bibitem[Hoopes et al.(1999)]{hoo99} 
1008: Hoopes, C. G., Walterbos, R. A. M., \& Rand, R. J. 1999, \apj, 522, 669 
1009: 
1010: \bibitem[Howk \& Savage(1997)]{how97} 
1011: Howk, J. C., \& Savage, B. D. 1997, \aj, 114, 2463
1012: 
1013: \bibitem[Howk \& Savage(1999)]{how99} 
1014: Howk, J. C., \& Savage, B. D. 1999, \aj, 117, 2077
1015: 
1016: \bibitem[Howk \& Savage(2000)]{how00} 
1017: Howk, J. C., \& Savage, B. D. 2000, \aj, 119, 644
1018: 
1019: \bibitem[Hummel et al.(1991)]{hum91}
1020: Hummel, E., Beck, R., \& Dettmar, R.-J. 1991, A\&AS, 87, 309
1021: 
1022: \bibitem[Hunter \& Thronson(1995)]{hun95}
1023: Hunter, D. A., \& Thronson, H. A., Jr., 1995, \apj, 452, 238
1024: 
1025: \bibitem[Irwin(1994)]{irw94} 
1026: Irwin, J. A. 1994, \apj, 429, 618 
1027: 
1028: \bibitem[Kenney \& Koopmann(1999)]{ken99}
1029: Kenney, J. D. P., \& Koopmann, R. A. 1999, \aj, 117, 181
1030: 
1031: \bibitem[Kennicutt(1998)]{ken98}
1032: Kennicutt, R. C., Jr. 1998, \araa, 36, 189
1033: 
1034: \bibitem[Kewley et al.(2001)]{kew01}
1035: Kewley, L. J., Heisler, C. A., Dopita, M. A., \& Lumsden, S. 2001, \apjs, 
1036: 132, 37
1037: 
1038: \bibitem[Koekemoer et al.(2005)]{koe05}
1039: Koekemoer, A. M., Fruchter, A. S., Hook, R. N., Hack, W., \& Hanley, C. 2005, 
1040: in: The 2005 HST Calibration Workshop, eds. A. M. Koekemoer, P. Goudfrooij, 
1041: L. L. Dressel, (Baltimore: STScI)
1042: 
1043: \bibitem[Lehnert \& Heckman(1995)]{leh95} 
1044: Lehnert, M. D., \& Heckman, T. M. 1995, \apjs, 97, 89
1045: 
1046: \bibitem[Mac Low \& Ferrara(1999)]{mac99}
1047: Mac Low, M.-M., \& Ferrara, A. 1999, \apj, 513, 142
1048: 
1049: \bibitem[Malkan et al.(1998)]{mal98}
1050: Malkan, M. A., Gorjian, V., \& Tam, R. 1998, \apjs, 117, 25
1051: 
1052: \bibitem[McClure-Griffiths et al.(2003)]{mcg03}
1053: McClure-Griffiths, N. M., Dickey, J. M., Gaensler, B. M., \& Green, A. J. 
1054: 2003, \apj, 594, 833
1055: 
1056: \bibitem[McClure-Griffiths et al.(2006)]{mcg06}
1057: McClure-Griffiths, N. M., Ford, A., Pisano, D. J., Gibson, B. K., 
1058: Staveley-Smith, L., Calabretta, M. R., Dedes, L., \& Kalberla, P. M. W. 
1059: 2006, \apj, 638, 196
1060: 
1061: \bibitem[Miller \& Veilleux(2003)]{mil03} 
1062: Miller, S. T., \& Veilleux, S. 2003, \apjs, 148, 383
1063: 
1064: \bibitem[Norman \& Ikeuchi(1989)]{nor89} 
1065: Norman, C. A., \& Ikeuchi, S. 1989, \apj, 345, 372
1066: 
1067: \bibitem[Normandeau et al.(1996)]{nor96}
1068: Normandeau, M., Taylor, A. R., \& Dewdney, P. E. 1996, \nat, 380, 687
1069: 
1070: \bibitem[Oey et al.(2007)]{oey07}
1071: Oey, M. S., et al. 2007, \apj, 661, 801
1072: 
1073: \bibitem[Otte et al.(2002)]{ott02}
1074: Otte, B., Gallagher, J. S., III, \& Reynolds, R. J. 2002, \apj, 572, 823
1075: 
1076: \bibitem[Pavlovsky et al.(2005)]{pav05}
1077: Pavlovsky, C., et al. 2005, ACS Data Handbook, Version 4.0 (Baltimore: STScI)
1078: 
1079: \bibitem[Pildis et al.(1994)]{pil94} 
1080: Pildis, R. A., Bregman, J. N., \& Schombert, J. M. 1994, \apj, 427, 160 
1081: 
1082: \bibitem[Rand(1996)]{ran96} 
1083: Rand, R. J. 1996, \apj, 462, 712
1084: 
1085: \bibitem[Rand(2000)]{ran00} 
1086: Rand, R. J. 2000, \apj, 537, L13
1087: 
1088: \bibitem[Rand et al.(1990)]{ran90} 
1089: Rand, R. J., Kulkarni, S. R., \& Hester, J. J. 1990, \apj, 352, L1
1090: 
1091: \bibitem[Rand et al.(1992)]{ran92} 
1092: Rand, R. J., Kulkarni, S. R., \& Hester, J. J. 1992, \apj, 396, 97 
1093: 
1094: \bibitem[Robitaille et al.(2007)]{rob07}
1095: Robitaille, T. P., Rossa, J., Bomans, D. J., \& van der Marel, R. P. 2007, 
1096: \aap, 464, 541 
1097: 
1098: \bibitem[Rossa \& Dettmar(2000)]{ros00} 
1099: Rossa, J., \& Dettmar, R.-J. 2000, \aap, 359, 433
1100: 
1101: \bibitem[Rossa \& Dettmar(2003a)]{ros03a} 
1102: Rossa, J., \& Dettmar, R.-J. 2003a, \aap, 406, 493 
1103: 
1104: \bibitem[Rossa \& Dettmar(2003b)]{ros03b} 
1105: Rossa, J., \& Dettmar, R.-J. 2003b, \aap, 406, 505 
1106: 
1107: \bibitem[Rossa et al.(2004)]{ros04} 
1108: Rossa, J., Dettmar, R.-J., Walterbos, R. A. M., \& Norman, C. A. 2004,
1109: \aj, 128, 674 
1110: 
1111: \bibitem[Sandage \& Tammann(1990)]{san90}
1112: Sandage, A., \& Tammann, G. A. 1990, \apj, 365, 1
1113: 
1114: \bibitem[Sanders et al.(2003)]{san03}
1115: Sanders, D. B., Mazzarella, J. M., Kim, D.-C., Surace, J. A., \& Soifer, 
1116: B. T. 2003, \aj, 126, 1607
1117: 
1118: \bibitem[Schlegel et al.(1998)]{sch98}
1119: Schlegel D. J., Finkbeiner, D. P., \& Davis, M. 1998, \apj, 500, 525
1120: 
1121: \bibitem[Shapiro \& Field(1976)]{sha76} 
1122: Shapiro, P. R., \& Field, G. B. 1976, \apj, 205, 762
1123: 
1124: \bibitem[Strickland et al.(2004)]{str04}
1125: Strickland, D. K., Heckman, T. M., Colbert, E. J. M., Hoopes, C. G., \&  
1126: Weaver, K. A. 2004, \apjs, 151, 193
1127: 
1128: \bibitem[Tasker \& Bryan(2006)]{tas06}
1129: Tasker, E. J., \& Bryan, G. L. 2006, \apj, 641, 878
1130: 
1131: \bibitem[Thompson et al.(2004)]{tho04}
1132: Thompson, T. W. J., Howk, J. C., \& Savage, B. D. 2004, \aj, 128, 662
1133: 
1134: \bibitem[T\"ullmann \& Dettmar(2000)]{tul00a} 
1135: T\"ullmann, R., \& Dettmar, R.-J. 2000, \aap, 362, 119
1136: 
1137: \bibitem[T\"ullmann et al.(2000)]{tul00} 
1138: T\"ullmann, R., Dettmar, R.-J., Soida, M., Urbanik, M., \& Rossa, J. 
1139: 2000, \aap, 364, L36
1140: 
1141: \bibitem[T\"ullmann et al.(2003)]{tul03} 
1142: T\"ullmann, R., Rosa, M. R., Elwert, T., Bomans, D. J., Ferguson, A. M. N., 
1143: \& Dettmar, R.-J. 2003, \aap, 412, 69
1144: 
1145: \bibitem[T\"ullmann et al.(2006)]{tul06} 
1146: T\"ullmann, R., Pietsch, W., Rossa, J., Breitschwerdt, D., \& Dettmar, R.-J. 
1147: 2006, \aap, 448, 43
1148: 
1149: \bibitem[Veilleux et al.(2005)]{vei05}
1150: Veilleux, S., Cecil, G., \& Bland-Hawthorn, J. 2005, \araa, 43, 769
1151: 
1152: \bibitem[Walterbos \& Braun(1994)]{wal94} 
1153: Walterbos, R. A. M., \& Braun, R. 1994, \apj, 431, 156  
1154: 
1155: %\bibitem[Wang et al.(2001)]{wan01} 
1156: %Wang, Q. D., Immler, S., Walterbos, R. A. M., Lauroesch, J. T., \& 
1157: %Breitschwerdt, D. 2001, \apj, 555, L99 
1158: 
1159: \bibitem[Wills et al.(1999)]{wil99}
1160: Wills, K. A., Redman, M. P., Muxlow, T. W. B., \& Pedlar, A. 1999, \mnras, 
1161: 309, 395
1162: 
1163: \bibitem[Wyse et al.(2006)]{wys06}
1164: Wyse, R. F. G., Gilmore, G., Norris, J. E., Wilkinson, M. I., Kleyna, J. T., 
1165: Koch, A., Evans, N. W., \& Grebel, E. K. 2006, \apj, 639, L13
1166: 
1167: \bibitem[Yoachim \& Dalcanton(2006)]{yoa06}
1168: Yoachim, P., \& Dalcanton, J. J. 2006, \aj, 131, 226
1169: 
1170: \end{thebibliography}
1171: 
1172: \clearpage
1173: 
1174: %%%%%%%%%%%%%%%
1175: % Figures
1176: %%%%%%%%%%%%%%%
1177: 
1178: %%% FIGURE %%%
1179: 
1180: \clearpage
1181: 
1182: \begin{figure}
1183: \includegraphics[angle=270,scale=0.7,clip=t]{f1.eps}
1184: \caption{Diagnostic DIG diagram \citep[updated and expanded from][]{ros03a} 
1185: with the four galaxies labeled. The horizontal line marks the division of 
1186: IRAS warm galaxies ($S_{60}/S_{100} \geq 0.4$). Filled symbols denote 
1187: galaxies with eDIG detections, whereas open symbols denote galaxies without 
1188: eDIG detections. Note that the Milky Way was labeled with an open symbol for 
1189: better visibility in this diagram, even though it reveals extraplanar 
1190: emission.\label{f:dddhst}}
1191: \end{figure}
1192: 
1193: %%% FIGURE %%%
1194: 
1195: \clearpage
1196: 
1197: \begin{figure}
1198: \includegraphics[angle=0,scale=1.0,clip=t]{f2.eps}
1199: \caption{Color composite images of NGC\,4634 (top left), NGC\,4700 (top right), 
1200: NGC\,5775 (bottom left) and NGC\,7090 (bottom right), as observed with {\it HST}/ACS. 
1201: The spatial resolution is $0\farcs05$\,pix$^{-1}$. The red color represents the 
1202: continuum subtracted H$\alpha$ image, the green color represents the H$\alpha$ image, 
1203: both of which trace the emission line gas, whereas the blue color represents the continuum 
1204: image (F625W = SDSS $r$), which indicates the stellar component. The orientation of the 
1205: individual figures is the usual, with north to the top and east to the left. 
1206: \label{f:colorfigures}}
1207: \end{figure}
1208: 
1209: %%% FIGURE %%%
1210: 
1211: \clearpage
1212: 
1213: \begin{figure}
1214: \includegraphics[angle=0,scale=1.0,clip=t]{f3.eps}
1215: \caption{{\it HST}/ACS continuum subtracted H$\alpha$ image of NGC\,4634. 
1216: The orientation is with N to the top and E to the left. The labeled boxes in the 
1217: overview image (top left) refer to the individual regions A, B, C, D and E, which are 
1218: shown as enlargements in the subsequent panels to reveal a more detailed view. The boxes 
1219: overdrawn on the image to the top left refer to the image areas shown in the following 
1220: subpanels: A = top right, B = middle left, C = middle right, D = bottom left, and E = bottom 
1221: right. \label{f:n4634sixpanel}}
1222: \end{figure} 
1223: 
1224: %%% FIGURE %%%
1225: 
1226: \clearpage
1227: 
1228: \begin{figure}
1229: \includegraphics[angle=0,scale=1.0,clip=t]{f4.eps}
1230: \caption{{\it HST}/ACS continuum subtracted H$\alpha$ image of NGC\,4700. 
1231: The orientation is with N to the top and E to the left. The labeled boxes in the 
1232: overview image (top left) refer to the individual regions A, B, C, D and E, which are 
1233: shown as enlargements in the subsequent subpanels to reveal a more detailed view. The boxes 
1234: overdrawn on the image to the top left refer to the image areas shown in the following 
1235: subpanels: A = top right, B = middle left, C = middle right, D = bottom left, and E = bottom 
1236: right. \label{f:n4700sixpanel}}
1237: \end{figure} 
1238: 
1239: %%% FIGURE %%%
1240: 
1241: \clearpage
1242: 
1243: \begin{figure}
1244: \includegraphics[angle=270,scale=0.8,clip=t]{f5.eps}
1245: \caption{Color composite image of the central region in NGC\,4700, as observed 
1246: with {\it HST}/ACS. A very complex DIG morphology is revealed, with many 
1247: individual superimposed filaments protruding from the disk region far into 
1248: the halo. The spatial resolution is $0\farcs05$\,pix$^{-1}$. The red color 
1249: reflects the continuum subtracted H$\alpha$ image, the green color reflects 
1250: the H$\alpha$ image, both of which trace the emission line gas, whereas the 
1251: blue color represents the continuum image (F625W = SDSS $r$), which indicates the 
1252: stellar component. The orientation is the usual way, with north to the top and east 
1253: to the left.\label{f:n4700acs}}
1254: \end{figure}
1255: 
1256: %%% FIGURE %%%
1257: 
1258: \clearpage
1259: 
1260: \begin{figure}
1261: \includegraphics[angle=0,scale=1.0,clip=t]{f6.eps}
1262: \caption{{\it HST}/ACS continuum subtracted H$\alpha$ image of NGC\,5775. 
1263: The orientation is with N to the top and E to the left. The labeled boxes in the 
1264: overview image (top left) refer to the individual regions A, B, C, D and E, which are 
1265: shown as enlargements in the subsequent panels to reveal a more detailed view. The boxes 
1266: overdrawn on the image to the top left refer to the image areas shown in the following 
1267: subpanels: A = top right, B = middle left, C = middle right, D = bottom left, and E = bottom 
1268: right.\label{f:n5775sixpanel}}
1269: \end{figure}
1270: 
1271: %%% FIGURE %%%
1272: 
1273: \clearpage
1274: 
1275: \begin{figure}
1276: \includegraphics[angle=0,scale=1.0,clip=t]{f7.eps}
1277: \caption{{\it HST}/ACS continuum subtracted H$\alpha$ image of NGC\,7090. 
1278: The orientation is with N to the top and E to the left. The labeled boxes in the 
1279: overview image (top left) refer to the individual regions A, B, C, D and E, which are 
1280: shown as enlargements in the subsequent panels to reveal a more detailed view. The boxes 
1281: overdrawn on the image to the top left refer to the image areas shown in the following 
1282: subpanels: E = top right, D = middle left, C = middle right, B = bottom left, and A = bottom 
1283: right. \label{f:n7090sixpanel}}
1284: \end{figure}
1285: 
1286: %%% FIGURE %%%
1287: 
1288: \clearpage
1289: 
1290: \begin{figure}
1291: \includegraphics[angle=0,scale=1.0,clip=t]{f8.eps}
1292: \caption{Unsharp-masked images revealing the extraplanar dust distribution 
1293: in NGC\,4634 (top left), NGC\,4700 (top right), NGC\,5775 (bottom left) and 
1294: NGC\,7090 (bottom right). The global orientation of the dust filaments in 
1295: NGC\,4634 is mostly perpendicular to the disk on both sides of the disk at 
1296: high $|z|$. At lower $|z|$ the orientation of the dust filaments is random. 
1297: Due to the lack of a strong dust lane in NGC\,4700 (and the general lack of 
1298: patchy dust in this galaxy), the unsharp-masked image does not reveal 
1299: much insight into the dust distibution. Also, the presence of many stars 
1300: that are clearly resolved within NGC\,4700, complicates the interpretation 
1301: of the distribution of extended dust filaments. Only the central part of 
1302: NGC\,5775 is shown here for better visibility. There is a filamentary web of 
1303: dusty structures revealed. The global orientation of the dust filaments are mostly 
1304: parallel to the disk on the one side of the disk, whereas the other side has mostly 
1305: an orientation perpendicular to the disk. Generally, though, at lower $|z|$ the orientation 
1306: is random. For NGC\,7090 also only the central part is shown here. Due to the lack 
1307: of a pronounced dust lane, the unsharp-masked image does not reveal much unsight 
1308: into the extraplanar dust distribution. \label{f:dust4}}.
1309: \end{figure}
1310: 
1311: %%% FIGURE %%%
1312: 
1313: \clearpage
1314: 
1315: \begin{figure}
1316: \includegraphics[angle=0,scale=0.95,clip=t]{f9.eps}
1317: \caption{The tidally disrupted dIrr galaxy, $\rm{J}\,124239.58$+141751.83, 
1318: in the halo of NGC\,4634, as revealed by {\it HST}/ACS. The left panel shows 
1319: the broadband image (F625W), the right panel shows the continuum subtracted 
1320: H$\alpha$ image. The field of view measures $17\farcs8 \times 12\farcs3$, which 
1321: corresponds to 1650\,pc$\times$1140\,pc at the distance to NGC\,4634. The orientation 
1322: is the usual with north to the top and east to the left. The diagonal color gradient 
1323: in the broadband image is not a flatfield artifact. It reflects the increase of galaxy 
1324: light from the halo region in the left side of the figure in this color stretch. 
1325: \label{f:dIrr}}
1326: \end{figure}
1327: 
1328: %%% END OF FIGURES %%%
1329: 
1330: \clearpage
1331: 
1332: %%%%%%%%%%%%%%%
1333: % Tables 
1334: %%%%%%%%%%%%%%%
1335: 
1336: %%% TABLE %%%
1337: 
1338: \begin{deluxetable}{lcccccc}
1339: \tabletypesize{\scriptsize}
1340: \tablecaption{Basic Galaxy Sample Properties\label{t:obs}}
1341: \tablewidth{0pt}
1342: \tablehead{
1343: Galaxy & R.A. (J2000.0) & Decl. (J2000.0) & Classification & $cz$ & $D$ & 
1344: $A_V$ \\
1345:  & & & & (km~s$^{-1}$) & (Mpc) & (mag) \\
1346: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} &
1347: \colhead{(5)} & \colhead{(6)} & \colhead{(7)} \\ 
1348: }
1349: \startdata
1350: \object{NGC\,4634} & 12 42 41.00 & $+$14 17 46.0 & SBcd & 221 & 19.1 & 0.093 \\
1351: \object{NGC\,4700} & 12 49 07.32 & $-$11 24 45.6 & SB(s)c & 1386 & 24.0 & 
1352: 0.156 \\
1353: \object{NGC\,5775} & 14 53 57.60 & $+$03 32 40.1 & SBc & 1768 & 26.7 & 0.139 \\
1354: \object{NGC\,7090} & 21 36 28.61 & $-$54 33 26.2 & SBc & 642 & 11.4 & 0.076 \\
1355: \enddata
1356: \tablecomments{Col.~(1): Galaxy name. Col.~(2) and~(3): Right ascension and 
1357: declination from the NASA Extragalactic Database (NED). Units of right 
1358: ascension are hours, minutes and seconds, and units of declination are 
1359: degrees, arcminutes and arcseconds. Col.~(4): Morphological classification 
1360: given by NED. Col.~(5): Systemic velocity corrected for Virgocentric infall 
1361: using the model of \citet{san90}, as taken from LEDA. Col.~(6): Distance, 
1362: taken as listed in \citet{ros00,ros03b}. Col.~(7): Galactic foreground 
1363: extinction \citep{sch98} in the $V$-band for $R_V = 3.1$ from NED.}  
1364: \end{deluxetable}
1365: 
1366: \clearpage
1367: 
1368: \begin{deluxetable}{lccccccccc}
1369: \tabletypesize{\scriptsize}
1370: \tablecaption{Journal of ACS Observations\label{t:exp}}
1371: \tablewidth{0pt}
1372: \tablehead{
1373: Galaxy & Data set ident. & Obs. Date & ACS-channel & Filter & 
1374: $\lambda_{\rm{cen}}$ & FWHM & $t_{\rm{int}}$ & Dither & P.A. \\
1375:  & & & & & ({\AA}) & ({\AA}) & (sec) & & ($\degr$) \\
1376: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} &
1377: \colhead{(5)} & \colhead{(6)} & \colhead{(7)} & \colhead{(8)} & 
1378: \colhead{(9)} & \colhead{(10)} \\
1379: }
1380: \startdata
1381: NGC\,4634 & j95x01 & 2005 Jun 24 & WFC & F625W & 6318 & 1442 & 2308.0 & 
1382: $2\times2$ & 125.596 \\
1383: NGC\,4634 & j95x01 & 2005 Jun 24 & WFC & F658N & 6584 & 78 & 6888.0 & 
1384: $3\times4$ & 125.596 \\
1385: NGC\,4700 & j95x02 & 2005 Jul 30 & WFC & F625W & 6318 & 1442 & 2308.0 & 
1386: $2\times2$ & 115.209 \\
1387: NGC\,4700 & j95x02 & 2005 Jul 30 & WFC & F658N & 6584 & 78 & 6888.0 & 
1388: $3\times4$ & 115.209 \\
1389: NGC\,5775 & j95x03 & 2005 Aug 21 & WFC & F625W & 6318 & 1442 & 2292.0 & 
1390: $2\times2$ & 101.544 \\
1391: NGC\,5775 & j95x03 & 2005 Aug 21 & WFC & F658N & 6584 & 78 & 6848.0 & 
1392: $3\times4$ & 101.544 \\
1393: NGC\,7090 & j95x04 & 2005 Jun 23 & WFC & F625W & 6318 & 1442 & 2508.0 & 
1394: $2\times2$ & $-$72.289 \\
1395: NGC\,7090 & j95x04 & 2005 Jun 23 & WFC & F658N & 6584 & 78 & 7496.0 & 
1396: $3\times4$ & $-$72.289 \\
1397: \enddata
1398: \tablecomments{Col.~(1): Galaxy name. Col.~(2) Data set identifier in the 
1399: {\it HST} Data Archive. Col.~(3): Date of observations. Col.~(4): ACS-aperture 
1400: (WFC: Wide Field Channel). Col.~(5), (6) and (7): Filter properties, namely 
1401: the identifier, central wavelength and full width at half maximum, 
1402: respectively. Col.~(8) Total exposure time, which is the sum of the individual 
1403: dithered images. Col.~(9) Used dither pattern (primary- and sub-pattern). 
1404: Col.~(10): Orientation angle at which the observations were performed.}  
1405: \end{deluxetable}
1406: 
1407: \begin{deluxetable}{lcccc}
1408: \tabletypesize{\scriptsize}
1409: \tablecaption{Selected Specific Galaxy Properties\label{t:prop}}
1410: \tablewidth{0pt}
1411: \tablehead{
1412: Property & NGC\,4634 & NGC\,4700 & NGC\,5775 & NGC\,7090 \\
1413: }
1414: \startdata
1415: $M_{\rm tot} [\rm{M}_{\sun}]$ & $2.72\times10^{10}$ & $1.91\times10^{10}$ & 
1416: $1.91\times10^{11}$ & $4.52\times10^{10}$ \\
1417: Disk morphology & regular? & irregular & regular & asymmetric \\
1418: eDIG morphology & diffuse & filamentary & diffuse & filamentary \\
1419: %e-dust morph.??? & & & & \\
1420: Interaction & yes & no & yes & no \\
1421: SF distribution & extended & extended & extended & part disk \\
1422: $SFR_{\rm tot} [\rm{M}_{\sun}\,\rm{yr}^{-1}]$ & 1.08 & 1.00 & 9.86 & 0.58 \\
1423: $S_{60}/S_{100}$ & 0.3720 & 0.5074 & 0.4240 &0.4129 \\
1424: $L_{\rm{FIR}}/D^2_{25} [10^{40}\,\rm{erg\,s^{-1}\,kpc^{-2}}]$  & 11.9 & 5.0 
1425: & 29.6 & 2.4\\
1426: X-ray halo & yes & ... & yes & yes? \\
1427: Radio halo & extended disk & yes & yes & yes \\
1428: $B_{\perp}$ & ... & ... & yes & yes \\
1429: \enddata
1430: \tablecomments{The total galaxy mass $M_{\rm tot}$ in units of solar masses, 
1431: listed in Row~(1) was calculated from \hbox{H\,{\sc i}} measurements using the 
1432: relation $M_i = 3\times10^{4}*A(0)*(\Delta v_0^i)^2$. A(0) is calculated 
1433: either from $D_{25}$ or D[\hbox{H\,{\sc i}}] and has units of kpc, and $\Delta 
1434: v_0^i$ is the inclination corrected full width W[\hbox{H\,{\sc i}}] in units 
1435: of km/s. Row~(2) describes the disk morphology and Row~(3) is the eDIG 
1436: morphology. Row~(4) describes whether the galaxy is interacting or not. 
1437: Row~(5) lists the SF distribution. Row~(6) lists the star formation rate. The 
1438: SFR's are calculated based on FIR data and using the relatation given by 
1439: \citet{ken98}. Row~(7) is the IRAS flux ratio at $60\mu$m and $100\mu$m. These 
1440: are based on the newly calibrated data from \citet{san03}. Row~(8) lists the 
1441: SFR per unit area, expressed as FIR luminosity ($L_{\rm{FIR}}$) divided by the 
1442: diameter of the optical 25th magnitude isophote diameter squared. Row~(9) and 
1443: (10) list whether the galaxy has detections of extended emission in the X-ray 
1444: and radio regime, respectively. The preliminary X-ray analysis of shallow 
1445: XMM-Newton data of NGC\,7090 shows the indication of an X-ray halo. This has 
1446: to be confirmed by deeper observations, which are already scheduled. For 
1447: NGC\,4700 X-ray observations are still pending. No radio halo was detected in 
1448: NGC\,4634, but based on low-resolution data obtained with the 100m Effelsberg 
1449: radio telescope at 2.8\,cm, at least an extended disk is revealed (J. Rossa, 
1450: unpublished). Finally, Row~(11) lists measured magnetic field vectors 
1451: perpendicular to the disk.}  
1452: \end{deluxetable}
1453: 
1454: \begin{deluxetable}{lccc}
1455: \tabletypesize{\scriptsize}
1456: \tablecaption{Description of the Observed DIG Morphology\label{t:morph}}
1457: \tablewidth{0pt}
1458: \tablehead{
1459: Galaxy & Region & Morphology & Ref. Figure \\
1460: \colhead{(1)} & \colhead{(2)} & \colhead{(3)} & \colhead{(4)} \\
1461: }
1462: \startdata
1463: NGC\,4634 & A & mostly diffuse & Figure~\ref{f:n4634sixpanel}b\\
1464: & & many compact SF regions &  \\
1465: & & few shells SF regions &  \\
1466:  & B & mostly diffuse & Figure~\ref{f:n4634sixpanel}c\\
1467: & & dIrr galaxy to the SW &  \\
1468:  & C & mostly diffuse & Figure~\ref{f:n4634sixpanel}d\\
1469: & & bright shells/supershells interspersed with dust clouds &  \\
1470:  & D & mostly diffuse & Figure~\ref{f:n4634sixpanel}e\\
1471: & & bright SF region (top) &  \\
1472: & & small-scale filaments surrounding bright shells &  \\
1473: & & prominent dust cloud &  \\
1474:  & E & two isolated shells/SF regions & Figure~\ref{f:n4634sixpanel}f\\
1475: & & off-planar feature (to the SW) & \\
1476: \hline \\
1477: NGC\,4700 & A & many superbubbles & Figure~\ref{f:n4700sixpanel}b\\
1478: & & filaments protruding from bright shells &  \\
1479:  & B & superbubbles & Figure~\ref{f:n4700sixpanel}c\\
1480: & & very filamentary &  \\
1481: & & bright supershell (to the SW) &  \\
1482:  & C & spectacular central ``radio lobe'' structure & Figure~\ref{f:n4700sixpanel}d\\
1483: & & filaments protruding from central region &  \\
1484: & & compact SF regions scattered (mostly off-planar) &  \\
1485:  & D & southern extension of central structure & Figure~\ref{f:n4700sixpanel}e\\
1486: & & filamentary + bright shells &  \\
1487:  & E & superbubbles & Figure~\ref{f:n4700sixpanel}f\\
1488: & & few filaments &  \\
1489: \hline \\
1490: NGC\,5775 & A & mostly diffuse & Figure~\ref{f:n5775sixpanel}b\\
1491: & & several bright SF regions of various sizes &  \\
1492:  & B & mostly diffuse & Figure~\ref{f:n5775sixpanel}c\\
1493: & & few bright SF regions of various sizes &  \\
1494:  & C & bright SF regions & Figure~\ref{f:n5775sixpanel}d\\
1495: & & patchy dust regions interspersed &  \\
1496:  & D & faint diffuse emission & Figure~\ref{f:n5775sixpanel}e\\
1497: & & few bright SF regions of various sizes &  \\
1498:  & E & very bright and complex SF region & Figure~\ref{f:n5775sixpanel}f\\
1499: & & diffuse surrounding breaking up into filaments &  \\
1500: \hline \\
1501: NGC\,7090 & A & few bright filamentary SF regions & Figure~\ref{f:n7090sixpanel}f\\
1502: & & many smaller SF regions interspersed &  \\
1503:  & B & complex filamentary SF region around bright star & Figure~\ref{f:n7090sixpanel}e\\
1504: & & few smaller SF regions interspersed &  \\
1505:  & C & very bright SF regions & Figure~\ref{f:n7090sixpanel}d\\
1506: & & copious filaments emanating from SF regions &  \\
1507: & & few compact off-planar SF regions  &  \\
1508:  & D & mainly bubbles and superbubbles & Figure~\ref{f:n7090sixpanel}c\\
1509: & & faint diffuse DIG superimposed &  \\
1510:  & E & few bright SF regions & Figure~\ref{f:n7090sixpanel}b\\
1511: & & bubbles, superbubbles, complex filaments &  \\
1512: \enddata
1513: \tablecomments{Col.~(1) lists the galaxy identifier. Col.~(2) lists the specific 
1514: region under study within each galaxy (labeled by identifiers A,...,E), and with 
1515: the described morphology in Col.~(3) as revealed in the corresponding figures, 
1516: indicated by Col.~(4). The reference figure labeling refers to the following subpanel 
1517: positions within the individual figures: b = top right, c = middle left, d = middle right, 
1518: e = bottom left, f = bottom right.}  
1519: \end{deluxetable}
1520: 
1521: 
1522: %%% END OF TABLES %%%
1523: 
1524: %%% END OF DOCUMENT %%%
1525: 
1526: \end{document}
1527: 
1528: 
1529: 
1530: