1: %\documentclass[preprint2]{aastex}
2: %% manuscript produces a one-column, double-spaced document
3:
4: \documentclass[preprint2]{emulateapj}
5: %\documentclass[12pt,preprint]{aastex}
6: %% preprint2 produces a double-column, single-spaced document
7:
8: \newcommand{\HI}{\ion{H}{1}}
9: \newcommand{\OII}{[\ion{O}{2}]}
10: \newcommand{\kms}{km\,s$^{-1}$}
11: \newcommand{\Msun}{M$_\sun$}
12:
13: \shorttitle{H{\sc i} observations of EA01A/B} \shortauthors{Buyle, De
14: Rijcke, Dejonghe}
15:
16: \begin{document}
17:
18: \title{EA01A/B : high-resolution H{\sc i} imaging of an interacting
19: pair of post-starburst (E+A) galaxies} \author{Pieter Buyle\altaffilmark{1},
20: Sven De Rijcke\altaffilmark{1}, Herwig Dejonghe\altaffilmark{1}}
21:
22: \altaffiltext{1}{Sterrenkundig Observatorium, Universiteit Gent,
23: Krijgslaan 281, S9, B-9000, Ghent, Belgium; Pieter.Buyle@UGent.be,
24: Sven.DeRijcke@UGent.be}
25:
26: \begin{abstract}
27: We present high spatial resolution 21~cm HI observations of EA01A and
28: EA01B, a pair of interacting post-starburst, or E+A, galaxies at $z =
29: 0.0746$. Based on optical HST/WFPC2 images, both galaxies are known to
30: display disturbed morphologies. They also appear to be linked by a
31: bridge of stars. Previous HI observations \citep{chang01} had already
32: uncovered sizable quantities of neutral gas in or near these galaxies
33: but they lacked the spatial resolution to locate the gas with any
34: precision within this galactic binary system. We have analysed deep,
35: high resolution archival VLA observations of the couple. We find
36: evidence for three gaseous tidal tails; one connected to EA01A and two
37: emanating from EA01B. These findings confirm, independently from the
38: optical imaging, that {\em (i)} EA01A and EA01B are actively
39: interacting, and that, as a consequence, the starbursts that occurred
40: in these galaxies were most likely triggered by this interaction, and
41: that {\em (ii)} $6.6\pm 0.9\times10^9$~{\Msun} of neutral gas are
42: still present in the immediate vicinity of the optical bodies of both
43: galaxies. The \HI\ column density is lowest at the optical positions
44: of the galaxies, suggesting that most of the neutral gas that is
45: visible in our maps is associated with the tidal arms and not with the
46: galaxies themselves. This might provide an explanation for the
47: apparent lack of ongoing star formation in these galaxies.
48: %The recession velocity of the gas is very close to that of
49: %the galaxies, showing that it is most likely still gravitationally
50: %bound and will remain available for future star formation events.
51: \end{abstract}
52:
53: \keywords{galaxies: evolution, galaxies: elliptical, galaxies: ISM, galaxies: fundamental parameters}
54:
55: \section{Introduction}\label{intro}
56:
57: Post-starburst galaxies (or PSGs, or E+A galaxies, or k+a/a+k
58: galaxies) are characterized by their optical spectra, independent of
59: any morphological or photometric considerations. PSGs contain a
60: sizable population of very young stars, producing very prominent
61: Balmer lines, but lack ongoing star formation, hence the absence of
62: detectable emission lines. This suggests that PSGs are indeed observed
63: very shortly, within less than 1~Gyr, after the end of a vigorous,
64: abruptly truncated starburst \citep{dres83,dres99,pog99}. Today, they
65: constitute only a small fraction of the galaxy cluster population
66: ($<1$\%, \citet{fab91}); at intermediate-redshifts, however, they
67: constituted a substantial cluster population \citep{bel95}. Although
68: the trigger and the abrupt end of the starburst is still not fully
69: understood, photometry of PSGs shows evidence for disturbed
70: morphologies, e.g. tidal tails, suggesting that in many cases the
71: trigger is most likely associated to a galaxy-galaxy merger or
72: interaction \citep{zab96,yang04,blake04,tran03,goto05,yang08}. Using
73: numerical simulations, \citet{bekki05} has shown that PSGs can be
74: formed via a major merger of two gas-rich spiral galaxies. In these
75: merger simulations, a starburst is triggered that consumes the
76: available gas within a timespan of roughly 1~Gyr, ending the starburst
77: abruptly. Recent \HI\ observations have uncovered large quantities of
78: neutral hydrogen gas in a significant fraction of PSGs
79: \citep{chang01,buyle06,helmboldt07,buyle08}. This opens up the
80: possibility of investigating the hypothesized merger origin of many
81: PSGs.
82:
83:
84: %This presence of \HI\can
85: %be explained in two ways. {\em (1)} An intensive burst of star
86: %formation can evaporate the dense molecular clouds \citep{f03}, the
87: %sites where star formation takes place, stopping further star
88: %formation even though a significant amount of \HI\ remains, possibly
89: %in a too rarefied state to lead to new star formation on short
90: %timescales. {\em (2)} Alternatively, star formation may still be
91: %ongoing but is obscured by large amounts of dust
92: %\citep{couch87,blake04}. The latter option, however, seems to
93: %contradict the radio-continuum observations of a sample of PSGs by
94: %\cite{mo01}, which showed that they have little or no radio emission
95: %coming from supernova remnants and hence no ongoing star formation.
96:
97: \begin{figure*}
98: \includegraphics[width=17cm]{f1.eps}
99: \caption{Channel maps overplotted on an optical HST/WFPC2 F702W image
100: of EA01A/B. The contour levels are 2$\sigma$, $3\sigma$,\ldots with
101: $\sigma$=0.2~mJy~beam$^{-1}$. The synthesized beam is depicted in the
102: bottom-right corner of the first map (top left).
103: \label{channels}}
104: \end{figure*}
105:
106: Indeed, high-resolution interferometric radio observations of the
107: neutral gas would provide valuable additional information. On the one
108: hand, \HI\ morphologies and velocity fields, if they are distorted,
109: could corroborate the merger hypothesis, and, moreover, could reveal
110: whether at least part of the neutral gas is still gravitationally
111: bound and will remain available for future star formation. We selected
112: the PSG binary EA01A/B as an interesting target for our study. Both
113: galaxies have a PSG-type optical spectrum and have the same recession
114: velocity. EA01A is the bluest galaxy of the PSG sample of
115: \cite{zab96}, suggesting it is also the youngest sample
116: member. HST/WFPC2 images in the F435W and F702W bands \citep{yang04}
117: provide strong evidence that the couple is interacting. EA01A,
118: positioned about 11$''$ east of EA01B, contains many stellar clusters
119: that are very conspicuous compared with the overall rather low surface
120: brightness of this galaxy. This diffuse appearance suggests it is on
121: the verge of disintegrating due to the injected orbital energy. EA01B,
122: on the other hand, is a bulge dominated early-type spiral galaxy; it
123: sports strongly asymmetric stellar arms. Both galaxies appear to be
124: connected by a stellar bridge. Previous HI observations
125: \citep{chang01} had already uncovered sizable quantities of neutral
126: gas in or near these galaxies but they lacked the spatial resolution
127: to locate the gas with any precision within this galactic binary
128: system.
129:
130: In section~\ref{obs}, we describe our data reduction and analysis of
131: deep archival VLA \HI\ observations of the EA01A/B galaxy binary. The
132: results are presented and discussed in section~\ref{discussion} and
133: summarized in section~\ref{conclusions}.
134:
135: \section{Observations and data reduction}\label{obs}
136:
137: \begin{figure*}
138: \includegraphics[width=9cm]{f2a.eps}
139: \includegraphics[width=9cm]{f2b.eps}
140: \includegraphics[width=2.5cm]{f2c.eps}
141: \includegraphics[width=2.5cm]{f2d.eps}
142: \includegraphics[width=2.5cm]{f2e.eps}
143: \includegraphics[width=2.5cm]{f2f.eps}
144: \includegraphics[width=2.5cm]{f2g.eps}
145: \includegraphics[width=2.5cm]{f2h.eps}
146: \includegraphics[width=2.5cm]{f2i.eps}
147: \caption{The observed \HI\ flux integrated map (left panel) from Chang
148: et al. (2001) (white contours) and from this paper (black contours)
149: and the \HI\ flux integrated map of a model that roughly reproduces
150: the most conspicuous features of the channel maps (right panel). The
151: synthesized beam is shown at the bottom left. The contour levels of
152: the observed flux map correspond to $2\sigma$, $3\sigma$,\ldots with
153: $\sigma=0.01$~Jy~beam$^{-1}$~\kms. The model channel maps are
154: displayed in the bottom panels.
155: \label{mom0}}
156: \end{figure*}
157:
158: \begin{table}
159: \caption{Data characteristics \label{char}}
160: \begin{tabular}{lc}
161: \tableline
162: channel width (km~s$^{-1}$) & 11.1 \\
163: synthesized beam (FWHM) & 8.87$'' \times$6.32$''$ \\
164: 1$\sigma$ noise level per channel & \\
165: \hspace{3em} flux density (mJy beam$^{-1}$) & 0.2 \\
166: \hspace{3em} brightness temperature (K beam$^{-1}$) & 2.1 \\
167: \hspace{3em} column density ($\times 10^{19}$ cm$^{-2}$ beam$^{-1}$) & 4.3 \\
168: integrated 1$\sigma$ noise level & \\
169: \hspace{3em} flux (Jy beam$^{-1}$ km s$^{-1}$) & 0.01 \\
170: \hspace{3em} brightness temperature (K beam$^{-1}$ km s$^{-1}$) & 71.1 \\
171: \hspace{3em} column density ($\times 10^{20}$ cm$^{-2}$ beam$^{-1}$) & 1.9 \\
172: \tableline
173: \tableline
174: \end{tabular}
175: \end{table}
176: We retrieved deep, high-resolution \HI\ 21~cm observations of EA01A/B
177: from the science archive of the National Radio Astronomy Observatory
178: (NRAO), made with the Very Large Array (VLA) in New Mexico (USA)
179: (project number AM0678). The observations were made with the B
180: configuration during the nights of March 11, 12, 14, 16 and 18,
181: 2001. At a central frequency of $1321.76$~MHz (L-band) and with a
182: bandwidth of $3.125$~MHz divided over 63 channels, the \HI\ in EA01A/B
183: was mapped with a velocity resolution of $11.1$~\kms. Alternating
184: observations of the phase calibrator 1127-145 with a 4 minute exposure
185: time and EA01A/B with an exposure time of 25 minutes were made,
186: followed at the end of each night by a 2.5 minute exposure of the flux
187: calibrator 3C286. This resulted in a total on-source integration time
188: of $32.6~$h. Standard flagging and calibration of the $u-v$ data was
189: performed with the Astronomical Image Processing Software (AIPS). The
190: continuum was subtracted by making a linear fit to the visibilities
191: over the line-free channels that were not affected by the edge effects
192: of the band. We created the final datacube using natural weighting to
193: optimize the sensitivity. This yields a beam-size of $8.87''\times
194: 6.32''$ or $12.3~h^{-1}$~kpc$\times 8.7~h^{-1}$~kpc at the
195: angular-size distance of EA1 (D$_A=285~h^{-1}$~Mpc, with
196: $H_0=75~h$~\kms~Mpc$^{-1}$). The rms noise per channel is
197: $0.2$~mJy~beam$^{-1}$, corresponding to a \HI\ column density of $4.3
198: \times 10^{19}$ cm$^{-2}$ averaged over the beamwidth. A summary of the
199: data characteristics can be found in Table \ref{char}.
200:
201: \section{Discussion}\label{discussion}
202:
203: We examined the final datacube and detected 21~cm emission in 12
204: adjacent channels, shown in Fig.~\ref{channels}. The \HI\ flux
205: integrated map of EA01A/B is presented in the left panel of
206: Fig. \ref{mom0}. All maps are overplotted onto a HST/WFPC2 F702W image
207: of the EA01A/B pair. It is clear from these maps that the
208: \HI\ emission spatially coincides with the optical position of EA01A/B
209: and is centered around the heliocentric velocity of the previous
210: \HI\ detection and the optical recession velocity of the pair
211: ($22327.9 -22450.0\, {\rm km}\,{\rm s}^{-1}$).
212:
213: The \HI\ profile of EA01A/B, obtained by summing the flux within a box
214: covering the radio emission of the galaxy is plotted in
215: Fig. \ref{specEA1}. We fitted a Gaussian to the \HI\ spectrum in order
216: to derive the total 21~cm line flux. We generated a large library of
217: mock \HI\ spectra using this Gaussian model and added noise to produce
218: synthetic data of the same quality as the original spectrum. We then
219: fitted Gaussian profiles to the mock data in order to estimate the
220: 1$\sigma$ uncertainties on all relevant parameters. We find a total
221: integrated flux of $S_\nu = 0.26\pm0.03$~Jy~\kms. This is equivalent to an
222: \HI\ mass of M$_{\rm HI}=6.6\pm 0.9\times10^9$~{\Msun}, adopting a
223: luminosity distance D$_L$=330~Mpc. The central velocity of the
224: Gaussian is $22357 \pm 9$~\kms\ and its FWHM is $133 \pm 20$~\kms. We
225: also re-reduced and analysed the low-resolution \HI\ observations of
226: \cite{chang01} (Fig. \ref{mom0}). Based on these data, we find a total integrated flux
227: of $S_\nu=0.29\pm0.08$~Jy~\kms, corresponding to an \HI\ mass of $7.5\pm
228: 2.1\times10^9$~\Msun. The central velocity is $22371 \pm 30$~\kms\ and
229: the FWHM is $191\pm 88$~\kms. This is in agreement with the
230: high-resolution VLA data presented here and with \cite{chang01}.
231:
232: \begin{figure}
233: \includegraphics[width=8cm]{f3.eps}
234: %\includegraphics[width=8cm]{specchangEA1.eps}
235: \caption{The \HI\ profile of EA01A/B, obtained by summing the flux
236: within a box around the flux integrated radio emission of the
237: galaxy. Rms noise levels are indicated by horizontal lines. A
238: Gaussian was fitted to the spectra to measure the velocity widths.
239: \label{specEA1}}
240: \end{figure}
241:
242: A plausible interpretation of the \HI\ flux integrated map and the
243: channel maps is that we are seeing one receding gaseous tidal arm
244: emanating from EA01A and two tidal arms connected to EA01B, with the
245: east arm approaching us and the west arm moving away from us. Little
246: of the detected gas appears to be directly connected with the galaxies
247: themselves, which show up as minima in the \HI\ column density. This
248: absence of dense gas may be related to the fact that these galaxies
249: are in a post-starburst phase. The EA01A arm runs in a south-eastern
250: direction and can be traced through 3 consecutive channel maps from
251: 22372.3~km~s$^{-1}$ to 22394.5~km~s$^{-1}$. The west EA01B ``arm''
252: shows up as emission in 3 consecutive channel maps from
253: 22372.3~km~s$^{-1}$ to 22394.5~km~s$^{-1}$. Following its position
254: through adjacent channel maps with recession velocities below $\approx
255: 22370$~km~s$^{-1}$, the east arm connected with EA01B appears to first
256: shift in the direction of EA01A and then to return a position slightly
257: north of EA01B before disappearing in the noise. At that point, the
258: gas is moving towards us with a radial velocity of about
259: 40~km~s$^{-1}$, relative to the systemic velocity of EA01A/B. The
260: curvature of the northern arm and the fact that it is approaching us
261: agrees with the curvature of the stellar arms.
262:
263: Since the data quality clearly does not allow for a detailed or
264: quantitative modeling of the gas distribution and kinematics, we
265: constructed a very simplified model for the \HI\ observations based on
266: the interpretation outlined above. Two gaseous arms were constructed
267: to emanate from EA01B, following the position and curvature of the
268: stellar arms. If the ellipticity $\epsilon = b/a \approx 0.5$ of the
269: outer isophotes of EA01B is indicative of its inclination, then this
270: early-type disk is inclined by about 30$^\circ$ from an edge-on
271: view. We assumed both arms to lie in a disk with the same
272: inclination. A third gas arm was connected with EA01A, curving in a
273: south-eastern direction. The \HI\ flux density of these model tidal
274: arms was assumed to decline exponentially with radius and they were
275: given solid-body rotation velocity fields in such a way that the mock
276: channel maps and the flux integrated map derived from this model
277: provided an acceptable match to the observed maps, taking into account
278: beam convolution and added noise. Despite its simplicity, this model
279: can give a fair account of the observations. The model flux map and
280: channel maps are presented in Fig. \ref{mom0} and can be compared with
281: Fig. \ref{channels}.
282:
283: With a 2$\sigma$ density limit of $4 \times 10^{20}$ cm$^{-2}$
284: beam$^{-1}$, averaged over a $12.3~h^{-1}$~kpc$\times 8.7~h^{-1}$~kpc
285: beam, we cannot expect to trace the tidal arms very far out. As a
286: comparison, under these conditions it would be impossible to trace the
287: southern arm emanating from the interacting pair NGC4038/4039 (``The
288: Antennae'') further out than $\sim 20$~kpc \citep{hi01}. At the
289: distance of EA01A/B, this corresponds to $\sim 15''$, less than two
290: beam diameters. At that distance, the velocity of the NGC4038/4039
291: tidal arm is less than $\sim 50$~km~s$^{-1}$, in rough agreement with
292: the largest velocities measured in the EA01A/B system.
293:
294: \section{Conclusions and summary}\label{conclusions}
295:
296: EA01A/B is a close pair of post-starburst (E+A) galaxies, surrounded
297: by some $7 \times 10^9$~{\Msun} of neutral gas. Most of this gas
298: resides in what appear to be three tidal arms, two of which are
299: connected with EA01B. Together with optical HST/WFPC2 images, these
300: observations show that EA01A and EA01B are actively interacting. The
301: galaxies themselves show up as minima in the \HI\ column density. This
302: lack of galaxy-bound dense neutral gas is most likely connected with the
303: fact that these galaxies are in a post-starburst phase.
304:
305: \acknowledgements We like to thank the referee, Tomotsugu Goto, for
306: his very useful comments. PB and SDR are post-doctoral fellows of the
307: Fund for Scientific Research - Flanders, Belgium (F.W.O.). The Very
308: Large Array is a facility instrument of the National Radio Astronomy
309: Observatory, which is operated by Associated Universities, Inc. under
310: contract from the National Science Foundation. This work has made use
311: of the NASA/IPAC Extragalactic Database (NED) which is operated by the
312: Jet Propulsion Laboratory, California Institute of Technology, under
313: contract with the National Aeronautics and Space Administration.
314:
315: \begin{thebibliography}{}
316:
317: \bibitem[Bekki et al.(2005)]{bekki05} Bekki, K., Couch, W.~J., Shioya, Y., \&
318: Vazdekis, A.\ 2005, \mnras, 359, 949
319:
320: \bibitem[Belloni et al. (1995)]{bel95} Belloni, P., Bruzual, A.G.,
321: Thimm, G.J. \& Roser, H.-J., 1995, A\&A, 297, 61
322:
323: \bibitem[Blake et al.(2004)]{blake04} Blake, C., Pracy, M. B., Couch,
324: W. J., Kenji, B., Lewis, I., Glazebrook, K., Baldry, I. K., Baugh,
325: b. M., Bland-Hawthorn, J., et al., 2004, \mnras, 355, 713
326:
327: \bibitem[Buyle et al.(2006)]{buyle06} Buyle, P., Michielsen, D., De Rijcke,
328: S., Pisano, D.J., Dejonghe, H. \& Freeman, K. 2006, \apj, 649, 163
329:
330: \bibitem[Buyle et al.(2008)]{buyle08} Buyle, P., De Rijcke, S.,
331: Pisano, D.J., Michielsen, D., Dejonghe, H. \& Freeman, K. 2008, \apj,
332: in preparation
333:
334: \bibitem[Chang et al. (2001)]{chang01} Chang, T., van Gorkom, J.H.,
335: Zabludoff, A., Zaritsky, D. \& Mihos, C., 2001, \aj, 121, 1965
336:
337: \bibitem[Couch \& Sharples (1987)]{couch87} Couch, W. J., Sharples,
338: R. M., 1987, MNRAS, 229, 423
339:
340: \bibitem[Dressler \& Gunn(1983)]{dres83} Dressler, A., Gunn, J. E.,
341: 1983, ApJ, 270, 7
342:
343: \bibitem[Dressler et al.(1999)]{dres99} Dressler, A., Smail, I.,
344: Poggianti, B. M., Butcher, H., Couch, W. J., Ellis, R. S., \& Oemler,
345: A., Jr., 1999, ApJS, 122, 51
346:
347: \bibitem[Fabricant et al. (1991)]{fab91} Fabricant, D.G., McClintock, J.E.\& Bautz, M.W., 1991, \apj, 381, 33
348:
349: \bibitem[Franco (2003)]{f03} Franco, J., 2003, RMxAC, 15, 149
350: \bibitem[Goto(2005)]{goto05} Goto, T.\ 2005, MNRAS, 357, 937
351:
352: \bibitem[Helmboldt (2007)]{helmboldt07} Helmboldt., J.F. 2007, \mnras,
353: 379, 1227
354:
355: \bibitem[Hibbard et al. (2001)]{hi01} Hibbard, J. E., van der Hulst,
356: J. M, Barnes, J. E., Rich, R. M., 2001, AJ, 122, 2969
357:
358: \bibitem[Miller \& Owen(2001)]{mo01} Miller, N.~A., \& Owen, F.~N.\ 2001,
359: \apj, 554, 125
360:
361: \bibitem[Poggianti et al.(1999)]{pog99} Poggianti, B. M., Smail, I.,
362: Dressler, A., Couch, W. J., Barger, J., Butcher, H., Ellis, E. S., \&
363: Oemler, A., Jr., 1999, ApJ, 518, 576
364:
365: \bibitem[Tran et al.(2003)]{tran03} Tran, K.-V.~H., Franx, M.,
366: Illingworth, G., Kelson, D.~D., \& van Dokkum, P.\ 2003, \apj, 599, 865
367:
368: \bibitem[Yang et al.(2004)]{yang04} Yang, Y., Zabludoff, A.~I., Zaritsky, D.,
369: Lauer, T.~R., \& Mihos, J.~C.\ 2004, \apj, 607, 258
370:
371: \bibitem[Yang et al.(2008)]{yang08} Yang, Y., Zabludoff, A.~I.,
372: Zaritsky, D.\ 2008, \apj, submitted, arXiv:0801.1190
373:
374: \bibitem[Zabludoff et al.(1996)]{zab96} Zabludoff, A.~I., Zaritsky, D., Lin,
375: H., Tucker, D., Hashimoto, Y., Shectman, S.~A., Oemler, A., \& Kirshner,
376: R.~P.\ 1996, \apj, 466, 104
377: \end{thebibliography}
378:
379: \end{document}
380:
381: %%% Local Variables:
382: %%% mode: latex
383: %%% TeX-master: t
384: %%% End:
385: