astro-ph0307004/ms.tex
1: %  Revised version for re-submission to ApJL  (July 25, 2003)
2: 
3: 
4: \documentclass[12pt,preprint]{aastex}
5: 
6: \newcommand{\simlt}{\mbox{$^{<}_{\sim}$}}
7: \newcommand{\simgt}{\mbox{$^{>}_{\sim}$}}
8: \newcommand{\cd}{cm$^{-2}$}
9: \newcommand{\Msun}{$M_{\odot}$}
10: \newcommand{\kms}{km~s$^{-1}$}
11: \def\h2{H$_{2}$}
12: \newcommand{\e}{$\times$10}
13: \newcommand{\fesc}{$f_{esc}$}
14: \newcommand{\ha}{H$\alpha$}
15: \def\n01{$N_{01}$}
16: \def\t01{$T_{01}$}
17: \def\tex{$T_{\rm ex}$}
18: \def\nd{\nodata}
19: \newcommand\col[2]{$#1 \times 10^{#2}$}%
20: \input epsf
21: 
22: %%%%% Title Page
23: 
24: %\received{2003 June 30}
25: \begin{document}
26: 
27: \title{THE MULTI-PHASE INTERGALACTIC MEDIUM TOWARDS PKS 2155-304 }
28: 
29: \author{J. MICHAEL SHULL\altaffilmark{1,2},
30: JASON TUMLINSON\altaffilmark{3}, and MARK L. GIROUX\altaffilmark{4}
31: } 
32: \altaffiltext{1}{CASA, Department of Astrophysical and Planetary Sciences, 
33:    University of Colorado, Boulder, CO 80309 (mshull@casa.colorado.edu)}
34: \altaffiltext{2}{Also at JILA, University of Colorado and National Institute 
35:    of Standards and Technology}
36: \altaffiltext{3}{Department of Astronomy and Astrophysics, University 
37:    of Chicago, Chicago, IL 60637 (tumlinso@oddjob.uchicago.edu)}
38: \altaffiltext{4}{Department of Physics and Astronomy, Box 70652,
39:    East Tennessee State University, Johnson City, TN 37614 (giroux@polar.etsu.edu)}
40: 
41: 
42: \begin{abstract}
43: 
44: We study the cluster of H~I and O~VI absorption systems and the 
45: claimed detection of O~VIII absorption from the intergalactic medium 
46: at $z \approx 0.0567$, associated with a group of galaxies toward 
47: the BL~Lac object PKS~2155-304. As measured by spectrographs on the 
48: {\it Hubble Space Telescope}, {\it Far Ultraviolet Spectroscopic Explorer}, 
49: and {\it Chandra}, this system appears to contain gas at a variety of 
50: temperatures.  We analyze this multi-phase gas in a clumpy-infall model. 
51: From the absence of C~IV and Si~III absorption in the Ly$\alpha$ clouds, 
52: we infer metallicities less than 2.5--10\% of solar values.  The only 
53: metals are detected in two O~VI absorption components, offset by 
54: $\pm 400$ km~s$^{-1}$ from the group barycenter ($cz \approx 16,600$ 
55: km~s$^{-1}$).  The \ion{O}{6} components may signify ``nearside" and 
56: ``backside" infall into the group potential well, which coincides with 
57: the claimed O~VIII absorption. 
58: If the claimed O~VIII detection is real, our analysis suggests that
59: clusters of strong Ly$\alpha$ and O~VI absorbers, 
60: associated with groups of galaxies, may be the ``signposts" of 
61: shock-heated, metal-enriched baryons.  Through combined UV and X-ray
62: spectra of H~I and O~VI, VII, and VIII, one may be able to clarify  
63: the heating mechanism of this multiphase gas.  
64: \end{abstract}
65: 
66: \keywords{intergalactic medium --- quasars: absorption lines --- 
67:     ultraviolet: general}
68: 
69: \section{INTRODUCTION}
70: 
71: With the development of high-throughput ultraviolet and X-ray spectrographs,
72: we can begin to account for a significant fraction of the low-redshift 
73: ``missing baryons" (Fukugita, Hogan, \& Peebles 1998; Shull 2003) in the 
74: intergalactic medium (IGM).  
75: Recent surveys by the {\it Hubble Space Telescope} (HST) and the
76: {\it Far Ultraviolet Spectroscopic Explorer} (FUSE) show that $32\pm6$\% 
77: of the baryons reside in the warm ($10^4$~K) Ly$\alpha$ forest (Penton, Stocke,
78: \& Shull 2003), while $5-10$\% lie in the shock-heated ($10^{5-6}$~K) 
79: IGM traced by O~VI absorption (Tripp, Savage, \& Jenkins 2000).    
80: A predicted 30-40\% of the IGM remains to be found in even hotter gas 
81: ($10^{6-7}$~K) through X-ray absorption lines of O~VII, O~VIII, and perhaps
82: Ne~IX (Fang et al.\ 2002; Nicastro et al.\ 2002).  
83: 
84: The total baryon density is known to 10\% accuracy from measurements
85: of light-element nucleosynthesis (D/H) and acoustic oscillations 
86: in the cosmic microwave background (CMB).  Recent estimates from these
87: techniques yield consistent values for $\Omega_b$, the fractional 
88: contribution of baryons to the closure density:   
89: $\Omega_b = (0.020 \pm 0.002) h^{-2}$ (D/H -- Burles \& Tytler 1998) 
90: and $\Omega_b = (0.0224 \pm 0.0009) h^{-2}$ (CMB -- Spergel et al.\ 2003; 
91: Netterfield et al.\ 2002) for a Hubble constant 
92: $H_0 = 100h$ km~s$^{-1}$~Mpc$^{-1}$. Even $H_0$ has become a
93: well-measured parameter, with values of $h = 0.71 \pm 0.08$
94: (Freedman et al.\ 2001) from the Cepheid Key Project with HST. 
95: Using the CMB values, the co-moving baryon density 
96: is $\rho_b = (4.2 \pm 0.2) \times 10^{-31}$ g~cm$^{-3}$ 
97: and the hydrogen density is $n_H = (1.90 \pm 0.2) \times 10^{-7}$
98: cm$^{-3}$ for primordial helium abundance $Y_p = 0.244$ by mass.   
99: 
100: Numerical simulations of the low-$z$ IGM  (Cen \& Ostriker 1999; 
101: Dav\'e et al.\ 2001) predict that the gas is distributed nearly 
102: equally ($30 \pm 10$\% each) in three ``phases": 
103: (1) warm photoionized gas ($10^4$~K); (2) warm/hot shocked gas 
104: ($10^{5-7}$~K); and (3) collapsed halos, galaxies, and clusters with much 
105: hotter gas ($T > 10^7$~K).  The warm phase is observable through UV absorption 
106: lines of H~I, C~III, C~IV, Si~III, and Si~IV, while the warm/hot phase
107: is detectable in O~VI (UV) and in X-ray lines of O~VII, O~VIII, 
108: and Ne~IX.   The warm-hot intergalactic medium (WHIM) is thought to be 
109: produced  by shock-heating during gravitational infall into dark-matter 
110: filaments co-evolving with the rise in cosmic metallicity.
111: These heavy elements are theorized to be expelled from galaxies 
112: by tidal stripping (Gnedin 1998) and ``starburst winds" as 
113: observed by Martin et al.\ (2002).  The metals are then assimilated 
114: into infalling clumps of IGM.  Details of this model remain unclear. 
115: Are the heavy elements expelled from galaxies in hot or warm gas?   How 
116: are the metals mixed with the IGM? What is the extent of their transport?  
117: Addressing these issues requires detecting and correlating many warm and hot 
118: IGM absorbers.
119: 
120: In this paper, we provide clear evidence for multi-phase IGM toward 
121: the BL Lac object PKS~2155-304.  This gas is associated 
122: with a small group of galaxies at $z \approx 0.0567$ and a cluster of 7 
123: strong Ly$\alpha$ absorbers (Shull et al.\ 1998) at velocities  
124: $cz = 16,283 - 17,570$ km~s$^{-1}$. This environment exhibits the diffuse 
125: warm phase (Ly$\alpha$, Ly$\beta$, Ly$\gamma$), the shocked (WHIM) phase 
126: (O~VI), and unconfirmed evidence (O~VIII) for much hotter, virialized gas 
127: residing at the bottom of the group's gravitational potential well.  
128: 
129: In \S~2, we describe our FUSE and HST observations of these absorbers in 
130: H~I and O~VI (and limits on C~IV, Si~III) 
131: and compare them to the claimed O~VIII absorption. 
132: In \S~3 we examine the absorber kinematics for the 
133: possibility of nearside/backside infall and ionization-state (O~VI/O~VIII) 
134: consistency.  We conclude in \S~4 with a discussion of an infall model 
135: that may be consistent with large-scale structure in this group.
136: In this picture, the H~I and O~VI absorbers arise from clumps of gas
137: falling at 200--400 km~s$^{-1}$ relative
138: to a much hotter ($10^{6.5-7.0}$~K) substrate that could be 
139: detectable in O~VII or O~VIII absorption. 
140: The metallicities may range from 0.02--0.1 solar.
141: All the phases predicted by simulations may be present in
142: this group, suggesting that clusters of strong Ly$\alpha$ and O~VI 
143: absorbers associated with groups of galaxies may be the 
144: ``signposts" of shock-heated, metal-enriched baryons. 
145: 
146: \section{OBSERVATIONS}
147: 
148: In HST Cycle 9, we obtained new data from the STIS Echelle (E140M)  
149: with 10 orbits (PID 8125, 28.5 ksec) and compared the absorber velocities to 
150: those in a small group of galaxies containing four large H~I galaxies 
151: and one dwarf seen with the VLA (see Fig.\ 3 in Shull et al.\ 1998). 
152: The PKS~2155-304 sightline passes near the center of this group,
153: none of these galaxies lies closer than 400 kpc to the sightline, and no 
154: other dwarf galaxies were seen down to $m_B \approx 19$.  The galaxy $1D$ 
155: velocity dispersion is $\sigma_{\rm gal} = 325$ km~s$^{-1}$, and the Ly$\alpha$ 
156: velocity dispersion is $\sigma_{\rm Lya} = 740$ km~s$^{-1}$.  We estimate the
157: galaxy overdensity as $\delta \sim 100$, based on the group of five galaxies.  
158: 
159: We also observed PKS~2155-304 with FUSE (Moos et al.\ 2000)
160: in three visits on 23-24 October 1999 (P1080701, $t_{\rm exp} = 19.2$ ksec, 
161: and P1080705, $t_{\rm exp} = 38.6$ ksec) and 18 June 2001 (P1080703, 
162: $t_{\rm exp} = 65.4$ ksec). All data were calibrated with CALFUSE 2.2.1. 
163: We base most of our analysis on the second visit, which obtained data for 
164: all of the FUSE detectors, whereas the earlier visits occurred before the 
165: SiC and LiF channels (Sahnow et al.\ 2000) were aligned. For the LiF1a 
166: segment (1000--1084 \AA), we roughly tripled the amount of available data 
167: by co-adding night-only reductions of the three observations. This method 
168: recovered the redshifted Ly$\gamma$ absorption (1027.6 and 1028.0~\AA) from 
169: below O~I airglow emission (1025.76 and 1026.47 \AA). We applied small wavelength 
170: offsets ($\leq 0.05$ \AA) to the FUSE data to align narrow interstellar lines 
171: and place the FUSE and STIS data on a common velocity scale set by Galactic 
172: 21-cm emission. 
173: 
174: The claimed X-ray detection of O~VIII absorption (Fang et al.\ 2002) 
175: with $4.5 \sigma$ significance was made with the {\it Chandra} Low-Energy
176: Transmission Grating Spectrometer. However, the redshifted O~VIII absorber
177: is not seen near 20~\AA\ in the spectrum taken by the {\it X-Ray Multi-Mirror
178: Observatory} (XMM) (see Fig. 1 and Table 1 in Rasmussen, Kahn, \& Paerels 2003).
179: Although the XMM velocity resolution is poorer than that of {\it Chandra}/LETG,
180: the XMM data have high S/N and detect oxygen absorption at $z = 0$ with
181: equivalent widths $16.3 \pm 3.3$ m\AA\ (O~VIII, 5.7 $\sigma$ significance) and
182: $9.0 \pm 2.7$ m\AA\ (O~VII, 8 $\sigma$ significance).  The non-confirmation of 
183: the $z = 0.0567$ O~VIII absorption remains a puzzle.
184:  
185: Figure 1 shows a velocity overlay of the Ly$\alpha$, Ly$\gamma$, O~VI, and
186: O~VIII absorbers between $cz = 15,000 - 20,000$ km~s$^{-1}$.
187: The detected lines and their properties are summarized in Table 1; 
188: the three strong Ly$\alpha$ absorbers are labeled A, B, and C for
189: clarity. We detect Ly$\beta$ at 16,243 km s$^{-1}$ (A), but Ly$\beta$ lines 
190: at 16,973 km~s$^{-1}$ (B) and 17,110 km s$^{-1}$ (C) are severely blended 
191: with the N~II $\lambda$1084 interstellar line. We detect Ly$\gamma$ in 
192: components B and C, but we derive a large error on their equivalent width
193: due to continuum placement uncertainty. We also detect O~VI in
194: components A and C, at 16,243 and 17,144 km~s$^{-1}$. 
195: 
196: 
197: 
198: \section{INTERPRETATION OF THE SPECTRA}
199: 
200: With our new HST/STIS E140M data, we can derive more accurate column densities 
201: and metallicities for the strong Ly$\alpha$ absorbers.
202: In our previous study (Shull et al.\ 1998) at 20 km~s$^{-1}$ 
203: (HST/GHRS) resolution, we estimated N$_{\rm HI} = (3-10) \times 10^{14}$ cm$^{-2}$ 
204: for the blended B and C components at $z \approx 0.0567$.  There, we set a 
205: metallicity limit $Z < 0.003 Z_{\odot}$ from the absence of Si~III 
206: and C~IV absorption. At the higher S/N and better spectral resolution of the
207: STIS/E140M, together with FUSE measurements of Ly$\gamma$, the derived H~I columns 
208: decreased considerably. The integrated Ly$\alpha$ optical depths give 
209: N(H~I) = 1.0, 2.2, and $0.94 \times 10^{14}$ cm$^{-2}$, respectively,
210: for absorbers at 16,185, 16,974, and 17,147 km~s$^{-1}$, 
211: consistent with the observed Ly$\gamma$ strengths.
212: 
213: We still did not detect any Si~III $\lambda 1206$ or C~IV $\lambda 1548$ 
214: absorption at the Ly$\alpha$ velocities (Table 1). The 6--12~m\AA\ equivalent 
215: width limits correspond to N(C~IV) $< 2.9 \times 10^{12}$ cm$^{-2}$ and
216: N(Si~III) $< 2.7 \times 10^{11}$ cm$^{-2}$.  
217: As described in Shull et al. (1998), we assume that the absorption arises in a 
218: uniform slab of gas photoionized by the metagalactic radiation background,   
219: with spectral shape for a background dominated by AGN with intrinsic 
220: spectral index of $\alpha = 1.8$ modified by IGM absorption (Shull et al.\ 1999).  
221: The ionization correction (conversion from C~IV/H~I to [C/H]) depends only on 
222: the ionization parameter, $\log U \propto I_0 / n_H$, where $I_0$ is the 
223: specific intensity at 1 ryd and $n_H$ is the volume density of the gas.  
224: For $-2.0 < \log U < -0.5$, the metallicity upper limit from C~IV/H~I ranges 
225: between 0.025--0.10 of the solar abundance in the CLOUDY photoionization code, 
226: (C/H)$_{\odot} = 3.55 \times 10^{-4}$ (Grevesse \& Noels 1993). 
227: A full discussion of the parameter dependence of the metallicity 
228: estimate appears in a subsequent paper.
229: 
230: The two detected O~VI systems have column densities of N(O~VI) = $(2.8 \pm 0.5) 
231: \times 10^{13}$ cm$^{-2}$ (A) and $(3.3 \pm 0.5) \times 10^{13}$ cm$^{-2}$ (C)
232: assuming a linear curve of growth.  At the temperature (log $T_{\rm OVI} = 5.45$)
233: of maximum ionization abundance ($b_{\rm OVI} = 17$ km~s$^{-1}$), the stronger
234: \ion{O}{6} $\lambda 1032$ lines in both components would still be unsaturated.
235: We identify these two O~VI absorbers with 
236: Ly$\alpha$ systems at 16,185 (A) and 17,147 km~s$^{-1}$ (C).  No O~VI is seen 
237: in the 16,974 km~s$^{-1}$ (B) absorber.  The O~VI lines are the 
238: only metals yet detected in these absorbers, but their metallicity
239: is uncertain, owing to an indeterminate ionization state and thermal phase 
240: (see Savage et al.\ 2002).  If the O~VI arises in photoionized gas, its column 
241: density and limits on C~IV imply $\log U > -0.8$, for C/O in a solar abundance 
242: ratio of $0.50 \pm 0.07$ (Allende Prieto, Lambert, \& Asplund 2002).
243: 
244: If the ionizing background has $I_0 \approx 10^{-23}$ 
245: ergs~cm$^{-2}$ s$^{-1}$ Hz$^{-1}$ sr$^{-1}$ (Shull et al.\ 1999), the 
246: density constraint is $n_H < 10^{-5.6}$ cm$^{-3}$, and the corresponding 
247: line-of-sight extent of the gas is $D > 1.1$~Mpc, comparable to the
248: extent of the galaxy group.  We believe that this low-density, large-absorber
249: model is highly unlikely.  Multiple absorbers of this extent could not fit 
250: within the 1 Mpc volume of the group, without undergoing collisions and
251: shredding.  However, if the O~VI arises in collisionally ionized gas, the 
252: requirement that the gas be so rarefied no longer holds.  
253: In addition, the fraction of O~VI in collisional 
254: ionization equilibrium peaks at $T \sim 10^{5.5}$ K, and it is tempting to 
255: associate the O~VI with the WHIM.  The tentative detection of O~VIII absorption 
256: makes collisionally ionized O~VI still more compelling.
257: The claimed strength of the X-ray feature (Fang et al.\ 2002) requires 
258: log N(O~VIII) $\approx 16.0 \pm 0.2$, which could represent hot gas at the 
259: group barycenter. However, redshifted O~VIII absorption was not confirmed by 
260: XMM, nor has O~VII absorption at $z = 0.0567$ been detected 
261: (Nicastro et al.\ 2002).
262: 
263: 
264: \section{DISCUSSION}
265: 
266: In order to understand the cluster of PKS~2155-304 absorbers at 
267: $z \approx 0.0567$, one must resort to a multiphase model. The H~I, O~VI, 
268: and O~VIII systems are unlikely to co-exist, owing 
269: to kinematic offsets and physical considerations.  For example, the
270: claimed O~VIII absorption centroid ($16,624 \pm 237$ km~s$^{-1}$) 
271: appears to differ from the observed H~I and O~VI absorbers (A, B, C 
272: components).  Also, the observed Ly$\gamma$ line widths 
273: (FWHM $\leq 30$ km~s$^{-1}$) rule out H~I temperatures greater than 
274: 20,000~K.   The measured O~VI line widths ($40 \pm 10$ km~s$^{-1}$ FWHM) 
275: make it unlikely, although marginally possible within errors, for O~VI 
276: and O~VIII to exist at the same temperature, $\log T \approx 6.25 \pm 0.1$,
277: where (O~VI/O~VIII) $\approx 6 \times 10^{-3}$ in collisional
278: ionization equilibrium (Sutherland \& Dopita 1993).  However, we 
279: discount this scenario, because of the observed velocity offsets
280: ($\sim 400$ km~s$^{-1}$).  For similar reasons, the strong H~I absorbers 
281: cannot co-exist with the hot O~VIII.  At the high temperatures 
282: needed to ionize O~VIII collisionally, the H~I ionization fraction
283: would be too small to be detected.
284: 
285: Therefore, we favor a more complex scenario in which the H~I and O~VI
286: absorbers arise in clumps of gas falling into a small-group potential. 
287: Any hot gas, visible in O~VII or O~VIII, would then exist at the barycenter
288: of the group (16,600 km~s$^{-1}$) which would be consistent
289: with the kinematic offsets.  If this model is correct, the two O~VI
290: absorbers (A and C) would arise from clumps
291: undergoing ``backside" and ``nearside" infall, respectively.  Because
292: production of collisionally ionized O~VI requires shock velocities 
293: $V_s \geq 130$ km~s$^{-1}$ (Shull \& McKee 1979; Dopita \& Sutherland 1996),
294: the infall model has consequences for which Ly$\alpha$ absorbers
295: contain associated O~VI.  In particular, the absence of detectable
296: O~VI in the strongest Ly$\alpha$ system (B) may be
297: a result of a low relative velocity between the infalling H~I clump
298: and the substrate.  From Figure 1, one sees that O~VI is
299: present in the H~I absorber at 17,147 km~s$^{-1}$, located 
300: approximately 170 km~s$^{-1}$ redward.  Thus, one might speculate
301: that these two blended Ly$\alpha$ absorbers, one with and one without 
302: O~VI, represent shocked and unshocked gas.
303: 
304: In the clump-infall model, the relative velocity between the two O~VI
305: components and the group barycenter is $\sim 400$ km~s$^{-1}$, which
306: would produce a post-shock temperature of $\sim 2 \times 10^6$~K,   
307: similar to that ($2.5 \times 10^6$~K)
308: at which the observed O~VI and claimed O~VIII would co-exist in
309: collisional ionization equilibrium.  The cooling time of such gas is
310: \begin{equation}
311:    t_{\rm cool} = \frac {(20~{\rm Gyr}) T_{6.5}} {n_{-4} \Lambda_{-23}} \; , 
312: \end{equation}
313: where $n_{-4} = n_H/(10^{-4}~{\rm cm}^{-3})$ and $T_{6.5} = (T / 10^{6.5}$~K). 
314: Here, $\Lambda_{-23}$ is the radiative cooling rate coefficient in units of 
315: $10^{-23}$ erg~cm$^3$~s$^{-1}$, typical of gas at $10^{6.5}$~K 
316: with 0.1 solar metallicity 
317: (Sutherland \& Dopita 1993). In the range $5.0 < \log T < 7.0$, 
318: $\Lambda(T) \propto T^{-1}$, so $t_{\rm cool}$ is longer 
319: than the age of the universe, unless the gas temperature can
320: be decreased ($T_{6.5} \ll 1$) or the density increased ($n_{-4} \gg 1$).   
321: Through shock-compression and cooling, the cooling will 
322: accelerate as $\Lambda(T)$ moves toward its $10^{5.3}$~K peak. 
323: 
324: We can relate the hydrogen density to the cosmological overdensity 
325: $\delta$ by: 
326: \begin{equation}
327:     n_H(z) = (1.90 \times 10^{-7}~{\rm cm}^{-3})(1+z)^3
328:        \left[ \frac {\Omega_b h_{70}^2}{0.046} \right] (1 + \delta)  \; . 
329: \end{equation}
330: At the redshift ($z_a = 0.0567$) of the strong absorbers, we have 
331: $n_{-4} \approx (\delta / 500)$.  If the infalling gas is shocked 
332: to $T_s = (1.24 \times 10^6~{\rm K})(V_s/300~{\rm km~s}^{-1})^2$
333: to form the WHIM phase, it must acquire a fairly high density
334: ($\delta > 500$ at $z \approx 0.1$) in order to cool, recombine, and form 
335: the observed O~VI. Adiabatic shocks can provide a compressive trigger, but
336: further radiative cooling is required, since adiabatic compression
337: lengthens the cooling time ($t_{\rm cool} \propto n^{1/3}$ for
338: $T \propto n^{2/3}$ and $\Lambda(T) \propto T^{-1}$).   Lower-velocity shocks, 
339: $V_s \approx 200$ km~s$^{-1}$, would cool more rapidly, with temperatures 
340: near the peak of the cooling curve.  
341: 
342: Two additional issues concern the ratios of C~IV, O~VI, and H~I.  
343: In the two detected O~VI absorbers, (C~IV/O~VI) $< 10^{-1.1}$, 
344: whereas models of radiatively cooling gas (Dopita \& Sutherland 
345: 1996; Indebetouw \& Shull 2003) find (C~IV/O~VI) $\approx 0.1-0.2$.  
346: {\it Where is the C~IV?}  Is the cooled, recombined layer truncated in the 
347: hot-gas environment?  Similarly, we do not understand the wide variation in 
348: [N(H~I)/N(O~VI)] $\approx$ 0.1--10 among other IGM absorbers (Shull 2003). 
349: The production of shocked O~VI requires 
350: {\it relative} velocities above $\sim 150$ km~s$^{-1}$ and O/H abundances
351: above a few percent solar. However, the production mechanism of O~VI remains 
352: unclear (Sembach et al. 2003). While shock-heating is the preferred mechanism, 
353: other processes have been proposed involving interfaces between hot substrates 
354: and moving clouds -- thermal conduction, shear instabilities, and turbulent
355: mixing (see Indebetouw \& Shull 2003 for a review).  These
356: processes all have problems explaining the observed components,
357: N(O~VI) $\approx 3 \times 10^{13}$ cm$^{-2}$.   Each conductive 
358: interface typically produces N(O~VI) $\sim 10^{13}$ cm$^{-2}$, while
359: shock heating requires supersonic flow and depends on the Mach number cubed.  
360: If $T_{\rm hot} \approx 10^{6.4}$~K (adiabatic sound speed 
361: $c_s \approx 240$ km/s), infalling clumps at 300-400 km~s$^{-1}$ have Mach 
362: numbers of just 1.25 -- 1.67.  
363: 
364: Resolving the O~VI heating paradox may require understanding the cloud-substrate 
365: interfaces at a deeper level.  Alternatively, if the O~VIII absorption
366: turns out to be non-existent, a virialized hot substrate may not have formed
367: yet. Intergroup gas at $T < 10^6$~K would produce sufficient shock-heating,
368: and the observed O~VI absorption may arise from the interactions of infalling
369: clumps with metals in a intergroup medium injected from the
370: galaxies.  With combined UV and X-ray spectra, one may be able to distinguish among  
371: these processes through the amount of shock heating and the relative abundances 
372: of O~VI, O~VII, and O~VIII.  
373: 
374: 
375: \acknowledgements
376: 
377: This work is based on data obtained for the Guaranteed Time Team by the 
378: NASA-CNES-CSA FUSE mission operated by the Johns Hopkins University. 
379: Financial support to U.S. participants has been provided by
380: NASA contract NAS5-32985. We were also supported by
381: grants GO-08571.01-A from the Space Telescope Science Institute and
382: NAG5-7262 from NASA/LTSA.  We thank Taotao Fang for providing the 
383: {\it Chandra} spectrum that appears in Figure 1 and John Stocke 
384: and Bill Blair for helpful discussions.
385: 
386: 
387: 
388: 
389: 
390: \newpage
391:  
392: \begin{references}
393: 
394: 
395: \reference{allende} Allende Prieto, C., Lambert, D. L., \& Asplund, M.
396:     2002, \apj, 573, L137
397: 
398: \reference{burles} Burles, S., \& Tytler, D. 1998, ApJ, 495, 80 
399: 
400: \reference{cen99} Cen, R., \& Ostriker, J. P. 1999, \apj, 519, L109
401: 
402: \reference{dave} Dav\'e, R., et al. 2001, \apj, 552, 473
403: 
404: \reference{dopita} Dopita, M. A., \& Sutherland, R. S. 1996, \apjs, 102, 161 
405: 
406: \reference{fang} Fang, T., Marshall, H. L., Lee, J. C., Davis, D. S.,
407:    \& Canizares, C. R. 2002, \apj, 572. L127
408: 
409: \reference{Freed} Freedman, W. L., et al. 2001, ApJ, 553, 47 
410: 
411: \reference{fuku} Fukugita, M., Hogan, C. J., \& Peebles, P. J. E. 1998, 
412:    \apj, 503, 518 
413: 
414: \reference{gnedin} Gnedin, N. Y. 1998, \mnras, 294, 407
415: 
416: \reference{grevesse} Grevesse, N., \& Noels, A. 1993, in Origin of the
417:    Elements, ed. N. Prantzos, E. Vangioni-Flam, \& M. Cass\'e,
418:    (Cambridge:  Cambridge Univ. Press), 15
419: 
420: \reference{indeb} Indebetouw, R., \& Shull, J. M. 2003, \apj, submitted
421: 
422: \reference{Martin} Martin, C. L., Kobulnicky, H. A., \& Heckman, T. M.
423:    2002, ApJ, 574, 663
424: 
425: \reference{moos} Moos, H. W., et al. 2000, ApJ, 538, L1
426: 
427: \reference{Nett} Netterfield, C.B., et al. 2002, ApJ, 571, 604
428: 
429: \reference{nicastro} Nicastro, F., et al. 2002, \apj, 573, 157
430: 
431: \reference{penton} Penton, S. V., Stocke, J. T., \& Shull, J. M. 2003,
432:    \apj, submitted 
433: 
434: \reference{rasmussen} Rasmussen, A., Kahn, S. M., \& Paerels, F. 2003,
435:     in The IGM/Galaxy Connection, (Dordrecht:  Kluwer Publ.), ed. 
436:     J. Rosenberg \& M. Putman, 109.  
437: 
438: \reference{sahnow}  Sahnow, D. J., et al. 2000, \apj, 538, L7 
439: 
440: \reference{savage02} Savage, B. D., Sembach, K. R., Tripp. T. M., \& 
441:     Richter, P. 2002, \apj, 564, 631
442: 
443: \reference{sembach} Sembach, K. R., et al. 2003, \apjs, 146, 165 
444: 
445: \reference{shull79} Shull, J. M., \& McKee, C. F. 1979, \apj, 227, 131
446: 
447: \reference{shull98} Shull, J. M., Penton, S. V., Stocke, J. T., Giroux, M. L.,
448:    van Gorkom, J. H., Lee, Y. H., \& Carilli, C. 1998, \aj, 116, 2094
449: 
450: \reference{shull99} Shull, J. M., Roberts, D., Giroux, M. L., Penton, S. V.,
451:      \& Fardal, M. A. 1999, AJ, 118, 1450
452: 
453: \reference{shull03} Shull, J. M., 2003, in The IGM/Galaxy Connection,
454:      (Dordrecht:  Kluwer Publ.), ed. J. Rosenberg \& M. Putman, 1
455:  
456: \reference{spergel} Spergel, D.N. et al. 2003, ApJ, in press (astro-ph/0302209)
457: 
458: \reference{sutherland} Sutherland, R. S., \& Dopita, M. A. 1993, \apjs, 88, 253
459: 
460: \reference{tripp00} Tripp, T. M., Savage, B. D., \& Jenkins, E. B. 2000,
461:    \apj, 534, L1
462: 
463: 
464: \end{references}
465: 
466: 
467: 
468: 
469: \newpage
470: 
471: \begin{deluxetable}{lccc}
472: \tablecolumns{4}
473: \tablewidth{0pc}
474: \tablecaption{PKS~2155-304 IGM Absorption Lines}
475: \tablehead{
476: \colhead{Line} & \colhead{Velocity}     & \colhead{$W_{\lambda}$}
477:               &  \colhead{$\Delta V$\tablenotemark{a} }  \\
478: \colhead{}    & \colhead{(km~s$^{-1}$)} & \colhead{(m\AA)} &
479:                 \colhead{(km~s$^{-1}$)} }
480: \startdata
481:   H I Ly$\alpha$ (A)   & $16185\pm 5$    & $331 \pm 12$      & $75 \pm 8$  \\
482:   H I Ly$\alpha$ (B)   & $16974\pm 5$    & $495 \pm 10$      & $76 \pm 5$  \\
483:   H I Ly$\alpha$ (C)   & $17147\pm 5$    & $353 \pm 8$       & $110 \pm 10$  \\
484:   H I Ly$\beta$  (A)   & $16243\pm 10$   & 85 $\pm$ 10       & $92 \pm 17$  \\
485:   H I Ly$\gamma$ (B)   & $16936\pm 10$   & 16 $\pm$ 4        & $<30$        \\
486:   H I Ly$\gamma$ (C)   & $17109\pm 10$   & 13 $\pm$ 3        & $<30$       \\
487:   H I Ly$\delta$\tablenotemark{b} & \nodata  & $<$12         & $<30$       \\
488:   O VI 1032 (A)        & $16243\pm 10$   & 37 $\pm$ 7        & $35 \pm 10$  \\
489:   O VI 1038 (A)        & $16252\pm 10$   & 25 $\pm$ 5        & $30 \pm 10$   \\
490:   O VI 1032 (C)        & $17144\pm 10$   & 44 $\pm$ 6        & $40 \pm 10$   \\
491:   O VI 1038 (C)        & $17116\pm 10$   & Fe II blend       &  ...          \\
492:   O~VIII 18.969        & $16624\pm 237$  & $14.0^{+7.3}_{-5.6}$  &  $<1380$   \\
493:   C III  977\tablenotemark{b}  & \nodata &  $<10$    &  \\
494:   C IV 1548\tablenotemark{b}   & \nodata &  $<12$    &  \\
495:   Si III 1206\tablenotemark{b} & \nodata &  $<6$     &  \\
496: \enddata
497: \tablenotetext{a}{Full width at half maximum (FWHM) from STIS/FUSE 
498:    profile fitting, or by converting {\it Chandra} line width of
499:    $\sigma({\rm O~VIII)} < 0.039$~\AA\ (Fang et al.\ 2002).  The doppler 
500:    parameter, $b_{\rm dopp} = {\rm FWHM}/[2 (\ln 2)^{1/2} ]$. }
501: 
502: \tablenotetext{b}{Upper limit at 4$\sigma$ significance.}
503: \end{deluxetable}
504: 
505: 
506: 
507: \newpage
508: 
509: \begin{figure*}
510:  \centerline{\epsfxsize=0.55\hsize{\epsfbox{f1.eps}}}
511:  \figcaption{Absorption lines of H~I (Ly$\alpha$ from HST/STIS, Ly$\gamma$
512:   from FUSE), O~VI 1032/1038 (FUSE), and claimed detection of O~VIII
513:   with {\it Chandra} (Fang et al.~2002) toward PKS~2155-304.  The three
514:   strongest Ly$\alpha$ absorbers are labeled (A,B,C -- see Table 1). We
515:   detect O~VI in the 16,243 km~s$^{-1}$ (A) and 17,144 km~s$^{-1}$ (C)
516:   absorbers, and Ly$\gamma$ in B and C absorbers; claimed O~VIII absorption
517:   is offset by $\sim400$ km~s$^{-1}$ from Ly$\alpha$ and O~VI
518:   (vertical dotted lines). Interstellar lines (Ly$\beta$, C~II, O~VI)
519:   are labeled in panel 3.  Lower curve in panel 4 shows O~VI~1038
520:   offset from O~VI 1032.  The Fe~II line shows an asymmetric profile
521:   that suggests presence of O~VI~1038 at 17,140 km~s$^{-1}$, but blending
522:   precludes an accurate measurement.}
523: \label{stackfig}
524: \end{figure*}
525: \vspace{0.1in}
526: 
527: 
528: 
529: \end{document}
530: 
531: