astro-ph0311632/ms.tex
1: %\documentclass{aastex}
2: %% preprint produces a one-column, single-spaced document:
3: %\documentclass[12pt,preprint]{aastex}
4: %% preprint2 produces a double-column, single-spaced document:
5: %\documentclass[preprint2]{aastex}
6: 
7: % Emulate ApJ document - for a nice 2 column compact form
8: \documentclass{aastex}
9: \usepackage{emulateapj5}
10: 
11: \slugcomment{To appear in ApJ Letters}
12: 
13: \shorttitle{UV and mid-IR emission in high-z IR galaxies}
14: \shortauthors{Charmandaris, Le Floc'h, \& Mirabel}
15: 
16: 
17: 
18: \begin{document}
19: 
20: \title{A bias in optical observations of high redshift luminous infrared galaxies}
21: 
22: 
23: \author{V. Charmandaris\altaffilmark{1}}
24: \affil{Astronomy Department, Cornell University, Ithaca, NY 14853, USA}
25: \email{vassilis@astro.cornell.edu}
26: 
27: \author{E. Le Floc'h\altaffilmark{2}, I.F. Mirabel\altaffilmark{3}}
28: \affil{CEA/DSM/DAPNIA, Service d'Astrophysique, F-91191 Gif-sur-Yvette, France}
29: \email{elefloch@as.arizona.edu,fmirabel@cea.fr}
30: 
31: %% Notice that each of these authors has alternate affiliations, which
32: %% are identified by the \altaffilmark after each name.  Specify alternate
33: %% affiliation information with \altaffiltext, with one command per each
34: %% affiliation.
35: 
36: 
37: \altaffiltext{1}{Chercheur Associ\'e, Observatoire de Paris, LERMA, 61 Av. de l'Observatoire, F-75014 Paris, France}
38: 
39: \altaffiltext{2}{Now at Steward Observatory, University of Arizona, 933 North Cherry Avenue, Tucson, AZ 85721, USA}
40: 
41: \altaffiltext{3}{Instituto de Astronom\'\i a y F\'\i sica del Espacio, cc 67, suc 28. 1428 Buenos Aires, Argentina}
42: 
43: \begin{abstract}
44: We present evidence for the dramatically different morphology between
45: the rest frame UV and 7$\mu$m mid-IR emission of VV\,114 and Arp\,299,
46: two nearby (z$\sim$0) violently interacting infrared luminous galaxies
47: (LIRGs). Nearly all LIRGs are interacting systems and it is currently
48: accepted that they dominate the IR emission at z$>$1. Luminous IR
49: galaxies located at z=1--2 could easily be detected as unresolved
50: sources in deep optical/near-IR ground based surveys, as well as in
51: upcoming 24$\mu$m surveys with the Space Infrared Telescope
52: Facility. We demonstrate that the spatial resolution of these surveys
53: will result in blending of the emission from unresolved interacting
54: components. An increased scatter will thus be introduced in the
55: observed optical to mid-IR colors of these galaxies, leading to a
56: systematic underestimation of their dust content.
57: 
58: \end{abstract}
59: 
60: \keywords{infrared: galaxies --  
61: 	ultraviolet: galaxies --
62: 	galaxies: individual (Arp\,299, VV\,114)
63: 	galaxies: high-redshift --
64: 	galaxies: starburst --
65: 	galaxies: irregular
66: }
67: 
68: 
69: \section{Introduction}
70: 
71: Over the past decade several studies on the physical characteristics
72: of high redshift galaxies have revealed a number of new exciting
73: results. In particular it has been established that the star formation
74: rate (SFR) of both individual galaxies, as well as the SFR per
75: co-moving volume, increases by more than an order of magnitude as we
76: move from systems of the local Universe to those found to a redshift
77: of z$\ge$2 \citep[e.g.,][]{Steidel96,Madau98}. The details on exactly
78: how the SFR evolves at even higher redshifts are still under debate as
79: they are strongly affected by small number statistics on the
80: properties of the population of the sources detected in sub-mm
81: surveys, but more importantly by the accurate modeling of the effects
82: of dust extinction. A number of deep surveys --- most notably the
83: Hubble Deep Field --- have also shown that a large number of faint
84: galaxies at intermediate and high redshift display irregular and/or
85: disturbed morphology as well as an excess of blue light emission
86: \citep[see][]{Williams96,Abraham96,LeFevre00}. A way to explain this
87: faint blue excess in the galaxy luminosity function is by introducing
88: a rapidly evolving population of star forming galaxies at those
89: redshifts \citep{Glazebrook95,Metcalfe01}. What are the properties
90: though, of this population? Several researchers suggest that these
91: distant galaxies are a new class of systems which are physically small
92: in size \citep{Pascarelle96}, while others speculate that they are
93: high redshift analogues of local interacting galaxies. The latter
94: scenario can be understood since those distant galaxies may look
95: peculiar because of the k-correction and band shifting of their short
96: wavelength (UV) emission to the optical HST filters \citep[][~and
97: references therein]{Hibbard97}.
98: 
99: Two additional independent findings seem to suggest that galaxy
100: interactions may indeed have significant implications to the blue
101: excess in the galaxy luminosity function. The first one is the ample
102: evidence that the galaxy interaction/merge rate increases as a power
103: law function of redshift $\sim$(1+z)$^m$
104: \citep{Carlberg90,Lavery96}. Even though a direct observational
105: measure of the interacting/merging galaxy pairs, and hence $m$, has
106: been difficult due to a number of potential biases in surveys, current
107: estimates indicate that $m$ can be as high as 3--4.5 \citep[see][~for
108: a review]{Lavery96,LeFevre00}. The second is that phenomena associated
109: with galaxy interactions such as the presence of starburst, infrared
110: luminous, and ultraluminous (LIRGs/ULIRGs) galaxies detected with IRAS
111: also increase with
112: redshift\citep[e.g.,][]{Lonsdale90,Sanders96,Xu01}. Tidally triggered
113: massive star formation in galaxies, which may eventually transform gas
114: rich galaxies into their ultraluminous phase, will be responsible for
115: increasing their rest frame UV flux and hence for the blue color
116: excess of the galaxies detected in deep surveys. It is also the reason
117: why large quantities ($\sim$10$^9$M$_{\sun}$) of molecular gas may be
118: compressed over an area of few hundred pc around their nuclei
119: \citep{Downes98,Bryant99} eventually obscuring them behind large
120: quantities of dust. As a result, wavelength dependent extinction by
121: dust grains distorts the morphology of interacting galaxies, limiting
122: our ability to accurately probe their properties .  A prime example of
123: this phenomenon was displayed in the study of NGC\,4038/39 (``The
124: Antennae'') by \citet{Mirabel98}. This system of interacting galaxies,
125: located at distance of 20\,Mpc is only moderately infrared
126: luminous\footnote{We use the standard definition of L$_{\rm
127: IR}$(8--1000\,$\mu$m) = 5.62$\times$10$^5$\,D(Mpc)$^2$\,(13.48{\it
128: f}$_{12}$+5.16{\it f}$_{25}$+2.58{\it f}$_{60}$+{\it
129: f}$_{100}$)\,L$_\sun$ \citep[see][]{Sanders96},
130: H$_0$\,=\,75\,km\,s$^{-1}$\,Mpc$^{-1}$, and q$_o$\,=\,0.5 to
131: facilitate comparison with earlier work.}  (L$_{\rm
132: IR}$=4.9$\times10^{10}$\,L$_{\sun}$) and still contains considerable
133: quantities of molecular gas dispersed over a large area, which may
134: become available for a more intense star formation event in the future
135: \citep{Wilson00}. Despite that, half of the mid-IR flux of the system
136: originates from the optically obscured overlap region between the two
137: nuclei \citep[see][]{Mirabel98}.
138: 
139: It is thus reasonable to pose the following questions. How would the
140: morphology of systems more IR luminous than NGC\,4038/39 be affected
141: by the presence of dust and what consequences this may have in
142: measurements of the {\em spatially unresolved} IR luminous galaxies
143: which are readily discovered at high (z$\gtrsim1.5$) redshift? More
144: specifically, how might this affect the conclusions of color
145: correlations between a number of optical and near-IR surveys that are
146: performed from the ground \citep[e.g.,][]{Jannuzi99}, and those
147: scheduled to be performed with the Space Infrared Telescope Facility
148: (SIRTF) \citep[e.g.,][]{Rieke00,Lonsdale03}?
149: 
150: We will base our analysis on high resolution UV and mid-IR
151: observations of Arp\,299 and VV\,114 that became recently
152: available. Both galaxies have IR luminosities almost an order of
153: magnitude higher than NGC\,4038/39 (L$_{\rm
154: IR}\sim$5$\times10^{11}$L$_\sun$ for Arp\,299 and
155: $\sim$4$\times10^{11}$L$_\sun$ for VV\,114) but the projected spatial
156: separation between their two interacting components is still
157: sufficiently large (8 and 6\,kpc respectively) for us to be able to
158: resolve them given their proximity (41 and 80\,Mpc respectively).
159: 
160: 
161: \section{Observations}
162: 
163: All observations used in this paper were retrieved from previously
164: published work and we refer the reader to those publications for
165: details on the data reduction. The mid-IR data were obtained with
166: ISOCAM on board the Infrared Space Observatory (ISO) and were
167: presented by \citet{LeFloch02} for VV\,114 and by
168: \citet{Charmandaris02} for Arp\,299. The UV data of both galaxies were
169: retrieved from the HST archive. An analysis of the STIS images of
170: VV\,114 and FOC images Arp\,299 is included in \citet{Goldader02} and
171: \citet{Meurer95} respectively.
172: 
173: 
174: \section{Discussion}
175: 
176: \subsection{Dust and interactions at high-z}
177: 
178: A detailed study of the apparent morphology of peculiar galaxies as a
179: function of redshift was already performed by \citet{Hibbard97} by
180: examining UV and optical images of a sample of 4 interacting
181: galaxies. These authors found that at z$>$1.5 it is no longer possible
182: to discern features such as long tidal tails and bridges that would
183: identify these systems as clearly ``interacting'', even in exposures
184: as deep as the Hubble Deep Field. These authors also alluded that the
185: redshifted rest UV light, which will be sampled in the optical if
186: those galaxies are at z$>$1.5, can not be used to accurately image the
187: evolved stellar population.
188: 
189: The limitations discussed by \citet{Hibbard97} become even stronger if
190: one considers dust extinction\footnote{We note that if we were to
191: trace dust using the usual extinction law A$_V$, then A$_{\rm
192: UV}$$\sim$7A$_{\rm H_\alpha}$$\sim$270A$_{7\mu m}$
193: \citep{Mathis90}.}. Even though numerous z$\sim$2--3 galaxies have
194: been detected as UV drop outs, the actual magnitude of extinction
195: effects in the detection of distant galaxies using their rest frame UV
196: emission is still under debate \citep[see][~for conflicting
197: views]{Meurer99,Adelberger00}. However, a number of recent results
198: clearly suggest the presence of considerable quantities of dust at
199: high redshifts. Based on a UV selected sample of 16 galaxies at
200: 2$\leq$z$\leq$2.6, \citet{Erb03} found that the SFR, measured using
201: the redshifted into the K-band H$\alpha$ line, is not only elevated
202: (16\,M$_{\sun}$yr$^{-1}$) but it is also a factor of $\sim$2.4 higher
203: than that estimated using their UV continuum (SFR$_{\rm UV}$). This
204: discrepancy in estimating the SFR is even higher for the systems with
205: low UV luminosity and can be easily understood due to the presence of
206: dust. Unfortunately no H$\beta$ line was detected from these galaxies
207: to measure their Balmer decrement, and consequently we can not be
208: certain whether the factor of 2.4 is just a lower limit. Furthermore,
209: UV selected samples may be biased towards systems with a low dust
210: content. A similar underestimate in SFR$_{\rm UV}$ compared to the one
211: calculated using the H$\beta$ line was found for a small sample of
212: galaxies at z=3 by \citet{Pettini01}.  Even at lower redshifts though
213: (z$\sim$1), where larger samples and better statistics are available,
214: the presence of dust is evident and new rather interesting trends
215: emerge. In an analysis of a large sample of galaxies located at
216: 0.4$\leq$z$\leq$0.8 \citet{Cardiel03} estimated their SFR using the IR
217: luminosity as well as their H$\alpha$ emission. They found a clear
218: trend indicating that as one moves from lower to higher luminosity
219: systems (L$_{\rm IR}=10^{11}$--$10^{12}$\,L$_{\sun}$) the SFR$_{\rm
220: H\alpha}$ is up to {\em an order of magnitude less} than the SFR$_{\rm
221: IR}$ (see their Fig. 15) in agreement with the original work by
222: \citet{Armus90} based on IRAS data.
223: 
224: The result of \citet{Cardiel03} further supports our concern that the
225: effects of dust in the z$\sim$2 IR luminous systems mentioned earlier
226: may not have been properly taken into account. It also stresses the
227: need to better quantify the correlation between the rest frame mid-IR
228: emission and the L$_{\rm IR}$ of IR luminous galaxies \citep{Chary01},
229: since a strong active galactic nucleus (AGN) can also contribute to
230: the mid-IR \citep[see][]{Laurent00,Tran01}.
231: 
232: 
233: \subsection{Arp\,299 and VV\,114 analogues in high-z surveys}
234: 
235: To better illustrate another aspect of the influence of dust in the
236: morphology of an IR luminous galaxy, we decided to compare the UV and
237: 7$\mu$m images of Arp\,299 and VV\,114, the two nearest interacting
238: pairs of galaxies with L$_{\rm IR}>10^{11}$, which at the same time
239: are sufficiently extended in the sky (see Fig.~1). Both systems show a
240: clearly resolved eastern and western component. As they have been
241: studied extensively in all wavelengths we had access to published high
242: resolution UV images with HST, mid-IR images with Keck
243: \citep{Soifer01}, as well as our own deep mid-IR images obtained with
244: ISOCAM \citep[see][ and references therein]{LeFloch02,Charmandaris02}.
245: 
246: The selection of the rest frame $\sim$0.2\,$\mu$m (UV) and
247: $\sim$7\,$\mu$m (mid-IR) bands was determined by the fact that at
248: z$\sim$2 these would be observed in R-/I-bands and at $\sim$24\,$\mu$m
249: respectively.  The ongoing NOAO Deep Wide-Field Survey\footnote{For
250: more information on the NOAO Deep Wide-Field Survey visit
251: http://www.noao.edu/noao/noaodeep} \citep{Jannuzi99} is covering those
252: optical bands and it is sufficiently sensitive to detect the rest
253: frame UV emission from Arp\,299 and VV\,114 (see Table 1), as well as
254: any L$^*$ star-forming galaxy at z$\ge$2 (detection limits of
255: $\sim$26\,mag in R and I-bands).  A number of deep mid-IR surveys have
256: also been scheduled to be performed with the Space Infrared Telescope
257: Facility (SIRTF) using the 24\,$\mu$m imaging using the Multiband
258: Imaging Photometer (MIPS\footnote{For more information on MIPS and its
259: sensitivity visit http://mips.as.arizona.edu.})
260: \citep{Rieke00,Lonsdale03}. SIRTF will readily detect
261: $\sim$10$^{11}$\,L$_{\sun}$ starburst and star forming galaxies at
262: z=2, since the rest frame 7.7\,$\mu$m Polycyclic Aromatic Hydrocarbon
263: (PAH) feature, which is very strong in starburst systems
264: \citep[see][]{Laurent00}, will move into the MIPS 24\,$\mu$m band.
265: 
266:  At a redshift z=2, though, one arcsecond on the sky, the typical
267: resolution of ground imagery, corresponds to a projected distance of
268: 8\,kpc. The spatial resolution of SIRTF is even poorer ($\sim$6.9$''$
269: at 24\,$\mu$m). As a result, systems similar to Arp\,299 and VV\,114
270: will be unresolved in these surveys.
271: 
272: Observing the two galaxies in Fig. 1 the striking difference between
273: the UV and 7$\mu$m emission is obvious. In Arp\,299 we note that
274: $\sim$20\% of the UV emission originates from the eastern component
275: (IC\,694) and $\sim$80\% is associated with NGC\,3690. However, at 7
276: and 15\,$\mu$m IC\,694 becomes progressively more luminous relative to
277: NGC\,3690, contributing $\sim$30\% and $\sim$70\% of the flux
278: respectively. IC\,694 eventually dominates the IR luminosity of
279: Arp\,299, contributing twice as much to it than NGC\,3690. As we can
280: see by the displaced centroid between the mid-IR and UV emission,
281: discussed in \cite{Charmandaris02}, all UV and optical emission
282: detected from IC\,694 is probably due to surface star formation
283: activity. The emission from the massive stars formed by the high
284: density molecular gas in the nucleus of IC\,694 only becomes visible in
285: the near-IR and mid-IR wavelengths. The situation in VV\,114 is even
286: more extreme. Although VV\,114E is 2.5 times brighter than VV\,114W in
287: the mid-IR it is practically invisible in the UV!  VV\,114E actually
288: displays strong evidence that a considerable fraction of its mid-IR
289: flux originates from an AGN \citep{LeFloch02,Soifer01}, which further
290: enhances its mid-IR emission. The relative contribution of both galaxy
291: components to the total emission of each system in the the UV and
292: mid-IR is presented in Table~1.
293: 
294: In Table 1 we also include the 7\,$\mu$m to UV flux density ratio,
295: $X^{MIR}_{UV}$, for each galaxy, as well as the global value for each
296: interacting pair. A scatter in those values is substantial. We observe
297: that the highly extinct galaxy of the pair has an $X^{MIR}_{UV}$ ratio
298: which can be 1.7 to 24 times greater than the corresponding ratio of
299: the whole Arp\,299 or VV\,114 respectively!  Furthermore, in both
300: cases the extreme characteristics of the redder/eastern galaxy, which
301: is actually responsible for the bulk of the L$_{\rm FIR}$ of the
302: system, remain inconspicuous if we restrict our observations to the
303: rest frame UV and optical observations.  When considered as a single
304: system the UV emission from the less obscured component dilutes the
305: signature of warm dust emission. Hence, if one is to use the
306: $X^{MIR}_{UV}$ reddening or even estimate the slope $\beta$ of the UV
307: spectrum to establish a $\beta$ to L$_{FIR}$ correlation as tracer of
308: extinction and total dust content \citep[i.e.,][]{Meurer99} in IR
309: luminous systems, a systematic underestimate of the dust content will
310: be introduced\footnote{Unfortunately, lack of high spatial resolution
311: UV and mid-IR maps for a statistically significant sample of local IR
312: luminous galaxies makes a quantitative prediction of this scatter
313: impossible.}.
314: 
315: This fact has serious implications when one tries to examine the
316: colors of distant high redshift sources using deep surveys with low or
317: moderate spatial resolution. All IR luminous galaxies are very likely
318: interacting systems. A significant fraction of them, similarly to the
319: two cases presented above, will remain as unresolved point
320: sources. Even though the actual value of the rest frame mid-IR to UV
321: ratio depends on the details of the filters used to observe the high-z
322: sources (R,I, and/or MIPS24), and they are beyond the scope of this
323: letter, one result is clear from this exercise. A simple diagnostic
324: which correlates integrated rest frame UV with mid-IR or radio
325: measurement is bound to introduce a systematic scatter not just due to
326: the unknown quantities of dust at high-z, but principally due to the
327: blending of the emission from the unresolved interacting
328: components. Depending on the stage of the interaction and the amount
329: of spatially extended star formation activity, there will be cases
330: (such as VV\,114) where this scatter may be extreme. The situation
331: will be even more complicated if an obscured AGN is present since this
332: will increase the 3--6\,$\mu$m mid-IR emission but it will have
333: minimal effects in the UV flux. To address these limitations,
334: follow-up observations with deep near-IR sub-arcsec imaging using the
335: Hubble Space Telescope or high quality ground facilities will be
336: necessary in order to reveal whether those systems are indeed
337: substantially disturbed and help us understand their physical
338: properties.
339: 
340: \subsection{Frequency of ``VV\,114''-type systems at high-z}
341: 
342: An interesting question to address would be how important/numerous
343: galaxies with properties similar to VV\,114 are at high-z. The fact
344: that the extreme characteristics of VV\,114 were realized based on
345: high resolution UV and mid-IR maps, which are available for only a
346: handful of the nearby LIRGs/ULIRGs, makes such a prediction
347: challenging. However, if we use the 1Jy sample of \citet{Kim02} as a
348: guide, we find that $\sim$10\% of the systems have interacting pairs
349: with similar spatial separation to VV\,114 and an R-K color of one
350: galaxy in the system being more than 2 mags different than the other
351: one. For VV\,114 this color difference is ~3.5\,mag. Even though LIRGs
352: and ULIRGs are rare in the local Universe contributing to less than
353: 6\% of the local infrared emission \citep[see][]{Sanders96}, they are
354: much more frequent at high-z. The ISOCAM deep surveys clearly
355: demonstrate that at z$\sim$1 nearly $\sim$65\% of the peak and
356: integrated cosmic infrared background originates from galaxies. Of
357: those galaxies, $\sim$75\% are LIRGs and ULIRGs \citep[see review
358: by][]{Elbaz03}. Furthermore, as we mentioned earlier, it is accepted
359: that LIRGs/ULIRGs are the result of galaxy merging and it has been
360: found that even though interacting systems comprise only 3--9\% of all
361: local galaxies \citep{Arp75,Struck99}, the merger rate increases
362: rapidly with redshift (1+z)$^{4-6}$ out to z$\sim$3
363: \citep{Conselice03} and most high-z galaxies appear somewhat
364: perturbed.  Based on the above results a simple scaling would suggest
365: that systems such as VV\,114 may contribute $\sim$5--10\% to the
366: galaxy counts in the infrared at z$\ge$1. So even though extremely
367: rare locally, systems such as VV\,114 may have profound effects in our
368: understanding of the early Universe.
369: 
370: \section{Conclusions}
371: 
372: Using archival images of VV\,114 and Arp\,299, two nearby violently
373: interacting infrared luminous galaxies, we show the strikingly
374: different morphology between their 0.2\,$\mu$m UV and 7\,$\mu$m mid-IR
375: emission. Even though the redshifted rest frame emission from both
376: wavelengths will be easily detected at z$\sim$2 by ground based
377: R-/I-band surveys and by the 24\,$\mu$m surveys of SIRTF respectively,
378: the fact that we will not be able to spatially resolve the different
379: components of each source will introduce a scatter in their measured
380: colors. This scatter will not only depend on the relative contribution
381: of each interacting component to the total IR luminosity, but it will
382: also cause a systematic underestimate of the integrated rest frame
383: mid-IR to UV colors leading to lower values of the total dust content
384: of those systems.
385: 
386: \acknowledgments
387: 
388: We thank the anonymous referee whose comments helped us improve this
389: paper. VC would like to thank J.R. Houck (Cornell), D. Elbaz
390: (CEA/Saclay), L. Armus (Caltech), J.Hibbard (NRAO), and J.D. Smith
391: (Arizona) for useful discussions, as well as to acknowledge the support
392: of JPL contract 960803.
393: 
394: \begin{thebibliography}{}
395: 
396: \bibitem[Abraham et al.(1996)]{Abraham96} 
397:         Abraham, R.~G., Tanvir, N.~R., Santiago, B.~X., Ellis, R.~S.,
398:         Glazebrook, K., \& van den Bergh, S.\ 1996, \mnras, 279, L47
399: 
400: \bibitem[Adelberger \& Steidel(2000)]{Adelberger00} 
401: 	Adelberger, K.~L.,~\& Steidel, C.~C.\ 2000, \apj, 544, 218
402: 
403: \bibitem[Armus et al.(1990)]{Armus90}	
404: 	Armus, L., Heckman, T. M., \& Miley, G. K.\ 1990, \apj, 364, 471
405: 
406: \bibitem[Arp \& Madore(1975)]{Arp75} 
407:         Arp, H.~A.,~\& Madore, B.~F.\ 1975, The Observatory, 95, 212
408: 
409: \bibitem[Bryant \& Scoville(1999)]{Bryant99} 
410: 	Bryant, P.~M.,~\& Scoville, N.~Z.\ 1999, \aj, 117, 2632
411: 
412: \bibitem[Cardiel et al.(2003)]{Cardiel03} 
413: 	Cardiel, N., Elbaz, D., Schiavon, R.~P., Willmer, C.~N.~A.,
414: 	Koo, D.~C., Phillips, A.~C., \& Gallego, J.\ 2003, \apj, 584,
415: 	76
416: \bibitem[Carlberg(1990)]{Carlberg90} 
417: 	Carlberg, R.~G.\ 1990, \apjl, 359, L1
418: 
419: \bibitem[Charmandaris et al.(2002)]{Charmandaris02}
420:         Charmandaris, V., Stacey, G. J., \& Gull, G. 2002, \apj, 571,
421:         282
422: 
423: \bibitem[Chary \& Elbaz(2001)]{Chary01} 
424: 	Chary, R.,~\& Elbaz, D.\ 2001, \apj, 556, 562
425: 
426: \bibitem[Conselice et al.(2003)]{Conselice03} 
427:         Conselice, C.~J., Bershady, M.~A., Dickinson, M., \& Papovich,
428: 	C.\ 2003, \aj, 126, 1183
429: 
430: \bibitem[Downes \& Solomon(1998)]{Downes98} 
431: 	Downes, D.,~\& Solomon, P.~M.\ 1998, \apj, 507, 615
432: 
433: \bibitem[Elbaz \& Cesarsky(2003)]{Elbaz03} 
434:         Elbaz, D.~\&   Cesarsky, C.~J.\ 2003, Science, 300, 270 
435: 
436: \bibitem[Erb et al.(2003)]{Erb03}
437:         Erb, D. W., Shapley, A. E., Steidel, C. C., Pettini, M.,
438:         Adelberger, K. L., Hunt, M. P., Moorwood, A. F. M., \& Cuby,
439:         J.-G. 2003, \apj, 591, 101
440: 
441: \bibitem[Glazebrook et al.(1995)]{Glazebrook95} 
442:         Glazebrook, K., Ellis, R., Santiago, B., \& Griffiths, R.\
443: 	1995, \mnras, 275, L19
444: 
445: \bibitem[Goldader et al.(2002)]{Goldader02} 
446:         Goldader, J.~D., Meurer, G., Heckman, T.~M., Seibert, M.,
447:         Sanders, D.~B., Calzetti, D., \& Steidel, C.~C.\ 2002, \apj,
448:         568, 651
449: 
450: \bibitem[Hibbard \& Vacca(1997)]{Hibbard97}
451:         Hibbard, J. E., \& Vacca, W. D. 1997, \aj, 114, 1741
452: 
453: \bibitem[Jannuzi \& Dey(1999)]{Jannuzi99} 
454: 	Jannuzi, B.~T.,~\& Dey, A.\ 1999, ASP Conf.~Ser.~191:
455: 	Photometric Redshifts and the Detection of High Redshift
456: 	Galaxies, 111
457: 
458: \bibitem[Kim et al.(2002)]{Kim02} 
459:         Kim, D.-C., Veilleux, S., \& Sanders, D.~B.\ 2002, \apjs, 143,
460: 	277
461: 
462: \bibitem[Lavery et al.(1996)]{Lavery96} 
463: 	Lavery, R.~J., Seitzer, P., Suntzeff, N.~B., Walker, A.~R., \&
464: 	Da Costa, G.~S.\ 1996, \apjl, 467, L1
465: 
466: \bibitem[Laurent et al.(2000)]{Laurent00}
467:         Laurent, O., Mirabel, I. F., Charmandaris, V., Gallais, P.,
468:         Sauvage, M., Vigroux, L., \& Cesarsky, C. 2000, \aap, 359, 885
469: 
470: \bibitem[Le F{\` e}vre et al.(2000)]{LeFevre00} 
471: 	Le F{\` e}vre, O.,~et al.\ 2000, \mnras, 311, 565
472: 
473: \bibitem[Lonsdale, et al.(1990)]{Lonsdale90} 
474: 	Lonsdale, C.~J., Hacking, P.~B., Conrow, T.~P., \&
475: 	Rowan-Robinson, M.\ 1990, \apj, 358, 60
476: 
477: \bibitem[Lonsdale(2003)]{Lonsdale03} 
478: 	Lonsdale, C.~J.,~et al.\ 2003, \pasp, 115, 897 
479: 
480: \bibitem[Le Floc'h et~al.(2002)]{LeFloch02}   
481: 	Le Floc'h, E., Charmandaris V., Laurent, O., Mirabel, I. F.,
482: 	Gallais, P., Sauvage, M., Vigroux, L., \& Cesarsky, C. 2002,
483: 	\aap, 391, 417
484: 
485: \bibitem[Madau et al.(1998)]{Madau98} 
486:         Madau, P., Pozzetti, L., Dickinson, M. 1998, \mnras, 498, 106
487: 
488: \bibitem[Mathis(1990)]{Mathis90}
489:    	Mathis, J. S. 1990, ARA\&A, 28, 37
490: 
491: \bibitem[Metcalfe et al.(2001)]{Metcalfe01} 
492:         Metcalfe, N., Shanks, T., Campos, A., McCracken, H.~J., \&
493: 	Fong, R.\ 2001, \mnras, 323, 795
494: 
495: \bibitem[Meurer et al.(1995)]{Meurer95} 
496: 	Meurer, G.~R., Heckman, T.~M., Leitherer, C., Kinney, A.,
497: 	Robert, C., \& Garnett, D.~R.\ 1995, \aj, 110, 2665
498: 
499: \bibitem[Meurer et al.(1999)]{Meurer99} 
500: 	Meurer, G.~R., Heckman, T.~M., \& Calzetti, D.\ 1999, \apj,
501: 	521, 64
502: 
503: \bibitem[Mirabel et al.(1998)]{Mirabel98}
504: 	Mirabel, I.~F., et al., 1998, \aap, 333, L1
505: 
506: \bibitem[Pascarelle et al.(1996)]{Pascarelle96} 
507:         Pascarelle, S.~M., Windhorst, R.~A., Keel, W.~C., \& Odewahn,
508: 	S.~C.\ 1996, \nat, 383, 45
509: 
510: \bibitem[Pettini et al.(2001)]{Pettini01} 
511: 	Pettini, M., Shapley, A.~E., Steidel, C.~C., Cuby, J.,
512: 	Dickinson, M., Moorwood, A.~F.~M., Adelberger, K.~L., \&
513: 	Giavalisco, M.\ 2001, \apj, 554, 981
514: 
515: \bibitem[Rieke et al.(2000)]{Rieke00} 
516: 	Rieke, M.~J., et al.\ 2000, AAS Meeting, 197,~\#77.13
517: 
518: \bibitem[Sanders \& Mirabel(1996)]{Sanders96}
519: 	Sanders, D. B., \& Mirabel, I. F. 1996, ARA\&A, 34, 749
520: 
521: \bibitem[Soifer et al.(2001)]{Soifer01}
522:    	Soifer, B. T., et al., 2001, \aj, 122, 1213
523: 
524: \bibitem[Steidel et al.(1996)]{Steidel96} 
525:         Steidel, C.~C., Giavalisco, M., Pettini, M., Dickinson, M., \&
526: 	Adelberger, K.~L.\ 1996, \apjl, 462, L17
527: 
528: \bibitem[Struck(1999)]{Struck99} 
529:         Struck, C.\ 1999, \physrep, 321, 1 
530: 
531: \bibitem[Tran et al.(2001)]{Tran01} 
532: 	Tran, Q.~D.,~et al.\ 2001, \apj, 552, 527
533: 
534: \bibitem[Williams et al.(1996)]{Williams96}
535:         Williams, R. E., et al. 1996, \aj, 112, 1335
536: 
537: \bibitem[Wilson et al.(2000)]{Wilson00}
538:         Wilson, C. D., Scoville, N. Z., Madden, S. C., \&
539:         Charmandaris, V.  2000, \apj, 542, 120
540: 
541: \bibitem[Xu et al.(2001)]{Xu01} 
542: 	Xu, C., Lonsdale, C.~J., Shupe, D.~L., O'Linger, J., \& Masci,
543: 	F.\ 2001, \apj, 562, 179
544: 
545: \end{thebibliography}
546: 
547: \newpage 
548: 
549: \clearpage
550: 
551: \begin{deluxetable}{lcccccc}
552: %\rotate
553: \tabletypesize{\scriptsize} 
554: \tablecaption{Rest Frame near-UV and mid-IR emission from Arp\,299 and VV\,114\label{tbl}}
555: \tablewidth{0pc}
556: \startdata \\
557: \tableline 
558: \tableline \\
559: 			& Arp\,299 & IC\,694 & 	NGC\,3690 & VV\,114 &
560: 			VV\,114E & VV\,114W \\
561: \tableline \\
562: UV ($\sim$0.22\,$\mu$m) (mag)\tablenotemark{a}	& 12.73   & 14.55 & 12.95  & 14.49	& 18.38    & 14.53 \\
563: mid-IR ($\sim$7\,$\mu$m) (mJy)\tablenotemark{b}	& 1032    & 325	  & 707	   & 202 	& 136      & 66 \\
564: $X^{MIR}_{UV}$=7\,$\mu$m / UV \tablenotemark{c}	        & 196	  & 330	   & 164	& 34	& 819      & 11 \\
565: \tablenotetext{a}{UV magnitudes based on Table 6 of \citet{Meurer95} and Table 3 of \citet{Goldader02}.}
566: \tablenotetext{b}{The 7$\mu$m (LW2) ISOCAM flux density over the same regions using data from \citet{Charmandaris02} and \citet{LeFloch02}.}
567: \tablenotetext{c}{The non dimentional $X^{MIR}_{UV}$ ratio is calculated after converting the UV mag to $f_{\nu}$ in mJy.}
568: \enddata
569: \end{deluxetable}
570: 
571: \clearpage 
572: \begin{figure} 
573: \figurenum{1} 
574: \plotone{f1.eps}
575: 
576:   \caption{a) An HST/STIS image of VV\,114 at 0.23\,$\mu$m, adapted
577:   from \citet{Goldader02} with an overlay of the 7$\mu$m emission from
578:   \citet{LeFloch02}. The images are uniform and the UV flux limit is
579:   26.6\,mag\,arcsec$^{-2}$. The contour levels are set with
580:   logarithmic spacing between 0.8 and 8.4\,mJy\,arcsec$^{-2}$ (
581:   1$\sigma$ $\sim$ 0.15\,mJy\,arcsec$^{-2}$). b) An archival HST/FOC
582:   (0.22\,$\mu$m) UV image of Arp\,299 with the 7$\mu$m emission
583:   contours from \citet{Charmandaris02}. The images are uniform and the
584:   UV flux limit is 21\,mag\,arcsec$^{-2}$.The contour limits are set
585:   to 1.2 and 30.7\,mJy\,arcsec$^{-2}$ (1$\sigma$ $\sim$
586:   0.27\,mJy\,arcsec$^{-2}$). Note the offset in the 7$\mu$m centroid
587:   of IC\,694 when compared to the underying UV emission. c) The same
588:   UV image of Arp\,299 with an overlay of the 38$\mu$m emission from
589:   \citet{Charmandaris02}.  The levels are 1.5\,Jy\,beam$^{-1}$
590:   beginning at 3\,Jy\,beam$^{-1}$ (6\,$\sigma$).  We can easily see
591:   that in both systems the rest frame UV light from the mid-IR
592:   dominant source is either completely suppressed (VV\,114) or
593:   marginally detected (Arp\,299), leading to the high scatter in the
594:   $X^{MIR}_{UV}$ values of Table 1. For reference we include in a) a
595:   horizontal bar which indicates the physical scale of 8\,kpc that
596:   1\,arcsec (the typical angular resolution of ground optical deep
597:   surveys) would cover if VV\,114 were at z=2. Note that the angular
598:   resolution of SIRTF at 8 and 24\,$\mu$m is $\sim$2.3 and
599:   $\sim$6.9\,arcsec respectively. \label{fig1}}
600: 
601: \end{figure} 
602: 
603: 
604: \end{document}
605: