1: %\documentclass[12pt,preprint]{aastex}
2: \documentclass[12pt]{emulateapj}
3:
4: \usepackage{graphicx}
5:
6: \newcommand{\hi}{{\sc Hi} }
7: \newcommand{\hip}{{\sc Hi}}
8: \newcommand{\cca}{cm$^{-2}$ }
9: \newcommand{\ccap}{cm$^{-2}$}
10: \newcommand{\AaA}{A\&A}
11: \newcommand{\ApJS}{ApJS}
12: \newcommand{\AJ}{AJ}
13: \newcommand{\MNRAS}{MNRAS}
14: \newcommand{\ApJ}{ApJ}
15: \newcommand{\Natur}{Nature}
16:
17: \begin{document}
18:
19: \title{First detection of dust emission in a High-Velocity Cloud}
20:
21: \author{M.-A. Miville-Desch\^enes\altaffilmark{1} and F. Boulanger}
22: \affil{Institut d'Astrophysique Spatiale, Universit\'e Paris-XI, 91405, Orsay Cedex, France}
23: \and
24:
25: \author{W. T. Reach \& A. Noriega-Crespo}
26: \affil{Spitzer Science Center, California Institute of Technology, MS 100-22, Pasadena, CA 91125, USA}
27:
28: \altaffiltext{1}{Senior Research Associate, Canadian Institute for Theoretical Astrophysics, University of Toronto,
29: 60 St-George st, Toronto, Canada}
30:
31:
32: \begin{abstract}
33: By comparing sensitive Spitzer Space Telescope (SST) infrared and Green Bank Telescope 21 cm observations,
34: we are able to report the first detection of dust emission in Complex C, the largest
35: High Velocity Cloud in the sky.
36: Dust in the region of Complex C studied here has a colder temperature
37: ($T=10.7^{+0.9}_{-0.8}$~K - 1$\sigma$)
38: than the local interstellar medium ($T=17.5$~K), in accordance with its great distance from the Galactic plane.
39: Based on the metallicity measurements and assuming diffuse Galactic interstellar medium dust properties
40: and dust/metals ratio, this detection could imply gas column densities more than 5 times higher
41: than observed in \hip. We suggest that the dust emission detected here comes from small molecular
42: clumps, spatially correlated with the \hi but with a low surface filling factor.
43: Our findings imply that the HVCs' mass would be much larger than inferred from \hi observations and that
44: most of the gas falling on the Milky Way would be in cold and dense clumps rather than in a diffuse phase.
45: \end{abstract}
46:
47: \keywords{Galaxy: Halo, ISM: Clouds, ISM: Dust}
48:
49: \section{Introduction}
50: Since their discovery in HI observations \citep{muller63}, High Velocity Clouds (HVCs)
51: have been the target of numerous studies (see \citet{van_woerden2004} for a thorough review)
52: but nevertheless remain puzzling. It is now widely considered that many of them
53: might be infalling clouds fueling the Galaxy with low metallicity gas.
54: This hypothesis received strong support from ultraviolet observations showing that HVCs have a subsolar
55: metallicity \citep{wakker99} and a D/H ratio compatible with primordial abundances of deuterium \citep{sembach2004}.
56:
57: The present study focuses on the Spitzer Extra-Galactic First Look Survey (XFLS) field,
58: a diffuse HI area ($N_{HI} \sim 2\times10^{20}$ cm$^{-2}$) at high Galactic latitude ($l=88^\circ.3$, $b=34^\circ.9$)
59: programmed to study galaxy populations and clustering \citep{fang2004}.
60: The XFLS field is located on the edge of Complex C, a high-velocity gas structure spanning more than
61: 1500 square degrees on the sky. Complex C has a subsolar metallicity ($\sim 0.1-0.3$) \citep{tripp2003},
62: it contains ionised gas observed in $H\alpha$ emission \citep{tufte98a}
63: and O VI absorption \citep{sembach2003} and is located at a distance greater than 5 kpc
64: from the sun \citep{wakker2001}.
65: Dust emission has been unsuccesfully looked for in HVCs using IRAS data \citep{wakker86}.
66: Upper limits were found to be compatible with their low \hi column densities,
67: low metallicity (and therefore low dust/gas ratio) and their large distance (i.e. fainter radiation field).
68: In this study we revisit the dust/gas correlation in HVCs using new infrared and \hi data.
69:
70: \section{Data}
71:
72: \label{section_data}
73:
74: In this study we used Spitzer Space Telescope (SST) MIPS 24 and 160~$\mu$m data of the XFLS.
75: The field was covered by MIPS in 8 scan maps that were combined and destriped,
76: then regridded to 16$''$ pixels and median-smoothed.
77: We complemented the SST data with improved IRAS 60 and 100~$\mu$m maps (IRIS - \citet{miville-deschenes2005}).
78: The FWHM angular resolution of the IRIS maps is $\sim 4'$ with a pixel size of 1'.5.
79: The calibration uncertainties on diffuse emission of the SST and IRIS maps is 10~\%.
80: Bright point sources were removed from all infrared maps (SST and IRIS)
81: before convolution by a Gaussian beam to bring all of them to the Green Bank Telescope (GBT) 9'.2 FWHM resolution.
82: Fig.~\ref{fig_infraredmaps} shows the convolved maps.
83:
84:
85: The \hi (21 cm) data we used in this study were obtained with the GBT,
86: previously published and kindly made available by \citet{lockman2005}.
87: These observations cover a $3^\circ\times 3^\circ$ area centered at (J2000)
88: $\alpha=17^h18^m$, $\delta=+59^\circ30'$, which encloses the SST observations.
89: The GBT data were corrected for stray-radiation contamination.
90: The sampling was done at 1'5, the FWHM beam is 9.2', the channel width is $dv=$0.51 km s$^{-1}$
91: and the noise level in a single channel is 0.05 K.
92:
93: \begin{figure}
94: \includegraphics[width=\linewidth, draft=false]{f1.eps}
95: \caption{\label{fig_infraredmaps} Infrared maps of the XFLS field at 24 and 160~micron from Spitzer
96: and 60 and 100 micron from IRIS. All maps were convolved
97: by a Gaussian beam to bring them to the GBT 9.2' FWHM resolution.}
98: \end{figure}
99:
100: \begin{figure}
101: \includegraphics[width=\linewidth, draft=false]{f2.ps}
102: \caption{\label{fig_spectreHI} Decomposition of the 21 cm data. The bottom curve shows the
103: 21 cm spectrum averaged over the whole field. The middle and top curves (shifted for clarity) are representative
104: 21 cm spectra where the IVC1 (top) and IVC2 (middle) components are observed. The shaded rectangles
105: indicate the velocity range used to extract the \hi components.}
106: \end{figure}
107:
108: \section{HI decomposition}
109:
110: Fig.~\ref{fig_spectreHI} presents the average 21 cm spectrum of the field and two typical lines-of-sight.
111: In addition to Complex C, the XFLS 21 cm data revealed two Intermediate-Velocity Clouds (IVC1 and IVC2)
112: and a local component typical of high-latitude \hi emission with a narrow cold component superimposed on
113: a broad warm component \citep{lockman2005}.
114: In this section we estimate the column density of each of these \hi components.
115: The HVC is well isolated in velocity space and its integrated emission was computed
116: by adding all the channels where $-210 < v < -132$ km s$^{-1}$.
117: On the other hand the two IVCs and the local component overlap in velocity.
118: The integrated emission of the two IVC components were computed by
119: adding all the channels between $-40 < v < -14$ km s$^{-1}$ for IVC1
120: and $-50 < v < -29$ km s$^{-1}$ for IVC2 from which we subtracted the local component
121: contribution estimated by interpolating a baseline on each spectrum (see Fig.~\ref{fig_spectreHI}).
122: The local component's integrated emission was computed by adding all channels between
123: $-48 < v < +25$ km s$^{-1}$
124: from which the two IVC contributions were removed. There is still contribution
125: from the local components in a few channels at $v < -48$~km s$^{-1}$
126: but they were not added, as the emission is very diffuse and its structure within the noise.
127:
128: In the optically-thin case
129: (i.e. when the 21 cm brightness temperature is much smaller than the \hi kinetic temperature),
130: the 21 cm integrated emission is proportional to the \hi column density:
131: $N_{HI}(x,y) = 1.823 \times 10^{18} \sum_v T_B(x,y,v) dv$.
132: The local component is dominated by a broad (i.e. warm and optically-thin) component but a fraction of the emission comes
133: from cold gas (narrow features) that needs to be corrected for opacity to properly estimate the column density.
134: To do so we estimated the fraction of warm and cold gas on each line-of-sight
135: using a Gaussian decomposition of each spectrum. For each Gaussian component
136: we estimated its opacity correction, considering that
137: thermal and turbulence broadening equally contribute to the velocity dispersion $\sigma_v$
138: (i.e. spin temperature is $T_s = 121 \sigma_v^2 /2$), instead of the usual isothermal
139: assumption (\citet{lockman2005} used $T_s=80$~K). This allowed us to compute a larger correction for
140: the narrow than for the broad components.
141: Overall the opacity correction for the local component is lower than 6~\%.
142: It reaches its maximum value in the bright filament (south-western part of the field,
143: see Fig.~\ref{fig_himaps}) where our estimate of the \hi kinetic temperature gets as low as 50~K.
144: The opacity correction for the two IVCs and the HVC is less than 1\% in accordance with their
145: broad lines ( FWHM$>8$~km~s$^{-1}$).
146: Fig.~\ref{fig_himaps} presents the \hi column density maps of the four \hi components.
147:
148:
149: \begin{figure}
150: \includegraphics[width=\linewidth, draft=false]{f3.eps}
151: \caption{\label{fig_himaps} \hi column density of the four \hi components (Local, HVC, IVC1 and IVC2).
152: All maps are in units of $10^{20}$~cm$^{-2}$. An opacity correction was applied to convert the
153: 21 cm integrated emission to \hi column density - see text for details.}
154: \end{figure}
155:
156:
157: \section{Infrared-HI correlation}
158:
159: The main goal of this study was to separate and determine the infrared emission of the
160: HI components identified with the 21~cm observations.
161: To do so we considered that the infrared map $I_\lambda(x,y)$ at wavelength $\lambda$
162: can be represented by the following model
163: \begin{equation}
164: \label{eq_regress}
165: I_\lambda(x,y) = \sum_i \alpha^{i}_\lambda N_{HI}^{i}(x,y) + C_\lambda(x,y)
166: \end{equation}
167: where $N_{HI}^i(x,y)$ is the column density of the $i^{th}$ HI component,
168: $\alpha^i_\lambda$ is the infrared-\hi correlation coefficient of component $i$
169: at wavelength $\lambda$ and $C_\lambda(x,y)$ is a residual term.
170:
171: To estimate the correlation coefficients $\alpha_\lambda^i$ we performed a chi-square minimisation
172: on the \hi and infrared data based on the model given by Eq.~\ref{eq_regress}.
173: The infrared-\hi correlation coefficients
174: are collected in Table~\ref{table_ir} and shown in Fig.~\ref{fig_sed}.
175: The standard deviation of the residual maps $C_\lambda(x,y)$ are also given
176: in Table~\ref{table_ir}.
177:
178: The uncertainties on $\alpha_\lambda^i$ given in Table~\ref{table_ir} include calibration and
179: statistical uncertainties, taking into account the data sampling.
180: To estimate the impact of our opacity correction we made the same analysis
181: with optically-thin \hi maps and obtained the same results within the uncertainties.
182:
183: The infrared-\hi correlation presented here shows that the model used (Eq.~\ref{eq_regress})
184: for the infrared emission is satisfying.
185: At each wavelength the standard deviation of the residual term $C_\lambda(x,y)$ is
186: in good concordance with the expected \citep{lagache2003} and measured \citep{miville-deschenes2002}
187: levels of the Cosmic Infrared Background (CIB) fluctuations (see Table~\ref{table_ir}),
188: except at 160~$\mu$m where we report a significant residual excess. As the residues at 60 and 100~$\mu$m
189: are in accordance with the CIB level, this analysis also leads us to conclude that dust
190: associated with the Warm Ionised Medium (WIM) \citep{lagache99}
191: does not significantly bias our decomposition of the infrared emission in this field.
192:
193: The striking result of this analysis is certainly the significant detection
194: of the infrared emission associated with an HVC at all wavelengths. Our result at 100~$\mu$m is
195: compatible with the upper limit given by \citet{lockman2005}.
196:
197: \section{Dust properties of Complex C}
198:
199: Using the infrared-\hi correlation coefficients we now want to put contrains on the properties of the dust,
200: especially in the HVC.
201:
202: We assume that the emission from big grains at thermal equilibrium with the radiation field is a modified black body :
203: \begin{equation}
204: I_\lambda = \tau_\lambda B_\nu(T_{BG})
205: \end{equation}
206: where $B_\nu$ is the Planck function, $T_{BG}$ is the big grains' equilibrium temperature and $\tau_\lambda$ the opacity,
207: often expressed as $\tau_\lambda = N_{HI}\epsilon_\lambda$ where $\epsilon_\lambda$ is the dust emissivity per H atom.
208: The emissivity is generally described as a power law : $\epsilon_\lambda = \epsilon_0 \lambda^{-\beta}$
209: where $\beta$ is the emissivity spectral index.
210: Based on FIRAS data, \citet{boulanger96} give the following expression for the big grains' emissivity law in
211: the local interstellar medium:
212: \begin{equation}
213: \epsilon_\lambda = 1.0 \times 10^{-25}(\lambda/250 \, \mu m)^{-2}\;\; \mbox{cm$^2$}.
214: \end{equation}
215:
216: In this scheme, and at wavelengths where the emission is dominated by big grains, the infrared-\hi correlation
217: coefficients can be expressed as $\alpha_\lambda = \epsilon_\lambda B_\nu(T_{BG})$.
218: Assuming that Very Small Grains (VSGs) do not contribute significantly in the far-infrared and
219: taking $\beta=2$ \citep{draine84,boulanger96}, we use
220: the 100 and 160~$\mu$m infrared-\hi correlation coefficients
221: to estimate the big grains' temperature $T_{BG}^i$ of each \hi component
222: \begin{equation}
223: \frac{\alpha_{100}^i}{\alpha_{160}^i} = \frac{ B_\nu(100\mu m, T_{BG}^i) 160^2}{ B_\nu(160\mu m, T_{BG}^i) 100^2}
224: \end{equation}
225: and the dust emissivities at 160~$\mu$m : $\epsilon_{160}^i = \alpha_{160}^i/B_\nu(160 \mu m, T_{BG}^i)$.
226: The big grain temperature and the 160~$\mu$m emissivity of each \hi component are given
227: in Table~\ref{table_dust}. In this table are also given the corresponding 1 and 3$\sigma$ ranges
228: on $T_{BG}^i$ and $\epsilon^i_{160}$ computed using 1 and 3$\sigma$
229: uncertainties on $\alpha_{100}^i$ and $\alpha_{160}^i$.
230:
231:
232: \begin{figure}
233: \includegraphics[width=\linewidth, draft=false]{f4.ps}
234: \caption{\label{fig_sed} \hi-infrared correlation coefficients for the Local, HVC, IVC1 and IVC2
235: components shown in Fig.~\ref{fig_himaps}. See Table~\ref{table_ir} for details.}
236: \end{figure}
237:
238: \begin{table*}
239: \begin{center}
240: \caption{\label{table_ir} Infrared-\hi correlation}
241: \begin{tabular}{ccccccc} \hline
242: $\lambda$ & $\alpha_\lambda^{Local}$& $\alpha_\lambda^{HVC}$ & $\alpha_\lambda^{IVC1}$ & $\alpha_\lambda^{IVC2}$ & $\sigma$(Residue) & CIB estimate\\
243: ($\mu$m) & & & & & (MJy sr$^{-1}$) & (MJy sr$^{-1}$)\\\hline
244: 24 & 0.017$\pm$0.002 & 0.007$\pm$0.002 & 0.039$\pm$0.006 & 0.014$\pm$0.005 & 0.004 & 0.004\\
245: 60 & 0.16$\pm$0.02 & 0.05$\pm$0.01 & 0.35$\pm$0.04 & 0.31$\pm$0.04 & 0.03 & 0.02\\
246: 100 & 0.80$\pm$0.08 & 0.055$\pm$0.015 & 0.70$\pm$0.09 & 0.9$\pm$0.1 & 0.09 & 0.07 \\
247: 160 & 1.7$\pm$0.2 & 0.8$\pm$0.1 & 2.7$\pm$0.4 & 1.4$\pm$0.4 & 0.30 & 0.11 \\\hline
248: \end{tabular}
249: \end{center}
250: Infrared-\hi correlation coefficients ($\alpha_\lambda^i$) given in MJy sr$^{-1}$ ($10^{20}$ H atoms)$^{-1}$ cm$^2$ for the four
251: \hi components (Local, HVC, IVC1 and IVC2) at 24, 60, 100 and 160~$\mu$m. The uncertainties on each
252: $e_\lambda^i$ is the 1$\sigma$ uncertainty taking into account the statistical variance
253: and instrumental uncertainties.
254: Column 6 gives the standard deviation of the residual map and Column 7 is an estimate of the Cosmic Infrared Background (CIB)
255: fluctuations based on the model of \citet{lagache2003}.
256: \end{table*}
257:
258: \begin{table}
259: \begin{center}
260: \caption{\label{table_dust} Dust parameters}
261: \begin{tabular}{lcccccc}\hline
262: & $T_{BG}$ & $1\sigma$ & $3\sigma$ & $\epsilon_{160}$ & $1\sigma$ & $3\sigma$\\ \hline
263: Local & 17.4 & 16.3-18.8 & 14.3-22.6 & 3.0 & 1.9-4.9 & 0.60-12.5 \\
264: HVC & 10.7 & 9.9-11.6 & 7.7-13.6 & 36.0 & 16.8-83.7 & 3.8-1442.8 \\
265: IVC1 & 14.6 & 13.6-15.8 & 11.8-19.4 & 13.2 & 7.0-24.1 & 1.6-79.6 \\
266: IVC2 & 19.4 & 17.1-23.1 & 14.1-107.4 & 1.5 & 0.5-3.6 & 0.003-15.5\\ \hline
267: \end{tabular}
268: \end{center}
269: Big Grain temperature ($T_{BG}$, given in Kelvin) and dust emissivity at 160~$\mu$m
270: ($\epsilon_{160} = \tau_{160}/N_{HI}$ given in $10^{-25}$ cm$^2$ per H atom) and their
271: corresponding 1 and 3~$\sigma$ ranges for each \hi component.
272: The values for the local interstellar medium deduced from the COBE data \citep{boulanger96} are
273: $T_{BG}=17.5$~K and $\epsilon_{160} = 2.4\times10^{-25}$~cm$^2$.
274: \end{table}
275:
276: \section{Discussion}
277:
278: The values of $T_{BG}^{local}=17.4^{+1.4}_{-1.1}$ and
279: $\epsilon_{160}^{local}=3.0^{+1.9}_{-1.1}\times 10^{-25}$~cm$^2$ per H atom found for the
280: local component are in very good agreement (within 1$\sigma$
281: uncertainties) with the average values found for the Solar neighborhood
282: by \citet{boulanger96} using COBE data : $T_{BG}^{soln}=17.5$~K and
283: $\epsilon_{160}^{soln}=2.4 \times 10^{-25}$ cm$^2$ per H atom.
284: This is an important validation of the photometric accuracy of the SST-MIPS data.
285: On the other hand dust in the HVC is found to be significantly colder than in the local interstellar medium,
286: even at the 3$\sigma$ limit ($T_{BG}^{HVC} = 10.7_{-3.0}^{+2.9}$),
287: which is consistant with a lower radiation field than in the Solar neighborhood
288: due to its distance to the Galaxy \citep{wakker86}.
289: In addition we report an emissivity $\epsilon_{160}$ higher in the HVC than in the local ISM
290: (see Table~\ref{table_dust}).
291: An eventual contribution of VSGs to the far-infrared emission
292: would have the effect of lowering $T_{BG}$ and increasing $\epsilon_{160}$ even more.
293:
294: For dust properties typical of the local interstellar medium,
295: $\epsilon_\lambda$ should be proportional to the dust-to-gas mass ratio which, for standard metal
296: depletion on grains, would scale with the metallicity.
297: The fact that $\epsilon^{HVC}_{160}/\epsilon^{soln}_{160} > 1.6$ (3$\sigma$) is greater than the metallicity
298: in Complex C ($ Z_{HVC}/Z_\odot = 0.2\pm0.1$ according to \citet{tripp2003}) could indicate a dust
299: (and therefore gas) column density in Complex C more than 5 times larger than given by the \hi emission.
300:
301: Could it be dust associated with ionised gas ?
302: H$\alpha$ and other ionised atomic lines have been detected in several HVCs. It is expected
303: to come from a photoionized layer at the surface of the clouds which plunge into the Galactic halo.
304: The column density of such a layer is $2-4 \times 10^{19}$ cm$^{-2}$ in Complex C \citep{wakker99}
305: which is similar to the \hi column density in the HVC seen in XFLS and therefore not enough
306: to explain the high infrared emissivity measured here. In addition only the ionised gas spatially
307: correlated with the \hi would contribute to the infrared emissivity.
308: Could it be dust associated with diffuse H$_2$ ?
309: Two detections of H$_2$ in absorption with FUSE have been reported but the column densities are much lower
310: than those of \hi \citep{richter2001}. Other attemps have been unsuccesful;
311: CO was not found in emission \citep{wakker97} nor HCO$^+$ in absorption \citep{combes2000}.
312: Furthermore abundance measurements in HVCs show no sign of iron depletion \citep{murphy2000,tripp2003}
313: which has been used as a strong argument in favor of very low dust in HVCs.
314:
315: To reconcile our findings with previous work we suggest that HVCs could be multiphase gas
316: including small and dense molecular clumps with a low surface filling factor (and therefore
317: undetected with pencil beam absorption observations) bathing in a diffuse phase seen in absorption.
318: The dust emission reported here would mainly originate in these clumps, spatially correlated with \hip,
319: which would have a larger iron depletion than in the diffuse phase of HVCs.
320: Evidence of the existence of such small molecular clumps was found in cirrus
321: clouds \citep{heithausen2004} and in the Magellanic Stream \citep{sembach2001}.
322: Additional observations of HVCs with the Spitzer Space Telescope are needed to confirm our result.
323:
324: If our result reveals a general property of HVCs, it would mean that their mass
325: is much larger than what can be inferred from \hi observations and that most
326: of the gas is in a dense phase. An infall of 1 M$_\odot$/year is needed to explain
327: the star formation rate, the distribution of stellar metallicities with age
328: and the presence of a bar in the Milky Way and in other galaxies.
329: The current idea is that this infalling gas is diffuse (ionised and neutral)
330: resulting from the cooling of very hot halo gas.
331: Our result changes this picture by showing that most of the infalling gas is in dense and cold clumps,
332: leaving open the question of the origin and stability of such structures.
333:
334: HVCs typically contribute to $<10\%$ of the \hi column density at high latitude, but
335: with its temperature lower than in the local interstellar medium,
336: dust in HVCs could contribute a much larger fraction to the sub-millimeter and millimeter emission.
337: Observations in this wavelength range, with Herschel and Planck for instance,
338: could be used as a tracer of the column density of HVCs and as a distance indicator,
339: using the dust temperature to put constraints on the radiation field strength.
340: If the detection presented in this study is confirmed and is typical of HVCs,
341: it could also have a strong impact on the analysis of observational cosmology data
342: (e.g. WMAP, Planck) that do not to this day take this foreground component into account.
343:
344:
345: The authors thank the anonymous referee for a careful reading of the manuscript
346: and very helpful comments. M.A.M.D. acknowledges support from The Canadian Space Agency.
347: This research has made use of NASA's Astrophysics Data System Bibliographic Services.
348:
349:
350: \begin{thebibliography}{{Miville-Deschenes} \& {Lagache}(2005)}
351:
352: \bibitem[{Boulanger} et~al.(1996)]{boulanger96}
353: {Boulanger}, F., et al.\
354: \newblock 1996, \AaA, 312, 256.
355:
356: \bibitem[{Combes} \& {Charmandaris}(2000)]{combes2000}
357: {Combes}, F. \& {Charmandaris}, V.
358: \newblock 2000, \AaA, 357, 75.
359:
360: \bibitem[{Draine} \& {Lee}(1984)]{draine84}
361: {Draine}, B.~T. \& {Lee}, H.~M.
362: \newblock 1984, ApJ, 285, 89
363:
364: \bibitem[{Fang} et~al.(2004)]{fang2004}
365: {Fang}, F. et~al.
366: \newblock 2004, \ApJS, 154, 35.
367:
368: \bibitem[Heithausen(2004)]{heithausen2004} Heithausen, A.\ 2004,
369: \apjl, 606, L13
370:
371: \bibitem[{Lagache} et~al.(1999)]{lagache99}
372: {Lagache}, G., {Abergel}, A., {Boulanger}, F., {Desert}, F.~X., {Puget}, J.~L.
373: \newblock 1999, \AaA, 344, 322
374:
375: \bibitem[{Lagache} et~al.(2003)]{lagache2003}
376: {Lagache}, G., {Dole}, H. \& {Puget}, J.~L.
377: \newblock 2003, \MNRAS, 338, 555.
378:
379: \bibitem[{Lockman} \& {Condon}(2005)]{lockman2005}
380: {Lockman}, F.~J. \& {Condon}, J.~J.
381: \newblock 2005, \aj, 129, 1968
382:
383: \bibitem[{Miville-Deschenes} \& {Lagache}(2005)]{miville-deschenes2005}
384: {Miville-Deschenes}, M.-A. \& {Lagache}, G.
385: \newblock 2005, \ApJS, 157, 302
386:
387: \bibitem[{Miville-Desch\^enes} et~al.(2002)]{miville-deschenes2002}
388: {Miville-Desch{\^e}nes}, M.~A., {Lagache}, G., \& {Puget}, J.-L..
389: \newblock 2002, A\&A, 393, 749.
390:
391: \bibitem[{Muller} et~al.(1963)]{muller63}
392: {Muller}, C.~A., {Oort}, J.~H. \& {Raimond}, E.
393: \newblock 1963, Comptes rendus de l'Acad\'emie des Sciences de Paris, 257, 1661.
394:
395: \bibitem[Murphy et al.(2000)]{murphy2000} Murphy, E.~M., et al.\
396: 2000, \apjl, 538, L35
397:
398: \bibitem[{Richter} et~al.(2001a)]{richter2001}
399: {Richter}, P., {Sembach}, K.~R., {Wakker}, B.~P. \& {Savage}, B.~D.
400: \newblock 2001a, \ApJ, 562, L181.
401:
402: \bibitem[{Richter} et~al.(2001b)]{richter2001a}
403: {Richter}, P., et al.\
404: \newblock 2001b, \ApJ, 559, 318.
405:
406: \bibitem[Sembach et al.(2001)]{sembach2001} Sembach, K.~R., Howk,
407: J.~C., Savage, B.~D., \& Shull, J.~M.\ 2001, \aj, 121, 992
408:
409: \bibitem[{Sembach} et~al.(2003)]{sembach2003}
410: {Sembach}, K.~R. et~al.
411: \newblock 2003, \ApJS, 146, 165.
412:
413: \bibitem[{Sembach} et~al.(2004)]{sembach2004}
414: {Sembach}, K.~R. et~al.
415: \newblock 2004, \ApJS, 150, 387.
416:
417: \bibitem[{Tripp} et~al.(2003)]{tripp2003}
418: {Tripp}, T.~M. et~al.
419: \newblock 2003, \AJ, 125, 3122.
420:
421: \bibitem[{Tufte} et~al.(1998)]{tufte98a}
422: {Tufte}, S.~L., {Reynolds}, R.~J. \& {Haffner}, L.~M.
423: \newblock 1998, \ApJ, 504, 773.
424:
425: \bibitem[{van Woerden} et~al.(2004)]{van_woerden2004}
426: {van Woerden}, H., {Wakker}, B.~P., {Schwarz}, U.~J. \& {De Boer}, K.~S.
427: \newblock 2004, {\em High Velocity Clouds}.
428: \newblock Kluwer Academic Publishers, Dordrecht.
429:
430: \bibitem[{Wakker} \& {Boulanger}(1986)]{wakker86}
431: {Wakker}, B.~P. \& {Boulanger}, F.
432: \newblock 1986, \AaA, 170, 84.
433:
434: \bibitem[{Wakker} et~al.(1997)]{wakker97}
435: {Wakker}, B.~P., {Murphy}, E.~M., {van Woerden}, H. \& {Dame}, T.~M.
436: \newblock 1997, \ApJ, 488, 216.
437:
438: \bibitem[{Wakker} et~al.(1999)]{wakker99}
439: {Wakker}, B.~P. et~al.
440: \newblock 1999, \Natur, 402, 388.
441:
442: \bibitem[{Wakker}(2001)]{wakker2001}
443: {Wakker}, B.~P.
444: \newblock 2001, \ApJS, 136, 463.
445:
446: \end{thebibliography}
447:
448: \end{document}
449: