1: \documentclass[12pt,preprint]{aastex}
2: \begin{document}
3: \title{A New {\it FUSE} Survey of Interstellar HD}
4:
5: \author{Theodore P. Snow\altaffilmark{1}, Teresa L. Ross\altaffilmark{1}, Joshua D. Destree\altaffilmark{1}, Meredith M. Drosback\altaffilmark{1}, Adam G. Jensen\altaffilmark{2}, Brian L. Rachford\altaffilmark{3}, Paule Sonnentrucker\altaffilmark{4}, and Roger Ferlet\altaffilmark{5}}
6:
7: \altaffiltext{1}{Center for Astrophysics and Space Astronomy, Department of Astrophysical and Planetary Sciences, University of Colorado at Boulder, Campus Box 389, Boulder, CO 80309-0391, USA; tsnow@casa.colorado.edu; teresa.ross@colorado.edu; destree@casa.colorado.edu; meredith.drosback@colorado.edu}
8: \altaffiltext{2}{NASA Postdoctoral Program Fellow, Goddard Space Flight Center, Code 665, NASA/GSFC, Greenbelt, MD 20771, USA; Adam.G.Jensen@nasa.gov}
9: \altaffiltext{3}{Department of Physics, Embry-Riddle Aeronautical University, 3700 Willow Creek Road, Prescott, AZ 86301, USA; rachf7ac@erau.edu}
10: \altaffiltext{4}{Department of Physics and Astronomy, Johns Hopkins University, 3400 North Charles Street, Baltimore, MD 21218, USA; sonnentr@pha.jhu.edu}
11: \altaffiltext{5}{Institut d'Astrophysique de Paris, UMR7095 CNRS, Universite Pierre \& Marie Curie, 98bis Boulevard Arago, 75014 Paris, France; ferlet@iap.fr}
12:
13:
14:
15:
16: \begin{abstract}
17: We have used archival {\it FUSE} data to complete a survey of interstellar
18: HD in 41 lines of sight with a wide range of extinctions. This follow up
19: to an earlier survey was made to further assess the utility of HD as a
20: cosmological probe; to analyze the HD formation process; and to see what
21: trends with other interstellar properties were present in the data. We
22: employed the curve-of-growth method, supported by line profile fitting, to
23: derive accurate column densities of HD. We find that the N(HD)/2N(H$_2$)
24: ratio is substantially lower than the atomic D/H ratio and conclude that
25: the molecular ratio has no bearing on cosmology, because local processes
26: are responsible for the formation of HD. Based on correlations with
27: E(B-V), H$_2$, CO, and iron depletion, we find that HD is formed in the
28: densest portion of the clouds; the slope of the logN(HD)/log(H$_2$)
29: correlation is greater than 1.0, caused by the destruction rate of HD
30: declining more slowly than that of H$_2$; and, as a sidelight, that the
31: depletions are density dependent.
32:
33: \end{abstract}
34:
35: \keywords{ISM: abundances}
36:
37: \section{Introduction}
38:
39: The H$_2$ isotopologue HD was first detected in {\it Copernicus} spectra by
40: Spitzer et al. (1973, 1974) and Morton (1975), and has subsequently been
41: observed in many sightlines by {\it FUSE}. The HD lines arising from the
42: lowest-lying rotational levels ($J$=0 and $J$=1) are far weaker than their
43: counterparts for H$_2$, but are in many cases still strong enough to be
44: saturated, requiring a curve-of-growth analysis for column density
45: determinations. To date HD analyses from {\it FUSE} spectra have been
46: published only for a few stars \cite{Ferlet,Lacour} though many more
47: detections reside in the {\it FUSE} archives. In addition, Lacour et
48: al. have re-analyzed {\it Copernicus} spectra for several stars, resulting
49: in a uniform survey of HD abundances in some 17 sightlines. The {\it
50: Copernicus} stars generally have A$_V$ values of 1.0 or less, while our
51: {\it FUSE} targets cover a range of A$_V$ from 0.1 to about 3 magnitudes.
52:
53: The ratio of HD to H$_2$ --- more specifically, N(HD)/2N(H$_2$) --- in the
54: Lacour et al. survey ranges from a few times 10$^{-7}$ to several times
55: 10$^{-6}$. This is somewhat higher than the values found earlier from {\it
56: Copernicus} data, but those values were based on two strong lines that were
57: almost always saturated.
58:
59: We follow the convention of Lacour et al. (2005), and assert that, in a
60: region where all hydrogen (including deuterium) is in molecular form, the
61: atomic D/H ratio is the same as the N(HD)/2N(H$_2$). The reasoning is that
62: the molecular fraction (N(HD)/N(H$_{total}$) = N(HD)/[N(H I) + 2N(H$_2$)])
63: reduces to N(HD)/2N(H$_2$) when H I can be neglected. For every HD
64: molecule, there is one D atom, so the atomic ratio D/H is equal to
65: N(HD)/2N(H$_2$). Note that this assumes that not only H I, but also atomic D,
66: can be neglected. This simplifying assumption is justified in cloud cores,
67: but it may break down in regions where hydrogen (and deuterium) is only
68: partially in molecular form, which may include some of the less-reddened
69: lines of sight in our survey. But even there, as shown in the next few
70: paragraphs, we can not use HD as an indicator of the D/H ratio anyway.
71:
72: It would be useful if the N(HD)/2N(H$_2$) ratio were a reliable indicator
73: of the atomic ratio D/H, because the latter is an important probe of the
74: early expansion rate of the universe and therefore valuable in cosmology.
75: An excellent summary of the observed atomic D/H ratio and its cosmological
76: implications can be found in Linsky et al. (2006). But for three
77: distinct reasons the comparison of HD to H$_2$ is difficult to apply to the
78: cosmological problem.
79:
80: First, H$_2$ begins to be self-shielding much before HD, as a function of
81: depth in a cloud. For example, H$_2$ becomes self-shielding at about
82: N(H$_2$) = 10$^{19}$ cm$^{-2}$ (which corresponds to N(HD) = 10$^{12-13}$
83: cm$^{-2}$) and a similar column density would be required for HD -- which is
84: incredibly unlikely. For HD to become self-shielding, the column density of
85: H$_2$ would have to be of order 10$^{25}$ cm$^{-2}$! Even infrared
86: observations of the quadrupole transitions at 2.4 $\mu$m would never reach
87: this column density. The fact that HD is not protected from radiative
88: dissociation acts to lower the N(HD)/2N(H$_2$) with respect to the atomic D/H
89: ratio in the denser portions of the lines of sight.
90:
91: Second, if formation of HD occurs on grain surfaces (in parallel with H$_2$
92: formation), the lower mobility of D atoms on grains as compared to H
93: atoms, greatly reduces the formation rate of HD as compared to H$_2$,
94: acting to reduce N(HD)/2N(H$_2$) ratio relative to the D/H ratio. If grain
95: formation of HD dominates, this would act to decrease the N(HD)/2N(H$_2$)
96: ratio over the atomic ratio. For a more complex and complete analysis of HD
97: formation on grain surfaces, see Cazaux et al. (2008).
98:
99: But third, HD has a gas-phase formation channel, through the "chemical
100: fractionation" reaction H$_2$ + D$^{+} \to$ HD + H$^{+}$ \cite{Watson,LePetit},
101: which tends to enhance the N(HD)/2N(H$_2$) ratio and depends on the
102: cosmic-ray ionization rate \cite{Black}. The comparison of the
103: N(HD)/2N(H$_2$) ratio to the atomic D/H ratio can tell us which processes
104: are most important.
105:
106: In any case, the N(HD)/2N(H$_2$) ratio should not be expected to reflect
107: the D/H ratio (Lacour et al. 2005). In general the N(HD)/2N(H$_2$) ratio
108: is not useful for determining the cosmological D/H ratio, although it may
109: be useful in constraining the rate of cosmic-ray ionization (which produces
110: D$^+$).
111:
112: This paper, then, presents a new and more complete survey of HD than was
113: available previously, to test formation theories and possible correlations
114: with other interstellar quantities. We have organized the paper into
115: sections as follows: Observation and Data Reduction, Data Analysis,
116: Determination of HD Column Densities, Correlations, Summary and Conclusions.
117:
118: \section{Observation and Data Reduction}
119:
120: The {\it FUSE} mission covered a
121: wavelength range (905-1188 \AA) rich in ground state transitions, making it
122: one of the most useful tools to study the interstellar medium. The 41
123: sightlines for this survey are archival {\it FUSE} spectra,
124: chosen to span different cloud types and physical conditions. The
125: sightline parameters used to select targets included color excess (0.11
126: $\le$ $E(B-V)$ $\le$ 0.83), the indicator of grain size R$_v$ (2.25 $\le$
127: $R_v$ $\le$ 4.76), molecular fraction (0.02 $\le$ f$_{H2}$ $\le$ 0.76),
128: and H$_2$ column density (19.07 $\le$ log N(H$_2$) $\le$ 21.11). Table
129: \ref{table1} lists all of the relevant physical properties for the
130: sightlines included in this study. Another selection criterion was a high
131: signal-to-noise ratio around the 1105.83 \AA\ HD line. This line is 100
132: times weaker than the rest of the HD absorption features measured, all of
133: which have similar oscillator strengths (HD f-values and wavelength data
134: from Abgrall \& Roueff 2006). In many cases this weak line's
135: equivalent width is the only one falling on the linear portion of the curve
136: of growth, thus providing unambiguous information about the column density
137: as Lacour et al. found in their study.
138:
139: Because HD is known to be correlated with H$_2$ abundances \cite{Lacour},
140: our target list includes sightlines with H$_2$ column densities in the
141: literature. H$_2$ column densities are taken from Shull et al. (in
142: preparation, 2008) and Rachford et al. (2002; in preparation, 2008). For
143: this study the {\it FUSE } data (Table \ref{table2}) were pre-calibrated
144: with version 3.1.4 or newer of the CALFUSE pipeline. For each observation
145: we used a cross correlation analysis to align individual exposures with
146: strong absorption features before co-adding them.
147:
148: In most cases we co-added the detector segments to increase the signal to
149: noise ratio before measuring the HD equivalent widths. When combining the
150: detector segments one significant systematic flaw was found in the
151: data. The raw data from the LiF2A detector segment revealed the same
152: systematic detector defect due to a Type I dead zone (as described in
153: section 9.1.6 of The {\it FUSE} Instrument and Data Handbook) in all
154: sightlines in the vicinity of the weak 1105.83 \AA\ HD line; thus we
155: excluded that segment and only used LiF1B.
156:
157:
158: \section{Data Analysis}
159:
160: To determine the column density of HD we measured equivalent widths for all
161: possible HD lines (see Table \ref{eq_width}) and performed a curve of
162: growth (COG) analysis. All of the HD lines measured are transitions from
163: the ground rotational state ($J$=0). While there are over 25 HD
164: absorption lines in the {\it FUSE} range, only seven are isolated enough
165: from other features to determine accurate equivalent widths. To measure
166: the equivalent width we first defined a continuum on both sides of the
167: feature and normalized it with low order (1 to 3) Legendre polynomials.
168: When there was no other absorption feature in the immediate vicinity we
169: integrated over the HD line while summing the flux errors provided by {\it
170: FUSE} in quadrature. If there was another absorption feature blended with
171: the HD line, we fit the HD and interfering line with Gaussians (see next
172: section for details). The errors in the width and depth of the Gaussian
173: were propagated into the equivalent width error. The statistical error on
174: the equivalent width was summed in quadrature with the continuum placement
175: error. The continuum error was up to an order of magnitude smaller than the
176: statistical error in all cases.
177:
178: Of the seven lines, six have similar oscillator strengths (see Table
179: \ref{absorption_lines}); the seventh and weakest line is located at 1105.86
180: \AA\ and is very important for constraining the COG fit. The weak HD
181: line is difficult to measure in most cases and in other cases it is
182: impossible to measure at all. Without a line to measure we used the
183: signal-to-noise ratio to set a two sigma upper limit on the equivalent
184: width using the equation:
185: \begin{equation}
186: \sigma = \frac { \Delta \lambda M^{1/2}} {(S/N)}
187: \end{equation}
188: where {$\Delta \lambda$} is the pixel scale, $S/N$ is the signal-to-noise
189: ratio, $M$ is the width of the feature (here {\it FUSE} resolution
190: elements ($\sim$9 pixels) are used since the width of the line will be
191: smeared out by the resolution), and $\sigma$ gives the one sigma error of the
192: equivalent width \cite{Jenkins}.
193:
194:
195:
196: \subsection{Modeling the C I* Line}
197: The first obstacle in measuring the weak HD line is a very crowded patch of
198: continuum. In a study of the sightline to $\zeta$ Oph, Morton (1978) found
199: an unidentified absorption feature at 1105.92 \AA. In the current study
200: we have been unable to confirm or deny the existence of a weak feature near
201: this wavelength; thus, if such a feature does exist it could be a source of
202: systematic error. Morton also rejected a feature at 1105.82 as being the
203: weak line of HD due to its large equivalent width. The conclusion of Lacour
204: et al. (2005) and the present study is that the feature at 1105.83 is
205: indeed HD and its strength is consistent with all of the other measured HD
206: features (see Figures \ref{plottwo}, \ref{plotthree}, \ref{plotfive} \&
207: \ref{plotsix} containing HD COG's). Shortward of the HD line, by 0.13 \AA,
208: is an absorption feature from the first excited state of carbon as
209: discussed by Lacour et al. In sightlines where the C I* and HD lines were
210: blended we modeled the C I* and divided it from the spectrum before
211: measuring the HD line.
212:
213: Removing this blended C I* line turned out to be more challenging than
214: originally thought. The f-value of the 1105.73 C I* line,
215: previously measured as 0.0113 by Morton (1978), did not fit the curve of
216: growth set by the other C I* lines. This is probably due to uncertainty in
217: the adopted f-value of this weak line. The previous value was based on a
218: single measurement of the line in the spectrum of $\zeta$ Oph. So to
219: accurately model the C I* line we used existing archival data to make an
220: empirical estimate of the f-value, based on more sightlines.
221:
222: To accomplish this task we found {\it Copernicus}, {\it FUSE}, and/or
223: STIS data in sightlines where the 1105 \AA\ C I* line was well
224: defined, not
225: heavily blended with other lines in the region (HD 12323, $o$ Per, HD
226: 207538, HD 210839) and measured the equivalent widths of all well defined C
227: I* lines, excluding the 1105 \AA\ line. For each sightline we then fit
228: these equivalent widths to a single component curve of growth varying both
229: $b$-value and column density (see Figure \ref{plotone} for an example).
230: Once a best fit curve of growth was found, the f-value of the 1105 \AA\
231: line was calculated using its measured equivalent width. Uncertainty in
232: the f-value was assigned based on errors in the measured equivalent width
233: of the 1105 \AA\ line. Calculated f-values and uncertainties for each
234: sightline can be found in Table \ref{tab_CIfvalue}. The results from all
235: sightlines were combined using a weighted average and the final f-value was
236: found to be 0.0062$^{+0.0015}_{-0.0010}$.
237:
238: The C I* lines used in modeling are in Table \ref{absorption_lines}. We
239: used STIS data when available, along with {\it FUSE} data to measure
240: the C I* absorption features. Using the best fit column density and
241: $b$-value of C I*, we generated a Voigt profile of the 1105.73 \AA\ C I*
242: line. This profile was then convolved with the Gaussian instrumental
243: profile (assuming a resolution of 20 km s$^{-1}$) and divided out of the
244: spectrum, allowing us to measure the HD 1105.83 \AA\ line.
245:
246: We began the current study using C I* line f-values from Morton's
247: compilation (1991). However, in light of newer theoretical f-values from
248: Zatsarinny \& Froese Fischer (2002) and Froese Fischer (2006), it was
249: important to determine whether differences due to revised f-values would
250: substantially affect our HD column densities. Repeating the calculation of
251: the C I* 1105.73 \AA\ line f-value yielded only a very small change (about
252: three percent) from the previously calculated f-value. All C I* column
253: densities were also recalculated using the newer f-values. Total column
254: densities did not vary greatly from those using the Morton (1991) f-values.
255: In all but three cases differences in C I* column density using the two
256: sets of f-values differed by 0.1 dex or less. The three most discrepant
257: cases were HD 73882, HD 101436 and HD 149404. For these three cases the
258: equivalent width of the weak HD line was remeasured and found to be well
259: within the 1$\sigma$ errors of the previous measurements. Similarly, in
260: all three cases rerunning the COG fit yielded only small changes in the HD
261: column density that were within the 1$\sigma$ error bars of the previous
262: measurements (in all cases the difference was less than 0.1 dex). Thus, we
263: did not think it necessary to repeat all of the C I* modeling, and
264: remeasure the weak HD line, as other errors far exceed any error introduced
265: by minor changes in some of the C I* f-values, and changes in C I* column
266: densities were not substantial in the vast majority of cases. All reported C
267: I* column densities are those using the newer f-values from Zatsarinny \&
268: Froese Fischer (2002) and Froese Fischer (2006).
269:
270:
271: \section{Determination of HD Column Densities}
272: Once the influence of C I* had been removed, we were able to do a
273: curve-of-growth analysis to determine the column density, but we used
274: profile fitting in a few cases as a check. We did two different COG
275: analyses; multiple component COG's for a sub-set of 13, and single
276: component COG's for all 41 sightlines. Both are described below in
277: subsection 4.1 'Curve of Growth Analysis'. The profile fitting is
278: described in subsection 4.2.
279:
280:
281: \subsection{Curve of Growth Analysis}
282: In many cases multiple cloud components exist along a single line of sight
283: increasing the difficulty in accurately measuring the column density.
284: Having knowledge of such structure is most important when the
285: equivalent widths fall on the flat portion of the curve. We can create a
286: multiple component COG, but first we need to find a suitable tracer of HD.
287:
288: HD has been found to be correlated to H$_2$ \cite{Lacour}. Similarly CH is
289: traced by H$_2$ because of the chemistry: the creation of CH in the diffuse
290: ISM is directed through a series of gas phase reactions with H$_2$
291: (Danks, Federman \& Lambert 1984; van Dishoeck \& Black 1989; Magnani et
292: al. 1998). Since CH and HD are both correlated to H$_2$ abundances, there
293: is an inferred correlation between CH and HD, which has been verified by
294: Lacour et al.
295:
296: For thirteen sightlines we had high resolution CH data (Welty, private
297: communication) which we used to define a velocity structure for HD. From
298: the measured CH line-of-sight velocity structure, we generated multiple
299: component COG's. The HD equivalent widths were fit to the curve varying
300: only the total column density. For comparison the data were also fit to a
301: single component COG that found a best fit of both $b$-value and column
302: density by minimizing the ${\chi}^2$. Figures \ref{plottwo} and
303: \ref{plotthree} show the side by side comparison of the COG analyses, and
304: Figure \ref{plotfour} compares the column densities and errors from the two
305: methods.
306:
307: The COG's from the two methods fit the data in all but three cases. The
308: side by side comparison show a clear mis-fit of the multiple component COG
309: for HD 149404, HD 185418, HD 192639, even though the column densities are
310: within one-sigma errors for the first two targets. The discrepancy arose
311: because a multiple component COG assumes the best fit velocity structure of
312: CH is the actual structure of the HD and does not allow for errors. Varying
313: the $b$-values on the multiple component COG within their one sigma errors
314: gave solutions that were consistent with the data and the single component COG.
315:
316: With the reasonable assurance that our single component COG model is good
317: we accounted for any other systematic errors that we could not quantify by
318: setting the reduced $\chi ^2$ of the best-fit curve equal to one. In 22
319: sightlines the reduced $\chi ^2$ was already less than one, thus the errors
320: were unchanged; for the remaining 19 cases the assumed good fit scaled up
321: the errors in equivalent width and column density.
322:
323: The consistency between the results of the single and multiple component
324: COG demonstrates that a single component model is a good approximation when
325: we do not have component information. Thus, we are justified in combining
326: the HD results for the 13 targets with component information and the 28
327: targets without; see Figures \ref{plotfive} and \ref{plotsix} for single
328: component COG's and Table \ref{colden} for column densities.
329:
330: Solving for column density in the COG analysis was more complicated when we
331: only had an upper limit on the 1105 \AA\ line. We used the limit on the
332: weak line to constrain the column density by throwing out solutions beyond
333: the upper limit. The limit was not used in calculating the $\chi ^2$ of
334: the fit.
335:
336: Lower limits on the $b$-value were determined when all the points fell on
337: the linear portion of the COG. The $b$-value controls where the curve will
338: transition from the linear to the saturated portion. Since the transition
339: cannot occur lower than the data points a two sigma lower limit is set (HD
340: 91824, HD 93204, HD 93205, HD 94493, HD 161807, and HD 201345.)
341:
342:
343: \subsection{Profile Fitting}
344:
345: Profile fitting of the HD lines was done for five lines of sight to verify
346: the COG column density measurements. All the available HD
347: lines were fit simultaneously so that a single best fit column density and
348: $b$-value would be measured. First the continuum surrounding each HD feature
349: was normalized by fitting a low order Legendre polynomial to the continuum
350: on both sides. Shifts in the central velocity of the HD features were
351: noticed from one line to another, (with each detector segment showing
352: different velocity shifts, usually on the order of $\sim$10 km s$^{-1}$) so
353: the lines were each fit with a single Gaussian and the central wavelength
354: corrected so that each feature would be at its rest wavelength. Profiles
355: were modeled assuming a single Voigt profile with column density $N$, line
356: width $b$ and central velocity $v$. These profiles were then convolved with
357: a Gaussian instrumental profile. We used the non-linear least-squares curve
358: fitting algorithm MPFIT by C. Markwardt\footnote{\tt
359: http://cow.physics.wisc.edu/$\sim$craigm/idl/idl.html} to find the best fit
360: parameter values. MPFIT is a set of routines that uses the
361: Levenberg-Marquardt technique to minimize the square of deviations between
362: data and a user-defined model. These routines are based upon the MINPACK-1
363: Fortran package by More' and collaborators\footnote{\tt
364: http://www.netlib.org/minpack}.
365:
366: Attempts were also made to fit multiple components when previous
367: measurements of CH existed. However, even when the velocity structure was
368: fixed by the CH structure, individual component column densities did not
369: converge to reasonable answers. Usually one component of the fit dominated
370: while the other component(s) were an order of magnitude or two lower in
371: column density (differing greatly from the relative strengths of the CH
372: components). Because of this, we report only the single component fits in
373: Table \ref{profile_vsCOG}.
374:
375: For all five sightlines we find that the column densities and $b$-values
376: for both the curve of growth and profile-fitting method are within standard
377: errors of one another (including HD 192639, our previously discrepant
378: case). We also find that the best fit profiles match the absorption
379: profiles in the data well (see Figure \ref{plotseven}). Thus, we are
380: confident that our column densities measured through curve of growth
381: fitting are reliable.
382:
383:
384: \subsection{Comparisons with previous results}
385:
386: We are able to compare our column densities with the previous HD
387: survey of Lacour et al. (2005). This survey had 7 stars observed
388: with {\it FUSE} and 10 more lines of sight observed years before, with
389: {\it Copernicus}. The comparison of our results are in Table
390: \ref{French_compare}.
391:
392: We re-analyzed the 7 {\it FUSE} targets that Lacour measured to check
393: whether our method was consistent with their results. Our method differs
394: in a few small ways. We re-derived the C I* f-value (as described above),
395: so our de-blending of the weak HD line is different. Also we weighted each
396: line equally in the COG fit, while the Lacour group doubly-weighted the
397: optically thin HD line. Even with these difference our column densities
398: are consistent within the uncertainties.
399:
400: We also did single-component curves of growth for every sightline,
401: which the Lacour group did not, and we found consistent column densities
402: with the multicomponent results. This allowed us to expand our survey with
403: confidence, in the end obtaining accurate column densities for 41 stars.
404:
405: Previous HD measurements made by Spitzer et al. (1974) are included in Figure
406: \ref{ploteight} and the correlation analysis. The plot includes the upper
407: limits for $\delta$ Per and $\lambda$ Ori and the measurements for $\pi$
408: Sco, $\delta$ Sco, $\sigma$ Sco, and $\gamma$ Ara.
409:
410:
411: \section{Correlations}
412: One way to derive information about the formation mechanism of HD, and to
413: test theoretical chemical models, is to see how various interstellar
414: quantities correlate with HD column densities. Before we do that, a few
415: words of caution are in order. Of necessity, we can measure only integrated
416: line-of-sight quantities, not localized to the same portion of the lines of
417: sight. For example, any correlations of molecular quantities versus dust
418: indicators or H I are probably not informative, because both dust and H I
419: are everywhere, inhabiting every region from diffuse to dark clouds. We
420: assume, with good reason, that the observable HD and H$_2$ (along with CH) are
421: confined to diffuse molecular and translucent clouds (as defined by Snow
422: \& McCall 2006), so we should not expect to find good correlations with either
423: dust or H I. If we do find good or decent correlations, that would indicate
424: that diffuse molecular and translucent clouds dominated the lines of sight
425: in our survey. On the other hand, those quantities that do correlate well
426: with each other would show that they arise in the same portion of the line
427: of sight. All correlations can be found in Table \ref{correlate},
428: including some not illustrated.
429:
430: The first correlation of interest is the one between HD and H$_2$, which is
431: shown in Figure \ref{ploteight}. There we see a very good correlation,
432: suggesting that the formation of HD depends on the local abundance of H$_2$,
433: or that both are formed by the same or similar processes. Because H$_2$
434: shoots up in column density due to the self-shielding about when N(H$_2$)
435: reaches 10$^{19}$ cm$^{-2}$, we checked for a change in slope around that
436: point. We see more scatter below log N(H$_2$) = 19.5 and no significant
437: change in slope (below 19.5, the slope is 1.55, based on a fairly narrow
438: range; above, it is 1.33). Considering all of our data points together
439: gives an overall slope of 1.25 and a correlation coefficient is 0.94.
440:
441: The significant departure from a slope of 1.00 in this correlation can be
442: interpreted in two different ways: either HD is being formed at a
443: faster rate than the formation rate of H$_2$ in the portion of the clouds
444: where both are forming; or HD is destroyed at a lesser rate than the
445: destruction rate of H$_2$ in those regions. This latter explanation may be
446: preferred, because of optical depth effects and how they affect the
447: photodestruction rates. For most of the observed range in A$_V$, the H$_2$
448: $J$ = 0 and 1 lines are damped, while the corresponding lines of HD lines are
449: becoming saturated - but not damped. In this range in A$_V$ (between A$_V$
450: about 0.1 and 0.8), the rate of HD destruction is rapidly declining, while the
451: photodestruction rate of H$_2$ is only slightly declining \cite{LePetit}.
452: This situation mimics the appearance of relative rate of formation favoring
453: HD, and that explains the slope of the N(HD)/N(H$_2$).
454:
455: The alternative explanation of the slope in Figure \ref{ploteight}, that HD
456: is being formed faster than H$_2$, could only be viable if some additional
457: formation process were at work, in addition to the standard gas-phase
458: reaction network. And that would imply HD formation on grain surfaces
459: competes with the gas-phase mechanism. But the detailed analysis of HD
460: formation on grains by Cazaux et al. (2008) explores conditions in which
461: grain formation of HD could be important, and our observed lines of sight
462: do not meet the criteria outlined by Cazaux et al. So our explanation, that
463: HD has started to be destroyed at a lesser rate, is more likely.
464:
465: This finding of a greater slope than 1.00 seems to be inconsistent with the
466: gas-phase models of Le Petit et al. (2002), who predict that the ratio of
467: HD over H$_2$ should decline over our entire range in molecular fraction.
468: Their Figure 4 shows the ratio N(HD)/2N(H$_2$) as a function of the molecular
469: fraction of hydrogen, and this figure shows a steady decline with
470: increasing molecular fraction. Only above a molecular fraction of 0.9
471: does the ratio start to increase. The study by Le Petit et al. only considers
472: a single cloud, whereas real sightlines cover many clouds in most cases.
473: But it is difficult to see how this helps reconcile the model with the
474: observations, because the situation only gets worse when multiple clouds,
475: which collectively form the total column density, are considered. The model
476: would be correct only if all the molecular regions along the lines of
477: sight, each having a molecular fraction $\ge$ 0.9, contained {\it all} the
478: HD and H$_2$, with no spillover to less dense regions.
479:
480: Figure \ref{plotnine} shows the correlation between the column densities of
481: HD and CO (the CO data were taken from Burgh et al. 2007). Here we see a very
482: good correlation (coefficient 0.93), which we take as an indication that
483: related processes account for both HD and CO. Since CO is formed by
484: gas-phase reactions (e.g Kaczmarczyk 2000a; 2000b; and references cited
485: therein), we again conclude that HD is also likely formed by gas-phase
486: reactions. The ion-neutral reaction H$_2$ + D$^+ \to$ HD + H$^+$
487: \cite{LePetit}, with the ion D$^+$ being created by cosmic-ray impacts, is
488: commonly thought to be the primary gas-phase formation channel for HD.
489:
490: To test our assertion that HD and H$_2$ do not inhabit the whole line of sight
491: in most cases, we show the correlations of both with quantities that are
492: expected to arise over the entire sightline. In Figure \ref{plotten} we show
493: the correlations of HD with the dust parameters total H (i.e., N(H I) +
494: 2N(H$_2$)); Figure \ref{ploteleven} shows the correlations of H$_2$ and HD
495: with E(B-V); and Figure \ref{plottwelve} shows the correlations of H$_2$
496: and HD with A$_V$. These correlations do not differ significantly from
497: each other, except there is a hint that both HD and H$_2$ go with E(B-V)
498: slightly better than they correlate with either total H, or A$_V$. Neither
499: HD or H$_2$ correlate significantly with R$_V$.
500:
501: The correlations of HD and H$_2$ with E(B-V) are better than those with H
502: I, telling us something surprising and perhaps useful: that the grains
503: responsible for differential extinction tend to be in the densest parts of the
504: observed lines of sight, where HD and H$_2$ are present. The grains
505: responsible for E(B-V) are thought to be comparable in size to the
506: wavelengths being scattered, so we conclude that other grains in the lines
507: of sight may have a different size distribution.
508:
509: We tested correlations of both H$_2$ and HD against the molecular fraction,
510: given by f = N(H$_2$)/N(H$_{Total}$), and found a fairly good correlation
511: (see Table \ref{correlate}). Neither should be a surprise, because they
512: just show that H$_2$ and HD exist primarily in the densest portions of the
513: lines of sight, as expected.
514:
515: Finally, we have looked at correlations between iron depletion, as shown by
516: the ratio of iron to total hydrogen, and we see several interesting
517: things. These correlations are shown in Figure \ref{plotthirteen}. We could
518: have considered correlations with depletions of other elements, but we have
519: more complete data on iron for our lines of sight, and we expect other
520: refractory elements to follow more or less the same trends as iron. The
521: iron column densities come from Jensen \& Snow (2007), and are listed in
522: Table \ref{table1}.
523:
524: The three correlations we considered are consistent with the idea that the
525: depletions are density dependent. Remembering that an increasingly negative
526: value of the Fe/H ratio means more depletion of iron, we see that the
527: depletion is growing with indicators of density such as the HD column
528: density, the fractional ratio of HD (i.e., N(HD)/N(H$_{Total}$); and the
529: ratio of HD to H$_2$ (see Figure \ref{plotthirteen}). The indicators we see
530: here, showing that depletion increases with density, is consistent with
531: the finding of Burgh et al. (2007), who found that the depletions correlate
532: with CO, another density indicator.
533:
534:
535: \section{Summary and Conclusions}
536:
537: We have derived column densities of interstellar HD for 41 lines of sight,
538: covering a wide range of total extinctions, total hydrogen column density,
539: and differential extinction E(B-V). We went to considerable effort to
540: reduce error in the derived column densities, which were ultimately found
541: using the curve-of-growth method, after measuring the equivalent
542: widths. One challenge we faced, and successfully resolved, was that the
543: weakest line of HD, detectable only toward highly-reddened stars, was
544: contaminated by a weak C I* line. We had to model and remove the C I*
545: line in those cases. Where possible, we checked our results against other
546: published column densities, with generally good agreement.
547:
548: We found values of the logarithmic ratio of N(HD)/2N(H$_2$) ranging from
549: -6.18 to -5.13 -- generally much lower than the galactic atomic D/H ratio
550: and even lower with respect to the extragalactic (or the primeval) ratios
551: \cite{Linsky}. This says two things: (1) the enhanced photodissociation of HD
552: over H$_2$ dominates and lowers the N(HD)/2N(H$_2$) ratio; and (2) our initial
553: assumption that we can not use N(HD)/2N(H$_2$) ratio as a cosmological test
554: is confirmed, and only the atomic ratio gives a true value of the
555: primordial production of deuterium (if that; see Linsky et al. 2006).
556:
557:
558: In order to see whether our results are consistent with extant chemical
559: models, we examined correlations of HD with other interstellar parameters,
560: finding the following: (1) HD is declining more slowly than the destruction
561: rate of H$_2$; (2) N(HD)
562: correlates well with N(CO); and (3) the depletion of iron (and probably
563: other refractory elements) is enhanced in the densest portions of the lines
564: of sight. The model of Le Petit et al. (2005) of the formation of HD is
565: inconsisten with the observations, unless individual dense clouds along the
566: lines of sight contain {\it all} the HD and H$_2$.
567:
568: \acknowledgements
569:
570: We are very grateful to the anonymous referee, who stimulated us to
571: revise this paper in many useful ways. We are grateful to several
572: colleagues (including especially Dan Welty and
573: Mike Shull) who gave us their advice, not to mention their data; and to
574: NASA, which funded this research through grants NNG05GA85G and NAS5-98043.
575: Also thanks to Ms. Lynsi Aldridge who helped with the early analysis of the
576: data.
577:
578: \begin{thebibliography}{99}
579: \bibitem[Abgrall \& Roueff 2006]{Abgrall} Abgrall, H., \& Roueff, E., 2006,
580: A\&A 445, 361
581: \bibitem[Black \& Dalgarno 1973]{Black} Black, J. H., \& Dalgarno, A., 1973,
582: \apj, 184, L101
583: \bibitem[Burgh et al. 2007]{Burgh} Burgh, E. B., France,
584: K., McCandliss, \& Stephan R., 2007, \apj, 658, 446
585: \bibitem[Cazaux et al. 2008]{Cazaux} Cazaux, S., Caselli, P., Cobut, V., \&
586: Le Bourlot, J., 2008, ArXiv e-prints, 802, arXiv:0802. 3319
587: \bibitem[Danks et al. 1984]{Danks} Danks, A. C., Federman, S. R., \&
588: Lambert, D. L., 1984, A\&A, 130, 62
589: \bibitem[Diplas \& Savage 1994]{Diplas} Diplas, A., \& Savage, B. D., 1994,
590: \apjs, 93, 211
591: \bibitem[Draine 2003]{draine}Draine, B. T. 2003, ARA\&A, 41, 241
592: \bibitem[Hipparcos]{hiptycho} ESA, 1997, The Hipparcos and Tycho
593: Catalogues, ESA SP-1200
594: \bibitem[van Dishoeck \& Black 1089]{vanDishoeck} van Dishoeck, E. F., \&
595: Black, J. H. 1989, \apj, 340, 273
596: \bibitem[Ferlet et al. 2000]{Ferlet} Ferlet, R., et al. 2000, \apj, 538, 69
597: \bibitem[Froese Fischer 2006]{Froese} Froese Fischer, C., 2006, JPhB, 39, 2159
598:
599: \bibitem[Jenkins et al. 1973]{Jenkins} Jenkins, E. B., Drake, J. F.,
600: Morton, D. C., Rogerson, J. B., Spitzer, L., \& York, D. G., 1973, \apj,
601: 181, L122
602: \bibitem[Jensen \& Snow 2007]{jensen} Jensen, A. G., \& Snow, T. P., 2007,
603: \apj, 669, 378
604: \bibitem[Kaczmarczyk 2000a]{Kac2000a} Kaczmarczyk, G., 2000a, \mnras, 312, 794
605: \bibitem[Kaczmarczyk 2000b]{Kac2000b} Kaczmarczyk, G., 2000b, \mnras, 316, 875
606: \bibitem[Lacour et al. 2005]{Lacour} Lacour, S., et al. 2005, A\&A, 430, 967
607: \bibitem[Le Petit et al. 2002]{LePetit} Le Petit, F., Roueff, E., \& Le
608: Bourlot, J. 2002, A\&A, 390, 369
609: \bibitem[Linsky et al. 2006]{Linsky} Linsky, J. L., et al. 2006, \apj, 647,
610: 1106
611: \bibitem[Magnani et al. 1998]{Magnani}Magnani, L., Onello, J. S., Adams,
612: N. G., Hartmann, D., \& Thaddeus, P. 1998, \apj, 504, 290
613: \bibitem[Morton 1975]{Morton75} Morton, D. C., 1975, \apj, 197, 85
614: \bibitem[Morton 1978]{Morton78} Morton, D. C., 1978, \mnras, 184, 713
615: \bibitem[Morton 1991]{Morton91} Morton, D. C., 1991, \apjs, 77, 119
616: \bibitem[Morton 2003]{Morton03} Morton, D. C., 2003, \apjs, 149, 205
617: \bibitem[Rachford et al. 2002]{Rachford} Rachford, B. L., et al. 2002,
618: \apj, 577, 221
619: \bibitem[Raboud et al. 1997]{Raboud} Raboud, D., Cramer, N., \& Bernasconi,
620: P. A., 1997 A\&A, 325, 167
621: \bibitem[Sahnow et al. 2002]{fuse}Sahnow, D., Hart, H., Dixon, V., Oegerle,
622: B., Murphy, E., \& Kriss, J., 2002, The {\it FUSE} Instrument and Data
623: Handbook, version 2.1, (Baltimore: Johns Hopkins University),
624: http://fuse.pha.jhu.edu/analysis/IDH/IDH.html
625: \bibitem[Savage et al. 1985]{Savage} Savage, B. D., Massa, D., Meade, M., \&
626: Wesselius, P. R., 1985, \apjs, 59, 397
627: \bibitem[Snow \& McCall 2006]{sno206} Snow, T. P., \& McCall, B. J., 2006,
628: ARAA, 44, 367
629: \bibitem[Sonnentrucker et al. 2003]{Sonnentrucker} Sonnentrucker, P.,
630: Friedman, S. D., Welty, D. E., York, D. G., \& Snow, T. P., 2003, \apj,
631: 596, 350
632: \bibitem[Spitzer et al. 1973]{Spitzer} Spitzer, L., Drake, J. F., Jenkins,
633: E. B., Morton, D. C., Rogerson, J. B., \& York D. G., 1973, \apj, 181 L116
634: \bibitem[Spitzer et al. 1974]{Spitzer74} Spitzer, L., Cochran, W. D., \&
635: Hirshfeld, A., 1974, \apjs, 28, 373
636: \bibitem[Valencic et al. 2004]{Valencic} Valencic, L. A., Clayton, G. C., \&
637: Gordon K. D., 2002, \apj, 616, 912
638: \bibitem[Watson 1973]{Watson} Watson, W. D., 1973, \apj, 182, L73
639: \bibitem[Zatsarinny \& Froese Fischer 2002]{Zats}Zatsarinny, O. \& Froese
640: Fischer, C., 2002, JPhB, 35, 4669
641:
642:
643:
644:
645: \end{thebibliography}
646:
647:
648: \clearpage
649:
650:
651: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Tables %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
652: \begin{deluxetable}{lccccccclcc}
653: \tablecolumns{11}
654: \tablewidth{0pc}
655: \tabletypesize{\scriptsize}
656: %\rotate{}
657: \tablecaption{Sightline Properties \label{table1}}
658: \tablehead{Star &E(B-V) &$R_V$ &Ref.\tablenotemark{a} &Molec. &log
659: N(H$_2$) &log N(CH) & log N(H$_{Tot}$) &Ref.\tablenotemark{a} &log N(Fe II)\tablenotemark{b} &log N(C I*)\tablenotemark{c} \\
660: (HD) &&&&Fract. &(cm$^{-2}$) &(cm$^{-2}$) &(cm$^{-2}$) &&(cm$^{-2}$) &(cm$^{-2}$) }
661:
662: \startdata
663:
664: 12323 &0.26 &2.75 &1 &0.21 &20.31 & ... &21.29 &4 &15.02 &... \\
665: 15558 &0.83 &2.94 &2 &0.32 &20.89 & ... &21.69 &4 &... &... \\
666: 24534 &0.45 &3.47 &3 &0.76 &20.92 &13.57 &21.34 &5 &14.63 &14.98 $\pm$ 0.02 \\
667: 27778 &0.38 &2.72 &3 &0.56 &20.79 &13.48 &21.34 &5 &... &15.15 $\pm$ 0.32 \\
668: 45314 &0.46 &4.42 &1 &0.42 &20.60 & ... &21.28 &4 &... &... \\
669: 53367 &0.74 &2.38 &3 &0.52 &21.06 & ... &21.65 &3 &15.08 &15.16 $\pm$ 1.83 \\
670: 73882 &0.72 &3.37 &3 &0.67 &21.11 &13.57 &21.58 &5 &15.15 &16.35 $\pm$ 0.64 \\
671: 74920 &0.35 &... &... &0.21 &20.26 &... &21.25 &4 &... &... \\
672: 90087 &0.30 &... &... &0.08 &19.77 & ... &21.18 &4 &15.15 &... \\
673: 91651 &0.30 &3.35 &1 &0.02 &19.07 & ... &21.16 &4 &15.23 &... \\
674: 91824 &0.26 &3.35 &1 &0.09 &19.84 & ... &21.19 &4 &... &... \\
675: 93204 &0.42 &... &... &0.04 &19.77 & ... &21.43 &4 &... &... \\
676: 93205 &0.37 &3.25 &1 &0.05 &19.75 & ... &21.35 &4 &15.35 &14.13 $\pm$ 0.05 \\
677: 93206 &0.39 &... &... &0.03 &19.52 & ... &21.35 &4 &... &14.13 $\pm$ 0.17 \\
678: 93222 &0.36 &4.76 &1 &0.03 &19.77 & ... &21.55 &4 &... &13.77 $\pm$ 0.04 \\
679: 94493 &0.20 &... &... &0.18 &20.12 & ... &21.17 &4 &15.38 &... \\
680: 101131 &0.34 &... &... &... &20.27 & ... &... &4 &... &... \\
681: 101190 &0.36 &2.48 &1 &0.27 &20.42 & ... &21.29 &4 &... &14.54 $\pm$ 0.11 \\
682: 101413 &0.36 &... &... &0.22 &20.38 & ... &21.34 &4 &... &... \\
683: 101436 &0.38 &... &... &0.22 &20.38 & ... &21.34 &4 &... &14.49 $\pm$ 0.22 \\
684: 104705 &0.26 &2.81 &1 &0.12 &19.93 & ... &21.16 &4 &15.22 &... \\
685: 110432 &0.40 &... &... &0.55 &20.64 &13.19 &21.20 &5 &... &14.55 $\pm$ 0.26 \\
686: 116852 &0.22 &2.42 &1 &0.12 &19.78 & ... &21.01 &4 &... &... \\
687: 147888 &0.52 &4.06 &3 &0.10 &20.46 & ... &21.71 &3 &14.89 &15.02 $\pm$ 0.10 \\
688: 148422 &0.28 &3.02 &1 &0.16 &20.13 & ... &21.23 &4 &... &... \\
689: 149404 &0.68 &3.28 &3 &0.33 &20.79 & ... &21.57 &3 &15.23 &14.86 $\pm$ 0.09 \\
690: 152233 &0.45 &2.95 &1 &0.17 &20.29 & ... &21.37 &4 &... &... \\
691: 152248 &0.45 &3.68 &1 &... &20.29 & ... &... &4 &... &... \\
692: 152723 &0.42 &3.36 &1 &0.13 &20.29 & ... &21.49 &4 &... &... \\
693: 161807 &0.11 &... &... &... &19.86 & ... &... &4 &... &... \\
694: 177989 &0.25 &2.83 &1 &0.23 &20.12 & ... &21.06 &4 &14.81 &14.52 $\pm$ 0.03 \\
695: 185418 &0.51 &2.32 &3 &0.47 &20.76 &13.12 &21.56 &5 &... &14.35 $\pm$ 0.08 \\
696: 192639 &0.66 &2.84 &3 &0.32 &20.69 &13.45 &21.47 &5 &... &14.54 $\pm$ 0.08 \\
697: 199579 &0.36 &2.95 &3 &0.38 &20.53 &13.36 &21.25 &5 &... &14.78 $\pm$ 0.12 \\
698: 201345 &0.17 &... &... &0.03 &19.43 & ... &20.91 &6 &... &... \\
699: 206267 &0.53 &2.67 &3 &0.42 &20.86 &13.41 &21.54 &5 &... &14.91 $\pm$ 2.09 \\
700: 207198 &0.62 &2.42 &3 &0.38 &20.83 &13.56 &21.55 &5 &... &15.18 $\pm$ 0.16 \\
701: 207538 &0.64 &2.25 &3 &0.43 &20.91 &13.63 &21.58 &5 &... &15.23 $\pm$ 0.15 \\
702: 224151 &0.42 &2.64 &2 &0.26 &20.57 & ... &21.45 &4 &15.38 &14.34 $\pm$ 0.06 \\
703: 303308 &0.43 &3.02 &1 &0.11 &20.24 &... &21.50 &4 &15.46 &14.38 $\pm$ 2.61 \\
704: 308813 &0.28 &... &... &0.22 &20.29 & ... &21.26 &4 &... &... \\
705:
706: \enddata
707: \tablenotetext{a}{References: (1) Valencic et al. 2004; (2) Rachford, private
708: communication; (3) Rachford in preparation, 2008;
709: \\(4) Shull in preparation, 2008; (5) Rachford et al. 2002; (6) Burgh et
710: al. 2007}
711: \tablenotetext{b}{Jensen \& Snow 2007}
712: \tablenotetext{c}{This study}
713:
714:
715: \end{deluxetable}
716:
717:
718: \begin{deluxetable}{lllllll}
719: \tablecolumns{7}
720: \tablewidth{0pc}
721: \tabletypesize{\scriptsize}
722: \tablecaption{Observation data \label{table2}}
723: \tablehead{Star &Stellar &V &Dist\tablenotemark{a} &Ref. &{\it FUSE} Data ID &STIS Data ID \\
724: (HD) &Type &(mag)&(pc)& && }
725:
726: \startdata
727: 12323 &O9V &8.92 &3900 &1 & P1020202000 &\\
728: 15558 &O5IIIf &7.95 &2187 &2 & P1170101000 &\\
729: 24534 &O9.5pe &6.10 &398 &2 & P1930201000 &o66p01020 \\
730: 27778 &B3V &6.33 &220 &3 & P1160301000 &o59s01010 \\
731: 45314 &O9pe &6.60 &799 &2 & P1021301000 &\\
732: 53367 &BoIV:e &6.94 &247 &3 & P1161101000 &\\
733: 73882 &O9III &7.24 &925 &1 & P1161302000 &\\
734: 74920 &O8 &7.54 &1497 &2 & P1022601000 &\\
735: 90087 &O9.5III&7.80 &2716 &2 & P1022901000 &\\
736: 91651 &O9Vn &8.87 &3500 &1 & P1023102000 &\\
737: 91824 &O7V &8.15 &4000 &3 & A1180802000 &\\
738: 93204 &O5Vf &8.48 &2630 &2 & P1023501000 &\\
739: 93205 &O3V &7.76 &2600 &1 & P1023601000 &o4qx01030 \\
740: 93206 &O9.7Ibn&6.24 &2512 &2 & P1023401000 &\\
741: 93222 &O7IIIf &8.11 &2900 &1 & P1023701000 &o4qx02030 \\
742: 94493 &B0.5Iab&7.27 &3327 &2 & P1024101000 &\\
743: 101131 &O6Vf &7.16 &709 &3 & P1024901000 &\\
744: 101190 &O6Vf &7.27 &2399 &2 & P1025001000 &\\
745: 101413 &O8V &8.33 &2399 &2 & P1025301000 &\\
746: 101436 &O6.5V &7.58 &2399 &2 & P1025401000 &o6lz51010 \\
747: 104705 &B0Ib &7.83 &3898 &2 & P1025701000 &\\
748: 110432 &B1IIIe &5.24 &301 &3 & P1161401000 &\\
749: 116852 &O9III &8.49 &4760 &2 & P1013801000 &\\
750: 147888 &B3 &6.78 &136 &3 & P1161501000 &o59s05010 \\
751: 148422 &B1Ia &8.64 &8836 &2 & P1015001001 &\\
752: &&&& & P1015002001 &\\
753: &&&& & P1015003001 &\\
754: 149404 &O9Iae &5.47 &1380 &2 & P1161702000 &\\
755: 152233 &O6III &6.59 &1905 &2 & P1026702000 &\\
756: 152248 &O7Ibnfp&6.11 &1758 &4 & P1026801000 &\\
757: 152723 &O6.5IIIf &7.31 &1905 &2 & P1027102000 &\\
758: 161807 &B0IIIn &7.01 &383 &3 & P1222302000 &\\
759: 177989 &B0III &9.34 &4909 &2 & P1017101000 &\\
760: 185418 &B0.5V &7.45 &950 &5 & P1162301000 &o5c01q010 \\
761: 192639 &O7Ib(f) &7.11 &1100 &1 & P1162401000 &o5c08t010 \\
762: 199579 &O6Ve &5.96 &794 &2 & P1162501000 &\\
763: 201345 &O9V &7.75 &1907 &2 & P1223001000 &\\
764: 206267 &O6.5V &5.62 &850 &1 & P1162701000 &\\
765: 207198 &O9.5Ib-II &5.96 &832 &2 & P1162801000 &o59s06010 \\
766: 207538 &O9V &7.31 &832 &2 & P1162902000 &o63y01010 \\
767: 224151 &B0.5II &6.05 &1355 &2 & P1224103000 &o54308010 \\
768: 303308 &O3Vf &8.21 &2630 &2 & P1221602000 &o4qx04010 \\
769: 308813 &O9.5V &9.32 &2398 &2 & P1221903000 &\\
770: \enddata
771: \tablenotetext{a}{Reference: (1) Savage et al. 1985; \ (2) Diplas \&
772: Savage 1994; \ (3) The Hipparcos and Tycho Catalogues;
773: \ (4) Raboud et al. 1997; \ (5) Sonnentrucker et al. 2003}
774:
775:
776: \end{deluxetable}
777:
778: \clearpage
779:
780:
781: \begin{deluxetable}{lcc}
782: \tablecolumns{3}
783: \tablewidth{0pc}
784: \tabletypesize{\scriptsize}
785: \tablecaption{Previously Reported HD Column Densities Comparison \label{French_compare}}
786: \tablehead{ &Previous Values\tablenotemark{a} & Our Values \\
787: Star &log N(HD) &log N(HD) \\
788: (HD) & (cm$^{-2}$) &(cm$^{-2}$)}
789:
790: \startdata
791: 27778 &15.51 $^{+0.30}_{-0.33}$ &15.84 $\pm$ 0.12 \\
792: 73882 &15.76 $^{+0.21}_{-0.38}$ &16.03 $\pm$ 0.34 \\
793: 110432 &15.28 $^{+0.14}_{-0.17}$ &15.30 $\pm$ 0.08 \\
794: 185418 &15.63 $^{+0.16}_{-0.13}$ &15.51 $\pm$ 0.32 \\
795: 192639 &15.18 $^{+0.20}_{-0.27}$ &15.57 $\pm$ 0.22 \\
796: 206267 &15.32 $^{+0.23}_{-0.28}$ &15.54 $\pm$ 0.10 \\
797: 207538 &15.70 $^{+0.31}_{-0.28}$ &15.82 $\pm$ 0.09 \\
798: \enddata
799: \tablenotetext{a}{Values from Lacour et al. 2005}
800: \end{deluxetable}
801:
802:
803:
804: \begin{deluxetable}{lccccccc}
805: \tablecolumns{8}
806: \tablewidth{0pc}
807: \tabletypesize{\scriptsize}
808: \tablecaption{Equivalent Widths of HD lines \label{eq_width}}
809: \tablehead{Star &W$_{959.82}$ &W$_{975.58}$ &W$_{1011.46}$ &W$_{1021.46}$ &W$_{1054.29}$ &W$_{1066.27}$
810: &W$_{1105.86}$ \\
811: (HD) &(m\AA)&(m\AA) &(m\AA) &(m\AA) &(m\AA) &(m\AA) &(m\AA) }
812:
813: \startdata
814:
815: 12323 &43.9 $\pm$ 3.0 &53.0 $\pm$ 3.8 &60.4 $\pm$ 3.3 &55.1 $\pm$ 2.5 &55.0 $\pm$ 2.3 &46.4 $\pm$ 3.0 & 8.7 $\pm$ 2.5 \\
816: 15558 &... &... &... &86.0 $\pm$ 19.0&90.1 $\pm$ 12.5&70.3 $\pm$ 10.2&26.9 $\pm$ 7.6 \\
817: 24534 &25.8 $\pm$ 5.2 &26.5 $\pm$ 10.7&26.2 $\pm$ 4.2 &28.4 $\pm$ 2.8 &27.3 $\pm$ 1.8 &32.3 $\pm$ 2.3 &18.8 $\pm$ 1.4 \\
818: 27778 &... & ... &34.7 $\pm$ 5.6 &... &35.4 $\pm$ 2.7 &32.9 $\pm$ 4.4 &22.0 $\pm$ 2.2 \\
819: 45314 &... &... &57.0 $\pm$ 1.7 &59.9 $\pm$ 1.4 &59.2 $\pm$ 1.2 &53.4 $\pm$ 1.1 & $<$ 25.7 \\
820: 53367 &53.8 $\pm$ 10.5 & ... &... &59.8 $\pm$ 15.8&51.9 $\pm$ 5.7 &61.2 $\pm$ 8.6 &26.3 $\pm$ 4.9 \\
821: 73882 &... &... &... &43.0 $\pm$ 10.1&44.2 $\pm$ 6.3 &45.6 $\pm$ 16.0&24.0 $\pm$ 8.0 \\
822: 74920 &37.0 $\pm$ 3.3 &38.9 $\pm$ 3.8 &43.7 $\pm$ 1.9 &42.2 $\pm$ 1.5 &42.4 $\pm$ 1.7 &36.0 $\pm$ 1.6 &13.0 $\pm$ 0.8 \\
823: 90087 &20.6 $\pm$ 2.0 &26.1 $\pm$ 3.6 &30.3 $\pm$ 1.4 &30.7 $\pm$ 1.5 &23.7 $\pm$ 1.2 &17.5 $\pm$ 1.1 & $<$ 4.9 \\
824: 91651 & 3.2 $\pm$ 1.4 &... &5.3 $\pm$ 0.9 & 6.2 $\pm$ 1.3 & 3.9 $\pm$ 0.8 & 2.9 $\pm$ 1.0 &... \\
825: 91824 &15.4 $\pm$ 2.3 &19.1 $\pm$ 2.6 &34.1 $\pm$ 6.9 &33.7 $\pm$ 2.3 &21.7 $\pm$ 1.7 & ... & $<$ 3.8 \\
826: 93204 &19.8 $\pm$ 5.0 &23.9 $\pm$ 4.1 &41.5 $\pm$ 4.4 &32.4 $\pm$ 6.8 &23.1 $\pm$ 2.2 &... & $<$ 4.5 \\
827: 93205 &19.8 $\pm$ 1.3 &26.6 $\pm$ 1.5 &37.6 $\pm$ 1.4 &... &27.4 $\pm$ 1.0 &... &1.6 $\pm$ 0.6 \\
828: 93206 &14.8 $\pm$ 1.5 &18.3 $\pm$ 2.2 &23.2 $\pm$ 1.5 &... &17.7 $\pm$ 0.6 &14.1 $\pm$ 0.8 & $<$ 3.1 \\
829: 93222 &14.8 $\pm$ 2.1 &17.4 $\pm$ 2.4 &25.4 $\pm$ 1.9 &24.4 $\pm$ 1.6 &20.1 $\pm$ 1.3 &13.3 $\pm$ 1.8 & $<$ 6.6 \\
830: 94493 &13.3 $\pm$ 2.0 &16.2 $\pm$ 3.6 &20.6 $\pm$ 1.6 &... &17.3 $\pm$ 1.1 &... & $<$ 6.2 \\
831: 101131 &19.7 $\pm$ 2.7 &24.5 $\pm$ 2.9 &25.7 $\pm$ 2.2 &26.6 $\pm$ 1.7 &28.3 $\pm$ 1.3 &24.7 $\pm$ 1.5 &4.2 $\pm$ 1.4 \\
832: 101190 &22.1 $\pm$ 2.6 &23.3 $\pm$ 2.4 &27.1 $\pm$ 2.0 &28.0 $\pm$ 2.1 &28.1 $\pm$ 1.1 &26.6 $\pm$ 1.4 &8.0 $\pm$ 1.0 \\
833: 101413 &20.7 $\pm$ 4.1 &... &25.4 $\pm$ 3.7 &25.2 $\pm$ 2.4 &25.7 $\pm$ 2.5 &21.1 $\pm$ 3.6 &6.8 $\pm$ 2.2 \\
834: 101436 &18.9 $\pm$ 3.1 &20.2 $\pm$ 2.9 &20.4 $\pm$ 2.6 &21.0 $\pm$ 2.8 &20.9 $\pm$ 1.4 &18.0 $\pm$ 1.6 &10.2 $\pm$ 1.0 \\
835: 104705 & 9.4 $\pm$ 1.5 &12.6 $\pm$ 2.9 &13.8 $\pm$ 1.2 &14.0 $\pm$ 1.6 &13.0 $\pm$ 0.7 &... & $<$ 3.9 \\
836: 110432 &25.9 $\pm$ 4.0 &27.1 $\pm$ 6.2 &30.4 $\pm$ 2.8 &30.2 $\pm$ 2.1 &34.8 $\pm$ 2.1 &32.1 $\pm$ 1.7 &13.2 $\pm$ 1.5 \\
837: 116852 & 8.9 $\pm$ 1.2 &13.0 $\pm$ 1.4 &15.3 $\pm$ 0.9 &... &13.7 $\pm$ 0.7 &10.6 $\pm$ 0.8 & $<$ 6.1\\
838: 147888 & ... & ... &15.6 $\pm$ 4.8 &15.1 $\pm$ 6.0 &15.6 $\pm$ 2.4 &14.2 $\pm$ 3.0 &8.8 $\pm$ 1.6 \\
839: 148422 &23.5 $\pm$ 5.0 &21.9 $\pm$ 8.5 &22.9 $\pm$ 3.3 &26.2 $\pm$ 4.6 &26.0 $\pm$ 2.6 &... &7.3 $\pm$ 2.1 \\
840: 149404 &47.3 $\pm$ 7.0 &56.7 $\pm$ 11.1&61.1 $\pm$ 7.4 &53.7 $\pm$ 3.5 &63.7 $\pm$ 3.9 &55.5 $\pm$ 4.6 &26.4 $\pm$ 2.5 \\
841: 152233 &28.5 $\pm$ 1.9 &... &31.1 $\pm$ 1.8 &35.4 $\pm$ 1.7 &32.0 $\pm$ 1.5 &29.1 $\pm$ 2.1 &6.5 $\pm$ 1.6 \\
842: 152248 &27.8 $\pm$ 3.9 &... &33.3 $\pm$ 2.2 &35.8 $\pm$ 1.8 &30.6 $\pm$ 1.8 &28.2 $\pm$ 2.2 &9.3 $\pm$ 2.1 \\
843: 152723 &... &41.2 $\pm$ 3.5 &46.1 $\pm$ 6.1 &46.3 $\pm$ 3.1 &49.5 $\pm$ 2.4 &43.0 $\pm$ 1.9 &6.6 $\pm$ 1.4 \\
844: 161807 &13.5 $\pm$ 3.0 &18.3 $\pm$ 4.6 &32.1 $\pm$ 2.7 &27.7 $\pm$ 2.7 &17.4 $\pm$ 1.8 &15.3 $\pm$ 1.8 & $<$ 6.9 \\
845: 177989 &25.5 $\pm$ 3.6 &33.3 $\pm$ 8.8 &33.6 $\pm$ 2.6 &... &26.5 $\pm$ 1.9 &26.6 $\pm$ 2.7 &6.3 $\pm$ 1.6 \\
846: 185418 &32.0 $\pm$ 6.8 &... &46.3 $\pm$ 6.4 &... &50.8 $\pm $3.9 &49.7 $\pm$ 4.2 &17.4 $\pm$ 4.9 \\
847: 192639 &48.2 $\pm$ 7.8 &... &51.9 $\pm$ 6.1 &54.2 $\pm$ 4.0 &54.7 $\pm$ 3.7 &53.4 $\pm$ 4.1 &20.1 $\pm$ 4.0 \\
848: 199579 &52.6 $\pm$ 4.3 &52.7 $\pm$ 5.5 &71.6 $\pm$ 4.5 &64.6 $\pm$ 5.3 &61.9 $\pm$ 2.6 &60.3 $\pm$ 2.7 &8.4 $\pm$ 1.2 \\
849: 201345 &11.5 $\pm$ 0.8 &15.9 $\pm$ 1.4 &22.2 $\pm$ 1.0 &22.3 $\pm$ 1.0 &14.8 $\pm$ 0.8 &10.6 $\pm$ 0.8 & $<$ 2.4\\
850: 206267 &42.5 $\pm$ 2.5 &48.5 $\pm$ 9.1 &46.1 $\pm$ 5.1 &52.0 $\pm$ 3.7 &49.5 $\pm$ 2.4 &45.6 $\pm$ 3.2 &24.1 $\pm$ 2.1 \\
851: 207198 &47.1 $\pm$ 7.8 &53.8 $\pm$ 12.3&57.8 $\pm$ 7.1 &47.3 $\pm$ 4.6 &56.9 $\pm$ 2.8 &51.1 $\pm$ 4.3 &34.7 $\pm$ 3.2 \\
852: 207538 &... &... &... &46.6 $\pm$ 6.7 &57.4 $\pm$ 5.4 &46.6 $\pm$ 8.3 &28.0$\pm$ 3.2 \\
853: 224151 &23.4 $\pm$ 2.7 &... &28.9 $\pm$ 2.0 &... &28.4 $\pm$ 1.5 &23.2 $\pm$ 1.5 &5.8 $\pm$ 0.7 \\
854: 303308 &22.6 $\pm$ 2.4 &25.3 $\pm$ 2.5 &... &... &24.6 $\pm$ 1.5 &24.2 $\pm$ 2.8 &5.3 $\pm$ 1.1 \\
855: 308813 &21.3 $\pm$ 2.5 &26.8 $\pm$ 3.9 &38.4 $\pm$ 2.4 &35.7 $\pm$ 3.6 &31.6 $\pm$ 2.6 &24.4 $\pm$ 2.4 &4.0 $\pm$ 1.5 \\
856:
857: \enddata
858: \end{deluxetable}
859:
860:
861:
862:
863: \begin{deluxetable}{llll}
864: \tablecolumns{4}
865: \tablewidth{0pc}
866: \tabletypesize{\scriptsize}
867: \tablecaption{Absorption Lines and Oscillator Strengths \label{absorption_lines}}
868: \tablehead{\multicolumn{2}{c} {HD Lines} &\multicolumn{2}{c} {C I*lines} \\
869: Wavelength (\AA) &log(f$\lambda$)\tablenotemark{a} (cm) &Wavelength (\AA)
870: & log(f$\lambda$)\tablenotemark{b} (cm) }
871:
872: \startdata
873: 959.7968 &-6.8520 &945.338 &-5.842 \\
874: 975.5524 &-6.7190 &1157.4056 &-7.759 \\
875: 1011.4439 &-6.5769 &1157.7697 &-6.993 \\
876: 1021.4436 &-6.5867 &1158.5443 &-8.107 \\
877: 1054.2800 &-6.7632 &1158.6744 &-7.770 \\
878: 1066.2636 &-6.9114 &1158.7321 &-7.702 \\
879: 1105.8335 &-8.0849 &1276.7498 &-7.405 \\
880: &&1279.0562 &-7.676 \\
881: &&1279.8907 &-6.737 \\
882: &&1280.4043 &-7.249 \\
883: &&1280.5975 &-7.045 \\
884: &&1656.2672 &-6.004 \\
885: &&1657.3792 &-6.228 \\
886: &&1657.9068 &-6.103 \\
887: \enddata
888: \tablenotetext{a} {HD log(f$\lambda $) values from Abgrall \& Roueff (2006)}
889: \tablenotetext{b} {C I* log(f$\lambda $) values from Morton (2003) }
890:
891: \end{deluxetable}
892:
893:
894:
895:
896:
897:
898: \begin{deluxetable}{lccr}
899: \tablecolumns{4}
900: \tablewidth{0pc}
901: \tabletypesize{\scriptsize}
902: \tablecaption{1105.73 \AA\ C I* f-value \label{tab_CIfvalue}}
903: \tablehead{Star & $W_{1105.73}$ & log N(C I*) &f-value\\
904: &(m\AA) &(cm$^{-2}$)&(10$^{-3}$)}
905: \startdata
906: HD 12323 & 16 $\pm$ 3 & 14.52 & 6.0$^{+1.7}_{-1.4}$ \\
907: $o$ Per & 17 $\pm$ 4 & 14.88 & 5.7$^{+8.0}_{-2.8}$ \\
908: HD 207538 & 32 $\pm$ 5 & 14.93 & 9.0$^{+6.3}_{-3.6}$ \\
909: HD 210839 & 29 $\pm$ 10 & 14.85 & 5.6$^{+3.5}_{-2.4}$ \\
910:
911: Weighted Avg & & & 6.2$^{+3.5}_{-1.0}$ \\
912: \enddata
913: \end{deluxetable}
914:
915:
916:
917: \begin{deluxetable}{llccc}
918: \tablecolumns{5}
919:
920: \tablewidth{0pc}
921: \tabletypesize{\scriptsize}
922: \tablecaption{HD Column Density \label{colden}}
923: \tablehead{Star (HD) & \multicolumn{3}{c} {Single Component COG}
924: &\multicolumn{1}{c} {Multiple Component COG} \\ &log N(HD) &$b$-value &
925: &log N(HD) }
926:
927: \startdata
928: 24534 &15.88 $\pm$ 0.40\tablenotemark{a} &1.8 $\pm$ 0.2 & &15.71 $\pm$ 0.07 \\
929: 27778 &15.89 $\pm$ 1.10 &2.2 $\pm$ 0.6 & &15.84 $\pm$ 0.12 \\
930: 53367 &15.74 $\pm$ 0.34 &4.0 $\pm$ 0.7 & &16.23 $\pm$ 0.17 \\
931: 73882 &15.81 $\pm$ 1.18 &2.9 $\pm$ 1.1 & &16.03 $\pm$ 0.34 \\
932: 110432 &15.43 $\pm$ 0.12\tablenotemark{a} &2.3 $\pm$ 0.2 & &15.30 $\pm$ 0.08 \\
933: 147888 &15.39 $\pm$ 1.47 &1.0 $\pm$ 0.4 & &15.30 $\pm$ 0.15 \\
934: 149404 &15.74 $\pm$ 0.12\tablenotemark{a} &4.1 $\pm$ 0.3 & &15.50 $\pm$ 0.18 \\
935: 185418 &15.51 $\pm$ 0.32\tablenotemark{a} &3.7 $\pm$ 0.6 & &15.87 $\pm$ 0.32 \\
936: 192639 &15.57 $\pm$ 0.22\tablenotemark{a}\tablenotemark{b} &4.0 $\pm$ 0.4 & &15.10 $\pm$ 0.14 \\
937: 199579 &15.08 $\pm$ 0.09\tablenotemark{a} &6.5 $\pm$ 0.7 & &15.09 $\pm$ 0.02 \\
938: 206267 &15.72 $\pm$ 0.14\tablenotemark{a} &3.4 $\pm$ 0.2 & &15.54 $\pm$ 0.10 \\
939: 207198 &16.12 $\pm$ 0.47\tablenotemark{a} &3.4 $\pm$ 0.4 & &15.84 $\pm$ 0.08 \\
940: 207538 &15.83 $\pm$ 0.25\tablenotemark{a} &3.6 $\pm$ 0.5 & &15.82 $\pm$ 0.09 \\
941: \hline \\ \hline
942:
943: 12323 &15.02 $\pm$ 0.18 &5.4 $\pm$ 0.8 & &\\
944: 15558 &15.64 $\pm$ 0.32 &6.4 $\pm$ 1.4 & &\\
945: 45314 &15.30 $\pm$ 0.37\tablenotemark{a} &4.9 $\pm$ 0.7 & &\\
946: 74920 &15.33 $\pm$ 0.04\tablenotemark{a} &3.2 $\pm$ 0.1 & &\\
947: 90087 &14.30 $\pm$ 0.07 &5.7 $\pm$ 2.6 & &\\
948: 91651 &13.43 $\pm$ 0.45 &$\ge$0.6 & &\\
949: 91824 &14.17 $\pm$ 0.07\tablenotemark{a} &$\ge$4.8 & &\\
950: 93204 &14.22 $\pm$ 0.10\tablenotemark{a} &$\ge$4.2 & &\\
951: 93205 &14.28 $\pm$ 0.07 &$\ge$5.7 & &\\
952: 93206 &14.19 $\pm$ 0.10 &4.1 $\pm$ 2.7 & &\\
953: 93222 &14.19 $\pm$ 0.12 &$\ge$3.2 & &\\
954: 94493 &14.24 $\pm$ 0.32 &$\ge$1.3 & &\\
955: 101131 &14.85 $\pm$ 0.19\tablenotemark{a} &2.4 $\pm$ 0.3 & &\\
956: 101190 &15.10 $\pm$ 0.12 &2.2 $\pm$ 0.1 & &\\
957: 101413 &15.01 $\pm$ 0.31 &2.0 $\pm$ 0.3 & &\\
958: 101436 &15.36 $\pm$ 0.17 &1.4 $\pm$ 0.5 & &\\
959: 104705 &14.26 $\pm$ 0.53 &1.5 $\pm$ 1.2 & &\\
960: 116852 &14.15 $\pm$ 0.13 &2.0 $\pm$ 1.5 & &\\
961: 148422 &15.07 $\pm$ 0.28 &2.0 $\pm$ 0.3 & &\\
962: 152233 &14.95 $\pm$ 0.19 &2.9 $\pm$ 0.3 & &\\
963: 152248 &15.15 $\pm$ 0.18\tablenotemark{a} &2.6 $\pm$ 0.3 & &\\
964: 152723 &15.01 $\pm$ 0.13\tablenotemark{a} &4.5 $\pm$ 0.5 & &\\
965: 161807 &14.12 $\pm$ 0.07\tablenotemark{a} &$\ge$4.5 & &\\
966: 177989 &14.92 $\pm$ 0.22\tablenotemark{a} &2.6 $\pm$ 0.4 & &\\
967: 201345 &14.00 $\pm$ 0.04 &$\ge$5.8 & &\\
968: 224151 &14.91 $\pm$ 0.09\tablenotemark{a} &2.4 $\pm$ 0.2 & &\\
969: 303308 &14.90 $\pm$ 0.14 &2.2 $\pm$ 0.3 & &\\
970: 308813 &14.48 $\pm$ 0.14 &5.2 $\pm$ 2.1 & &\\
971:
972: \enddata
973: \tablenotetext{a}{Assumed a good fit then scaled errors to a reduced chi
974: square of one.}\tablenotetext{b}{The single COG is a better solution and
975: agrees with the Profile fitting column densities within errors while
976: the multiple component COG is not within errors and misfit the data (see Figure
977: \ref{plotthree}).}
978:
979: \end{deluxetable}
980:
981:
982:
983: \begin{deluxetable}{lcccc}
984: \tablecolumns{5}
985: \tablewidth{0pc}
986: \tabletypesize{\scriptsize}
987: \tablecaption{Column Density Comparison \label{profile_vsCOG}}
988: \tablehead{Star &\multicolumn{2}{c} {Profile Fitting} &\multicolumn{2}{c}{Curve of Growth}\\
989: &log N(HD) &$b$-value &log N(HD) &$b$-value\\
990: & (cm$^{-2}$)&&(cm$^{-2}$)& }
991:
992: \startdata
993: 90087 &14.36 $\pm$ 0.05 &3.59 &14.30 $\pm$ 0.07 &5.7 \\
994: 101436 &15.12 $\pm$ 0.08 &1.33 &15.36 $\pm$ 0.17 &1.4 \\
995: 177989 &14.92 $\pm$ 0.07 &2.59 &14.92 $\pm$ 0.22 &2.6 \\
996: 110432 & 15.32 $\pm$ 0.04 &2.32 &15.43 $\pm$ 0.12 &2.3\\
997: 192639 & 15.52 $\pm$ 0.07 &3.70 &15.57 $\pm$ 0.22 &4.0 \\
998: \enddata
999: \end{deluxetable}
1000:
1001:
1002:
1003: \begin{deluxetable}{lcccc}
1004: \tablecolumns{5}
1005: \tablewidth{0pc}
1006: \tabletypesize{\scriptsize}
1007: \tablecaption{Correlation Coefficients \label{correlate}}
1008: \tablehead{Quantities &$r$\tablenotemark{a}& $\rho$ & $\#$ &Slope}
1009: \startdata
1010: log N(HD) vs. log N(H$_2$) &0.94 &0.89 : 0.96 &55 &1.25 $\pm$ 0.03 \\
1011: log N(HD) vs. log N(CO) &0.93 &0.77 : 0.98 &13 &0.49 $\pm$ 0.02 \\
1012: log N(HD) vs. log N(H$_{Tot}$) &0.67 &0.48 : 0.80 &48 &1.80 $\pm$ 0.05 \\
1013: log N(HD) vs. E(B-V) &0.78 &0.64 : 0.87 &51 &3.94 $\pm$ 0.09 \\
1014: log N(H$_2$) vs. E(B-V) &0.75 &0.59 : 0.85 &51 &2.34 $\pm$ 0.04 \\
1015: log N(HD) vs. A$_V$ &0.56 &0.27 : 0.75 &35 &...\\
1016: log N(H$_2$) vs. A$_V$ &0.53 &0.23 : 0.73 &35 &...\\
1017:
1018:
1019:
1020: log N(Fe II/H$_{Tot}$) vs. log N(HD) &-0.78 &-0.92 : -0.44 &15 &-0.30 $\pm$ 0.07 \\
1021: log N(Fe II/H$_{Tot}$) vs. log N(HD/H$_{Tot}$) &-0.72 &-0.90 : -0.33 &15 &-0.33 $\pm$ 0.09 \\
1022: log N(Fe II/H$_{Tot}$) vs. log N(HD/2H$_2$) &-0.80 &-0.93 : -0.48 &15 &-0.85 $\pm$ 0.17 \\
1023: \hline $\qquad$Correlations not shown\\ \hline
1024: log N(HD) vs. log N(H I) &0.34 &0.07 : 0.57 &48 &... \\
1025: log N(H$_2$) vs. log N(H I) &0.33 &0.05 : 0.56 &48 &... \\
1026: log N(HD) vs. Rv &-0.25\tablenotemark{b} & ... : -0.54 &35 &... \\
1027: log N(H$_2$) vs. Rv &-0.26\tablenotemark{b} & ... : -0.54 &35 &... \\
1028: $f_{H2}$ vs. log N(HD) &0.74 &0.58 : 0.85 &48 &... \\
1029: $f_{H2}$ vs. log N(H$_2$) &0.81 &0.68 : 0.89 &48 &... \\
1030:
1031:
1032: \enddata
1033: \tablenotetext{a}{r is the Pearson correlation coefficient, $\rho$ is the
1034: 95$\%$ confidence interval(uncertainties in the correlation coefficients),
1035: $\#$ is the number of data points used in the correlation, and
1036: Slope is the slope of the weighted least-squares line fit.}
1037: \tablenotetext{b}{Correlations not 95$\%$ significant}
1038:
1039: \end{deluxetable}
1040:
1041:
1042: \clearpage
1043:
1044:
1045: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% COG's %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
1046: \begin{figure}
1047: \begin{center}
1048: \scalebox{.5} [.5]{\includegraphics{f1.eps}}
1049: \figcaption{Carbon I* Curve of growth. \label{plotone}}
1050: \end{center}
1051: \end{figure}
1052:
1053:
1054:
1055: \clearpage
1056:
1057:
1058:
1059: \begin{figure}
1060: \scalebox{.75} [1]{\includegraphics[angle=90]{f2.eps}}
1061: \caption{Single Component (left) and Multiple Component (right) Curves of
1062: Growth for HD 24534 through HD 149404. \label{plottwo}}
1063: \end{figure}
1064:
1065: \clearpage
1066:
1067: \begin{figure}
1068: \scalebox{.75} [1]{\includegraphics[angle=90]{f3.eps}}
1069: \caption{Single Component (left) and Multiple Component (right) Curves of
1070: Growth for HD 185418 through HD 207538. \label{plotthree}}
1071: \end{figure}
1072:
1073:
1074:
1075:
1076: \begin{figure}
1077: \begin{center}
1078: \scalebox{.4} [.36]{\includegraphics[angle=90]{f4.eps}}
1079: \caption{Column Density solutions from a multiple component curve of growth
1080: compared to a single component curve of growth. \label{plotfour}}
1081: \end{center}
1082: \end{figure}
1083:
1084:
1085:
1086:
1087: \clearpage
1088:
1089:
1090: \begin{figure}
1091: \scalebox{.8} [1]{\includegraphics[angle=90]{f5.eps}}
1092: \caption{Single Component Curves of Growth with column densities and $b$-values
1093: for HD 12323 through HD 152248. \label{plotfive}}
1094: \end{figure}
1095:
1096: \clearpage
1097:
1098: \begin{figure}
1099:
1100: \scalebox{1} [.9]{\includegraphics{f6.eps}}
1101: \caption{Single Component Curves of Growth with column densities and $b$-
1102: values for HD 152723 through HD 308813. \label{plotsix}}
1103: \end{figure}
1104:
1105:
1106:
1107:
1108:
1109:
1110: \begin{figure}
1111: \begin{center}
1112: \scalebox{.4} [.36]{\includegraphics{f7.eps}}
1113: \caption{Profile fit of HD 110432's seven absorption lines, simultaneously
1114: solving for column density (log N(HD) = 15.32) and $b$-value (2.32). The
1115: velocity is set to the rest wavelength due to varying velocity solutions
1116: from different detector segments. \label{plotseven}}
1117: \end{center}
1118: \end{figure}
1119:
1120:
1121:
1122:
1123: \clearpage
1124:
1125: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% science plots %%%%%%%%%%%%%%%%%
1126: \begin{figure}
1127: \scalebox{.4} [.4]{\includegraphics[angle=90]{f8.eps}}
1128: \caption{{Correlation plots of log N(HD) versus log N(H$_2$). The line is
1129: a weighted least-squares fit. Square plot symbols represent {\it
1130: Copernicus} data analyzed by Lacour et al. 2005, diamonds represent this
1131: study, asterisk and upper limit represent {\it Copernicus} data reported by
1132: Spitzer et al. 1974.} \label{ploteight}}
1133: \end{figure}
1134:
1135: \begin{figure}
1136: \scalebox{.32} [.32]{\includegraphics[angle=90]{f9.eps}}
1137: \caption{{Correlation of log N(CO) with log N(HD). CO data from Burgh et
1138: al. 2007}\label{plotnine}}
1139: \end{figure}
1140:
1141: \begin{figure}
1142: \scalebox{.32}[.32]{\includegraphics[angle=90]{f10.eps}}
1143: \caption{{Correlation plots of log N(HD) versus log N(total Hydrogen). Square
1144: plot symbols represent {\it Copernicus} data analyzed by Lacour et
1145: al. 2005, diamonds represent this study.} \label{plotten}}
1146: \end{figure}
1147:
1148: \begin{figure}
1149: \scalebox{.32} [.32]{\includegraphics[angle=90]{f11a.eps} \quad \includegraphics[angle=90]{f11b.eps}}
1150: \caption{{log N(HD) and log N(H$_2$) versus reddening,
1151: E(B-V). Symbols same as Figure \ref{plotten}.} \label{ploteleven}}
1152: \end{figure}
1153:
1154:
1155:
1156: \begin{figure}
1157: \scalebox{.32} [.32]{\includegraphics[angle=90]{f12a.eps} \quad \includegraphics[angle=90]{f12b.eps}}
1158: \caption{{log N(HD) and log N(H$_2$), versus extinction (A$_V$). Symbols same as Figure \ref{plotten}.}
1159: \label{plottwelve}}
1160: \end{figure}
1161:
1162:
1163:
1164: \begin{figure}
1165: \begin{center}
1166: \scalebox{.32} [.32]{\includegraphics[angle=90]{f13a.eps} \quad \includegraphics[angle=90]{f13b.eps}}
1167: \vspace{.5 cm}
1168: \scalebox{.32} [.32]{\includegraphics[angle=90]{f13c.eps}}
1169: \caption{{Correlations with iron depletions, a density indicator. Iron
1170: data in Table \ref{table1} originally reported by Jensen \& Snow 2007.}
1171: \label{plotthirteen}}
1172: \end{center}
1173: \end{figure}
1174:
1175:
1176:
1177:
1178: \end{document}
1179: