1: \documentclass[12pt,preprint]{aastex}
2: %%\documentclass[manuscript]{aastex}
3: %%\documentclass[preprint2]{aastex}
4: %%\documentclass[preprint2,longabstract]{aastex}
5:
6: \def\lapprox{\;\rlap{\lower 2.5pt % approx. lesser
7: \hbox{$\sim$}}\raise 1.5pt\hbox{$<$}\;}
8:
9: \shorttitle{Interstellar Krypton Abundance Variations}
10: \shortauthors{Cartledge et al.}
11:
12: \begin{document}
13:
14: \title{Interstellar Krypton Abundances: The Detection of Kiloparsec-scale
15: Differences in Galactic Nucleosynthetic History\footnote{Based on observations
16: with the NASA/ESA {\it Hubble Space Telescope (HST)} and the NASA-CNES-CSA {\it
17: Far-Ultraviolet Spectroscopic Explorer (FUSE)}. {\it HST} spectra were obtained
18: at the Space Telescope Science Institute, which is operated by the Association
19: of Universities for Research in Astronomy, Inc. under NASA contract No. NAS
20: 5-26555; {\it FUSE} is operated for NASA by the John Hopkins University under
21: NASA contract NAS-32985.}}
22:
23: \author{Stefan I.B. Cartledge}
24: \affil{Dept. of Physics and Astronomy, Louisiana State University, Baton Rouge,
25: LA 70803}
26: \email{scartled@gmail.com}
27:
28: \author{J. T. Lauroesch}
29: \affil{Department of Physics and Astronomy, University of Louisville,
30: Louisville, KY 40292}
31: \email{jtlaur01@louisville.edu}
32:
33: \author{David M. Meyer}
34: \affil{Department of Physics and Astronomy, Northwestern University, Evanston,
35: IL 60208}
36: \email{davemeyer@northwestern.edu}
37:
38: \author{Ulysses J. Sofia}
39: \affil{Department of Astronomy, Whitman College, Walla Walla, WA 99362}
40: \email{sofiauj@whitman.edu}
41:
42: \and
43:
44: \author{Geoffrey C. Clayton}
45: \affil{Dept. of Physics and Astronomy, Louisiana State University, Baton Rouge,
46: LA 70803}
47: \email{gclayton@fenway.phys.lsu.edu}
48:
49: \begin{abstract}
50: We present an analysis of \ion{Kr}{1} $\lambda$1236 line measurements from 50
51: sight lines in the {\it Hubble Space Telescope} Space Telescope Imaging
52: Spectrograph and Goddard High Resolution Spectrograph data archives that have
53: sufficiently high resolution and signal-to-noise ratio to permit reliable
54: krypton-to-hydrogen abundance ratio determinations. The distribution of Kr/H
55: ratios in this sample is consistent with a single value for the ISM within 5900
56: pc of the Sun, log$_{10}$(Kr/H) = $-9.02\pm0.02$, apart from a rough annulus
57: from between $\sim$600 and 2500 pc distant. The Kr/H ratio toward stars within
58: this annulus is elevated by approximately 0.11 dex, similar to previously noted
59: elevations of O/H and Cu/H gas-phase abundances beyond $\sim$800 pc. A
60: significant drop in the gas-phase N/O ratio in the same region suggests that
61: this is an artifact of nucleosynthetic history. Since the physical scale of the
62: annulus' inner edge is comparable to the radius of the Gould Belt and the outer
63: limit of heliocentric distances where the D/H abundance ratio is highly
64: variable, these phenomena may be related to the Gould Belt's origins.
65: \end{abstract}
66:
67: \keywords{Galaxy: abundances --- ISM: abundances --- ultraviolet: ISM}
68:
69: \section{Introduction}
70: Where krypton is detectable, its interstellar abundance has the capacity to be
71: a singularly reliable gauge of the degree of nucleosynthesis in a given region
72: of the interstellar medium (ISM). Krypton is a noble gas whose outer electron
73: shell is spherically symmetric. Consequently, it is not prone to forming either
74: mechanical or chemical bonds in diffuse interstellar clouds where, given its
75: ionisation potential relative to that of hydrogen (13.999 eV versus 13.598 eV),
76: it should be predominantly in neutral form. These properties imply that a
77: constant krypton to hydrogen interstellar abundance ratio is synonymous with a
78: well-mixed ISM on the length scale over which the gas is sampled. Conversely,
79: any significant departures from an established large-scale mean represent
80: nucleosynthetic artifacts that mixing has not yet erased. Identifying such
81: artifacts is crucial to understanding topics as varied as the details of dust
82: composition and the large-scale processes of galactic chemical evolution. In
83: particular, analyses of abundance departures from interstellar means constrain
84: the observed efficacy of mixing at different Galactic length scales, the
85: influence of isolated and/or prolonged but localized star formation events on
86: elemental abundance ratios, the yields of supernovae, and the amount and
87: composition of material available for dust formation. Because it is undepleted
88: in the diffuse ISM, elemental abundance ratios involving krypton can be very
89: sensitive probes for these fields of study.
90:
91: Krypton to hydrogen abundance ratios have been well-studied within a kiloparsec
92: of the Sun. The krypton line at 1235.838 {\AA} was first detected using the
93: Goddard High Resolution Spectrograph (GHRS) aboard {\it Hubble Space Telescope
94: (HST)} during observations of $\zeta$ Oph \citep{car91}. A comprehensive study
95: of GHRS detections of \ion{Kr}{1} $\lambda$1236 later demonstrated that the
96: Kr/H abundance ratio within 500 pc of the Sun is constant at the level of
97: log$_{10}$(Kr/H) = $-9.02\pm0.02$ \citep{car97}. This conclusion was consistent
98: with GHRS results from oxygen \citep{mey98} and nitrogen \citep{mey97}, two of
99: the most abundant elements in the ISM, which also exhibited a constant local
100: abundance relative to hydrogen. Moreover, since the interstellar krypton
101: abundance ratio is roughly 50\% of the Solar value ($-8.72\pm0.08$;
102: \citealt{lod03}), this study established that the deficit in the interstellar
103: krypton abundance with respect to its Solar value was similar to the deficits
104: for other elements (e.g., carbon, oxygen, and nitrogen), using the then-accepted Solar abundances. Myriad spectral observations using the Space Telescope Imaging Spectrograph (STIS) and revisions both to oscillator strengths and Solar abundances have lessened the similarity between the interstellar abundance deficits for these elements; nevertheless the uniformity of their elemental abundances in the local ISM has only {\it gained} further support from the new data. Indeed, since krypton does not readily deplete into grains or form molecules, the constancy of its abundance relative to hydrogen, independent of molecular hydrogen fraction [$f$(H$_2$)] or mean hydrogen sight line density [${\langle}n_{\rm H}\rangle$], has been used to establish that the diffuse ISM is well-mixed on lengthscales of a few hundred parsecs \citep{car03,car06}. This result is independently supported by other analyses, specifically some involving the composition of stellar atmospheres \citep{red03} and pre-solar meteoritic dust grains \citep{nit05}. As a result, it can be concluded that significant deviations from the mean interstellar Kr/H abundance ratio identify sites or regions within the Galaxy where the nucleosynthetic history differs from that of the local ($d \lapprox$ 500-700 pc) ISM.
105:
106: In order to apply this tool to exploring the limits of this well-mixed
107: interstellar region and to probe for variations in the Kr/H abundance ratio
108: based on physical conditions not found locally, it was necessary to search for
109: krypton along a wider variety of paths than were previously studied. Thus, we
110: have plumbed the depths of the MAST Archive at STScI for detections of
111: \ion{Kr}{1} $\lambda$1236, measuring the krypton abundance for each sight line
112: with a significant absorption feature. In this paper, we discuss the relevant
113: GHRS, STIS, and {\it Far Ultraviolet Spectroscopic Explorer (FUSE)}
114: observations (\S~\ref{observations}), then examine the uniformity of the Kr/H
115: abundance ratio with respect to key sight line properties
116: (\S~\ref{krypton_distribution}). Finally, in \S~\ref{distance_effect},
117: variations in Kr/H, O/H, O/Kr, N/O, and Cu/H are compared, in search of
118: reinforcement for the identified Kr/H trends, and possible explanations.
119:
120: \section{Observations and Abundance Measurements}
121: \label{observations}
122: The study of interstellar krypton abundances using GHRS initially established
123: our knowledge of the uniformity of the Kr/H ratio within 500 pc of the Sun \citep{car97}. When
124: STIS data were added to the analysis, the case for a singular interstellar Kr/H
125: abundance ratio was broadened to include 26 sight lines covering a range of
126: nearly five orders of magnitude in $f$(H$_2$) and over two orders of magnitude
127: in ${\langle}n_{\rm H}\rangle$, as only two exceptions to the remarkably narrow
128: distribution around log$_{10}$(Kr/H) = $-9.02\pm0.02$ were identified
129: \citep{car03}. Since the majority of the combined GHRS and STIS sample were
130: relatively short sight lines ($<$ 600 pc) and the two exceptions (HD 116852 and
131: HD 152590) were the only paths to penetrate neighboring spiral arms, these
132: deviations were thought to be related to this circumstance. Seeking to
133: investigate this hypothesis, we proposed to use STIS to observe eight Galactic
134: OB stars (GO9855), at distances up to 4.8 kpc, that intersect neighboring
135: spiral arms. The targets included both HD 116852 and HD 152590, so that the
136: previous krypton measurements could be compared with results derived from new
137: higher signal-to-noise ratio spectra. Each observation was designed to generate
138: spectra with S/N = 50 per resolution element for the E140H echelle grating
139: centered at 1271 {\AA} using the {0.2\arcsec$\times$0.9\arcsec} aperture. The
140: equivalent width of an unsaturated feature detected at 5$\sigma$ would thus be
141: 1.1 m{\AA}, permitting the measurement of log$_{10}$(Kr/H) values as small as
142: $-9.60$; this limit is a factor of three smaller than the smallest ratio
143: reported to that date.
144:
145: Unfortunately, STIS ceased to operate in mid-2004, after only four of our
146: planned targets were observed. Although we were now able to re-examine both
147: previously observed targets, only two new sight lines, one each directed toward
148: and away from the Galactic centre, were added. To achieve the stated goals of
149: GO9855, we were forced to search the STIS archive for all observations that
150: might possess unreported detections of krypton in either the 1164.8672 or
151: 1235.838 {\AA} resonance lines. Since this search was well-matched to our
152: archival proposal to investigate the abundances of rare elements (AR10643;
153: \citealt{lau08}), particularly those produced by slow and rapid neutron capture
154: processes, the searches were conducted concurrently. The relevant search
155: criteria were sufficient spectral resolution to distinguish components from
156: different spiral arms and limit contamination from unresolved saturation (no
157: datasets with resolution less than $\lambda/\Delta\lambda$ = 45000 were used),
158: and wavelength coverage that included one or both of the krypton resonance
159: lines previously mentioned. Among the archival datasets, 35 sight lines exhibit
160: some evidence of a krypton absorption feature associated with \ion{Kr}{1}
161: $\lambda$1236; observational details for the 26 sight lines with reliable
162: detections and the four GO9855 targets are given in Table~\ref{datalist}. The
163: table includes specifications for all GHRS, STIS, and {\it FUSE} observations
164: of these stars upon which the abundance measurements are based. Unfortunately,
165: the poor signal-to-noise ratios for spectral orders covering 1164.8672 {\AA} do
166: not permit reliable column density determinations from these data. Please note
167: that seven sight lines with marginal krypton detections (HD 3827, HD 93205, HD
168: 93222, HD 168941, HD 202347, HD 218915, and CPD $-$59 2603) have been rejected
169: because in each case the absorption feature's equivalent width was less than a
170: factor of 4 greater than the associated uncertainty. HD 36408B and HD 121968
171: have better than 5$\sigma$ features, but these are omitted for the reasons
172: outlined below.
173:
174: The spectral type for HD 36408 B, B7IV, is unsuitable for the direct
175: measurement of an interstellar atomic hydrogen abundance, so this quantity must
176: instead be gauged using the well-known empirical linear relationship between
177: $E(\bv)$ and $N$(H) (e.g., see \citealt{spi78}). Using the \citet{rac02}
178: calibration indicates an oxygen abundance of log$_{10}$(O/H) = $-3.17\pm0.30$,
179: about 0.24 dex above the interstellar mean for low-density sight lines
180: \citep{car04}. The gas-phase abundances of other elements, notably \ion{Mg}{2},
181: \ion{P}{2}, \ion{Ni}{2}, \ion{Cu}{2}, \ion{Ge}{2}, and especially \ion{Kr}{1},
182: range from 0.30 dex to 0.60 dex higher than their respective low-density
183: interstellar means, while \ion{Mn}{2} is roughly at the interstellar value (for
184: interstellar abundance references at log$_{10}{\langle}n_{\rm H}\rangle \approx
185: -0.11$, see \citealt{car06}). Due to the lack of $N$(H) and the unusual
186: depletion pattern along this sight line compared with sight lines analyzed by
187: \citet{car06}, we identify this path as a subject for future study but omit it
188: from the current analysis.
189:
190: We also identify HD 121968 as a sight line whose ambiguous krypton abundance
191: warrants further study. A strong absorption component is evident near
192: $v_{heliocentric}= -10$km/s in line profiles of \ion{O}{1} $\lambda$1355,
193: \ion{Si}{1} $\lambda$1425, \ion{P}{2} $\lambda$1301, all \ion{Mn}{2} lines,
194: \ion{Cu}{2} $\lambda$1358, \ion{Ga}{2} $\lambda$1414, \ion{Ge}{2}
195: $\lambda$1237, all \ion{C}{1} lines, and \ion{Cl}{1} $\lambda$1347. The
196: \ion{Kr}{1} $\lambda$1235 profile contains a similar feature at the same
197: helocentric velocity; however, an absorption spike of equal strength is
198: situated 5 km/s to the blue. In none of the other line profiles mentioned above
199: does a comparable component appear, which would make this sight line unique
200: among those studied to date if the krypton profile were accepted without
201: further consideration. Several nickel lines are the only profiles to exhibit
202: strong absorption at velocities removed from the central component; the feature
203: most closely approximating the second krypton absorption spike is at
204: $v_{heliocentric}\approx -18$km/s with nearly 75\% depth relative to the main
205: component. However, this feature is merely one of the strongest in a very
206: complex profile that bears very little resemblance to the line profile of
207: krypton or any other element. Lastly, there is no blueward component to match
208: the hint of absorption that corresponds to the main absorption feature in the
209: the \ion{Kr}{1} $\lambda$1164 profile, although this spectral order is
210: compromised by noise spikes comparable to the profile depth. Consequently, we
211: conjecture that the apparent \ion{Kr}{1} $\lambda$1235 absorption profile is
212: contaminated by noise, and omit the HD 121968 sight line from further analysis.
213:
214: Adding 30 new krypton gas-phase abundance measurements to previous
215: detections, there now exists a database of 56 Galactic sight lines with
216: published krypton column densities. Unfortunately, for three of these paths
217: there are no observations at short enough wavelengths to assess the amount of
218: molecular hydrogen intersected by the line of sight. Toward HD 37021 and HD
219: 37061, however, \citet{car01} argued that the low \ion{Cl}{1} abundance toward
220: each star implies that these sight lines do not have large H$_2$ column
221: densities. In the current sample, HD 208947 lacks this evidence for a small
222: $f$(H$_2$) but its krypton abundance is near the low end of the distribution
223: and we include it provisionally in the Kr/H abundance ratio analysis. However,
224: six stars whose spectra exhibit krypton absorption features also have
225: interstellar Lyman-$\alpha$ absorption profiles that are contaminated by
226: significant stellar absorption, hence the sample of Kr/H abundance ratios
227: contain values for only 50 of the 56 sight lines with krypton. The krypton
228: abundance for each path was determined using the methods established by
229: \citet{car01}, with a slight alteration to the calibration procedure. The new
230: spectra were calibrated using the standard STSDAS STIS data reduction routine
231: CALSTIS (v.2.13), with the two-dimensional scattered-light correction algorithm
232: SC2DCORR applied automatically. Although this procedure produces slightly
233: different calibrated spectra than the stisextract method \citep{how00} we used
234: in previous krypton analyses \citep{car01,car03}, a comparison of the
235: abundances derived when the data are calibrated using each method indicates
236: that the results generally agree within 0.02 dex. In view of this accord, and
237: since it is more efficient to process the data through a single algorithm, we
238: calibrated the new data using CALSTIS alone.
239:
240: In order to alleviate concerns of unresolved saturation, we followed the same
241: abundance determination procedures used in our previous work. Specifically,
242: both apparent optical depth (AOD) and profile-fitting methods were applied to
243: the data to check consistency, then the profile-fitting results were adopted
244: while AOD uncertainties were folded into the listed column density errors to
245: explicitly account for continuum placement uncertainty. For the majority of
246: the new sight lines, the spectral coverage included wavelengths from shortward
247: of 1200 {\AA} to longward of 1250 {\AA} so that the neutral hydrogen column
248: density could be assessed by continuum reconstruction near the Ly-$\alpha$
249: line. {\it FUSE} data were also available for nearly all of these paths, so
250: the same method was used to determine the molecular hydrogen column density
251: toward each target star. The krypton and hydrogen abundance data for the newly
252: identified krypton sight lines are given in Table~\ref{KrHdata}. For more
253: detail on the techniques mentioned here in brief, please see \citet{car03} and
254: references therein. All previously-published krypton and hydrogen abundances
255: for sight lines where \ion{Kr}{1} $\lambda$1236 has been detected are compiled
256: in Table 2 of that paper.
257:
258: \section{The Gas-Phase Kr/H Ratio Distribution}
259: \label{krypton_distribution}
260: The mean interstellar krypton gas-phase abundance level that \citet{car97}
261: derived, represented by the Kr/H ratio log$_{10}$(Kr/H) = $-9.02\pm0.02$ [Kr/H
262: = $(0.96\pm0.05)\times10^{-9}$], was remarkable in that each and every
263: measurement of the ratio agreed with this value within its 1$\sigma$ error bar.
264: Nine of the ten spectra compiled in that study were observations of stars
265: within 500 pc of the Sun, indicating that the ISM within this region is very
266: well mixed. This evidence for an homogeneous local ISM was supported by the
267: analysis of GHRS observations of two \ion{N}{1} absorption lines near 1160 {\AA}
268: and an \ion{O}{1} line at 1356 {\AA} in the spectra of stars with similar
269: distances, many of them the same stars toward which krypton was detected.
270: Gas-phase abundance levels of log$_{10}$(N/H) = $-4.12\pm0.02$ \citep{mey97}
271: and log$_{10}$(O/H) $-3.46\pm0.02$ (\citealt{mey98}; this value has been
272: adjusted to reflect updates in the oxygen line's oscillator strength) were
273: identified by these studies and although both nitrogen and oxygen, unlike
274: krypton, are susceptible to depletion from the gas phase, no deviation from the
275: interstellar means by more than 1.5$\sigma$ were evident. These data indicated
276: that the degree of mixing in the local ISM is independent of the elements'
277: different nucleosynthetic origins.
278:
279: \citet{car01} added 10 STIS-observed spectra into the discussion of krypton
280: abundance, probing denser and longer sight lines. Regardless of the much
281: broader volume of parameter space being studied, in terms of total column
282: density, ${\langle}n_{\rm H}\rangle$, extinction, and distance, each of their
283: new Kr/H data points also agreed with the unchanged overall interstellar mean
284: within 1$\sigma$. Most recently, \citet{car03} added 12 more paths to the
285: sample. Their data indicated that among the 22 sight lines less than $\sim$800
286: pc in length, all but one were consistent with a singular interstellar Kr/H
287: ratio, while the exception deviated by no more than 1.3$\sigma$. Of the 4 paths
288: between 1.5 and 4.8 kpc in length, however, only the two directed along the
289: curve of the Orion Spur (the local spiral arm situated between the
290: Carina/Sagittarius and Perseus Arms) agreed with the otherwise ubiquitous Kr/H
291: ratio already established. The exceptional sight lines, toward HD 116852 and
292: HD 152590, both intersected the Carina/Sagittarius Arm and it was suggested that
293: their roughly solar krypton abundances might be related to this property. It
294: should be noted, however, that the abundance measurements for these paths were
295: among the most uncertain, given complications with distinguishing weak
296: absorption components from the noise in the HD 116852 spectrum and the unusually
297: steep curvature of the HD 152590 stellar background near 1236 {\AA}.
298:
299: Both stars were re-observed for GO9855, with the following results:
300: log$_{10}$(Kr/H)$_{\rm HD 116852}$ = $-8.97^{+0.08}_{-0.09}$;
301: log$_{10}$(Kr/H)$_{\rm HD 152590}$ = $-8.93^{+0.06}_{-0.07}$. The new
302: observations have provided spectra integrated for more than 15 and 7 times the
303: previous exposure times, respectively. As a result, the difficulties with the
304: first measurements of krypton abundance along these sight lines were
305: dramatically reduced. Perhaps unsurprisingly, the Kr/H ratios for both sight
306: lines are closer to the interstellar mean than previously determined; the
307: krypton abundance toward HD 116852 matches log$_{10}$(Kr/H)$_{ISM}$ =
308: $-9.02\pm0.02$ within error, while the HD 152590 path is only the second, after
309: HD 37367, to deviate by as much as an amount approaching 1.3$\sigma$. The
310: remaining two sight lines successfully observed for the GO9855 program, HD
311: 40893 and HD 165246, agree within their respective uncertainties with the
312: interstellar mean krypton abundance (see Table~\ref{KrHdata}). So before the
313: results from the STScI Archive (by way of AR10643) are considered, the picture
314: we have constructed is of a remarkably uniform sample of 28 Kr/H ratios which
315: differ from a single value by no more than 1.3$\sigma$ (see
316: Figure~\ref{krnhplot}). Since krypton is undepleted along the sight lines that
317: can be probed using ultraviolet absorption features, this tight distribution is
318: strong evidence that the diffuse ISM is very well mixed on length scales of
319: several hundred parsecs. Combining the krypton result with data for other
320: elements (e.g., O, Mg, P, Mn, Ni, Cu, Ge) has implied that the composition of
321: the diffuse ISM is homogeneous to a limit of order 0.04 dex \citep{car06} on
322: this same spatial scale.
323:
324: However, this krypton abundance sample of 28 sight lines is dominated by short
325: paths extending less than a kiloparsec through the Galactic disk. To more fully
326: comment on the singular krypton level, and hopefully identify regions with
327: distinct nucleosynthetic histories, it became necessary to consider all
328: possible unpublished \ion{Kr}{1} $\lambda$1236 detections in the Archive. The
329: data from Table~\ref{KrHdata} and all previous krypton abundance measurements
330: have been encapsulated in Figures~\ref{krnhplot}, \ref{krh2plot},
331: \ref{krrgplot} and \ref{krdpplot}, which describe the fluctuations in Kr/H with
332: the sight line properties mean sight line hydrogen density [${\langle}n_{\rm
333: H}\rangle$], molecular hydrogen fraction [$f$(H$_2$)], Galactocentric radius
334: ($r_{\rm G}$), and sight line pathlength ($d_\ast$), respectively.
335:
336: The first two plots serve to re-iterate and emphasize previous conclusions
337: regarding the state of krypton in the interstellar medium. In
338: Figure~\ref{krnhplot}, the Kr/H gas-phase abundance ratio distribution does not
339: depend on ${\langle}n_{\rm H}\rangle$, the mean hydrogen sight line density.
340: Previously-published data are more uniform than the newer data in their
341: collective agreement, differing by less than 1.3$\sigma$ from the mean on a
342: point-by-point basis, but the 26 Archive-based Kr/H ratios were derived from
343: spectra not designed to measure krypton abundance and so generally suffer from
344: larger uncertainties and are more widely distributed. Nevertheless, upon
345: including these new data the overall interstellar mean shifts only slightly to
346: log$_{10}$(Kr/H)$_{ISM}$ = $-9.00\pm0.02$ [or Kr/H =
347: $(1.00\pm0.04)\times10^{-9}$] from the previous determination, and there are no
348: coherent trends with ${\langle}n_{\rm H}\rangle$. Figure~\ref{krh2plot} clearly
349: shows that the measured Kr/H ratio is also independent of a sight line's
350: molecular hydrogen fraction. This property is evident whether the molecular
351: hydrogen fraction is expressed strictly as a number, emphasizing the uniform
352: distribution of Kr/H ratios for $f$(H$_2$) $> 0.1$, or in logarithmic form,
353: which stresses the singular nature of the Kr/H ratio across several orders of
354: magnitude to very small values of $f$(H$_2$). In other words, the shielded
355: conditions that at the same time favor the enhancement of H$_2$ abundances
356: relative to \ion{H}{1} and elemental depletion into dust grains do not affect
357: the gas-phase abundance of krypton. Thus krypton is undepleted in the ISM,
358: within the density range and spatial regions probed to date by Galactic sight
359: lines.
360:
361: The broad density-parameter strokes used to define this portrait of
362: kiloparsec-scale krypton abundance uniformity and lack of depletion into dust,
363: however, mask some subtler features. In particular, the scatter among the older
364: measurements is only 0.06 dex, compared with a typical Kr/H ratio uncertainty
365: of 0.09 dex. Encompassing the new and Archival data into the sample increases
366: this scatter to 0.10 dex. And although the Archival Kr/H ratios are in general
367: less precise than earlier measurements, it should be noted that they are also
368: largely derived from longer paths through the Galactic disk. In fact, according
369: to Figures~\ref{krrgplot} and \ref{krdpplot}, the increased Kr/H ratio scatter
370: is not due solely to greater inherent uncertainty in the Archival measurements,
371: but is perhaps dominantly produced by variations dependent on spatial sight
372: line properties.
373:
374: Consider Figure~\ref{krrgplot}, which examines Kr/H as a function of the
375: galactocentric radius ($r_G$) of the observed OB star. In studies of stellar
376: atmospheres (e.g., \citealt{rol00}) and planetary nebulae (e.g.,
377: \citealt{mar02}), a clear abundance trend with $r_G$ is apparent. As a star's,
378: or planetary nebula's, distance from the Galactic center increases, elemental
379: abundances steadily fall. For example, the oxygen radial abundance gradient is
380: determined by these studies to be about $-0.07$ dex/kpc, attributed to enhanced
381: nucleosythesis in the central regions of the Galaxy relative to outlying areas.
382: It is more difficult to measure such trends using abundances determined from UV
383: absorption line profiles because these profiles are convolutions of features
384: from individual clouds along the line of sight to which the distances are
385: rarely known. Nevertheless, the Kr/H data are plotted as a function of $r_G$ in
386: Figure~\ref{krrgplot} to search for this trend. Unfortunately, only a few stars
387: are situated beyond 500 pc from the Sun's radial location in the disk, either
388: closer to or farther from the Galactic center (the adopted Solar galactocentric
389: radius for this analysis is 8.5 kpc). Moreover, the bulk of stars outside the
390: annulus are to be found nearer the Galactic centre than away from it. A
391: weighted linear regression to the radial dependence of Kr/H gives a slope of
392: $-0.013\pm0.014$ dex/kpc, which becomes a krypton radial abundance gradient of
393: $-$0.027 dex/kpc under the crude approximation that all absorbing material is
394: situated at the midpoint of each sight line (i.e., the radial gradient is
395: double the measured slope). The slope, however, includes zero within its
396: 1$\sigma$ uncertainty and the trend is not very visually compelling.
397: Consequently, these data are also consistent with a flat gradient.
398:
399: The small apparent slope with $r_G$ may simply be a secondary result of the
400: striking spatial pattern in krypton abundance that becomes clear when Kr/H is
401: plotted against the distance to the observed star, as in Figure~\ref{krdpplot}.
402: Previously published data are consistent with krypton abundance being
403: independent of the target distance, yet they conform to a pattern of elevated
404: Kr/H ratios for 750 pc $< d_* <$ 2500 pc when combined with measurements of
405: features found in Archival spectra. Note that the four GO9855 sight lines are
406: those black diamonds limned in red and that they are consistent with both
407: patterns. There is some overlap for targets at distances between 550 and 750 pc
408: from the Sun, where a fraction of the sight lines exhibit elevated krypton
409: abundances and the remainder match the level of the previously published
410: interstellar mean. The dot-dashed lines in this plot indicate the mean Kr/H
411: ratios for each distance interval. If the sight lines with pathlengths between
412: 550 and 750 pc are allotted by their krypton abundances to the groups with
413: pathlengths less than 550 pc and those between 750 and 2500 pc long, then paths
414: with endpoints in the defined annulus are elevated in Kr/H by 0.11 dex with
415: respect to the local ISM mean (see Table~\ref{aratmeans}). Notably, the Kr/H
416: ratios interior and exterior to the annulus are identical within their errors,
417: each differing from the value within the annulus by roughly 3$\sigma$.
418:
419: Although the difference between weighted Kr/H ratio means for the annulus and
420: non-annulus sets provide strong evidence that the Kr/H abundance ratio changes
421: with heliocentric distance, the effect is just at the 3$\sigma$ level and the
422: sample sizes are small. To quantitatively confirm the reality of the annulus of
423: elevated krypton abundances, we have applied the Student's t-test to the data.
424: The t-test is a statistical measurement that determines the likelihood that two
425: datasets are merely random samplings of the same underlying population---in
426: this case that of Kr/H abundance ratios. Expressing the data as sets of
427: log$_{10}$(Kr/H) values, we find that the annulus and non-annulus groups are
428: both reasonable approximations of normal distributions, that the probability
429: that they are derived from the same underlying population is 0.000034\%, and
430: that at the 95\% confidence level the difference between the means for these
431: two samples is at least 0.09 dex. Therefore, we conclude that these groups of
432: sight lines indeed represent distinct populations. This analysis blends
433: seamlessly with a more qualitative description of the distinctions between the
434: two distributions. Specifically, the breadth of the krypton abundance
435: distribution interior to the annulus is characterized by a scatter of 0.06 dex,
436: indicating a particularly tight grouping given the typical uncertainty of 0.09
437: dex in this measurement. The scatters for the groups of sight lines both ending
438: within the annulus and exterior to it are larger, reflecting the larger typical
439: uncertainties for the corresponding paths, but the annulus group is still
440: scattered to a lesser degree than would be indicated by the typical abundance
441: ratio uncertainty. Thus, an annulus of inner radius roughly 600 pc centered
442: near the Sun defines a region of enhanced krypton abundance relative to the
443: interstellar mean for the Galactic disk within 6 kpc. See
444: Figure~\ref{krradplot} for an illustration of this region in terms of Kr/H
445: abundance ratios, looking down on the Galactic plane from above.
446:
447: \section{Is the Effect Only Evident for Krypton?}
448: \label{distance_effect}
449: This annulus of krypton abundance enhancement has previously been detected in
450: the abundance data for other elements, although its interpretation in these
451: instances has been different. Two independent compilations of O/H abundance
452: ratios based on the weak $\lambda$1356 intersystem line \citep{and03,car04}
453: noted that this quantity is generally constant except for two effects: enhanced
454: depletion along dense sight lines and an elevation in the ratio for paths
455: longer than about 800 pc and less dense than ${\langle}n_{\rm H}\rangle = 1.0
456: {\rm cm}^{-3}$. \citet{car06} showed that copper to hydrogen abundance ratios
457: also exhibit the latter dependence on a sight line's length. Unfortunately, the
458: database of element-to-hydrogen abundance ratios that they compiled for the
459: other elements being studied (Mg, P, Mn, Ni, and Ge) did not include enough
460: short low-density sight lines to rigorously establish that these elements also
461: conformed to the same trend, and including denser sight lines would have
462: confused any distance effect with the observed depletion enhancement with
463: density. Nevertheless, there was some support for elevated abundance ratios
464: beyond 800 pc in these data as well.
465:
466: The interpretation of the heliocentric distance trend for these studies and
467: the earlier GHRS investigations has been guided by simple spatial arguments.
468: When the first high spectral resolution data from GHRS were compiled, the goal
469: was to explain the apparent constancy of interstellar C, O, N, and Kr
470: abundance levels within 500 pc of the Sun at a common level relative to their
471: Solar values. \citet{mey98} posited three possible explanations: early
472: solar system enrichment by a local supernova; a recent infall of metal-poor gas
473: in the local Milky Way; or an outward diffusion of the Sun from a smaller
474: Galactocentric distance. Each possibility would have the effect of reducing
475: elemental abundances in the Solar neighborhood by a common factor relative to
476: the Sun, provided that radial abundance gradients for these elements were also
477: similar. For the effects observed by both oxygen compilations and
478: \citet{car06}, the abundance reduction is not tied to the Sun, so the simplest
479: mechanism to reduce abundances by a common factor over a large region, and
480: moreover reduce the abundances of several elements by the same factor, is
481: recent infall of metal-poor gas.
482:
483: For other abundance analyses, however, alternate explanations have also been
484: considered. The abundance of nitrogen, ranking along with oxygen as one of the
485: five most abundant elements in the galaxy, also exhibits unexpected variations
486: in the local ISM, for which possible explanations may be nucleosynthetic or
487: related to differential mixing. Specifically, GHRS observations of the weak
488: \ion{N}{1} 1160 {\AA} doublet described a constant N/H ratio within 500 pc of
489: the Sun, with negligible depletion into dust grains \citep{mey97}, but {\it
490: Interstellar Medium Absorption Profile Spectrograph}, STIS, and {\it FUSE} data
491: have altered this picture by enlarging the sample to probe more diverse
492: environments. From these data, \citet{kna03} found that N/H varied
493: significantly as a function of $N$(H) without showing any evidence of N$_2$
494: formation or depletion into dust. Combining nitrogen and oxygen abundances for
495: the most robustly determined of these N column densities, \citet{kna06}
496: determined that locally the N/O ratio is 0.217$\pm$0.011, but that beyond 500
497: pc the ratio drops to 0.142$\pm$0.008 . Eight of the 13 sight lines from which
498: this pattern is deduced are also among those with krypton abundances, and the
499: two paths added in proof to that paper bring the total to ten whose Kr/H ratios
500: can be checked against the distance pattern noted here. Toward each star that
501: was reported by \citet{kna06} as being situated within 500 pc of the Sun, the
502: krypton abundance is in agreement with the local mean [log$_{10}$(Kr/H)$_{ISM}$
503: = $-9.02\pm0.02$]. And furthermore, those paths assigned lengths between 750
504: and 2500 pc agree with log$_{10}$(Kr/H)$_{ISM}$ = $-8.91\pm0.02$, the mean
505: abundance for the annulus of enhancement. In this paper, however, we have
506: adopted a distance of 3.06 kpc for HD 99857 rather than the sub-500 pc {\it
507: Hipparcos} value used by \citet{kna06}. The parallax error on the latter value
508: is of order 80\%, and the distance we adopt is more consistent with the star's
509: spectroscopic parallax and previous literature estimates of its distance. But
510: regardless of whether the path is less than 500 pc or longer than 2.5 kpc, the
511: krypton abundance means in these two regions are in agreement (see
512: Table~\ref{aratmeans}).
513:
514: A similar distinction between local ISM abundances and those determined using
515: paths longer than a few hundred pc has been noted for deuterium, although the
516: transition distance cannot be as cleanly defined. \citet{woo04} compiled D/H
517: measurements from a variety of instruments, finding that both within 100 pc of
518: the Sun and beyond about 600 pc the ratio was well-defined but that the levels
519: were roughly a factor of two (0.30 dex) different. Between 100 and 600 pc,
520: there is a complicated distribution including data spanning a much wider range
521: than is delineated by the near and far average D/H values. Subsequent analyses
522: have expanded upon this description with new data, considering scenarios of
523: infall, variable astration, and density-dependent depletion as explanations for
524: the significant D/H ratio variability among paths between 100 and 600 pc long
525: (e.g., \citealt{oli06,lin06}). However, the possibility of infall has been
526: regarded as doubtful for that situation because it would need to be confined to
527: particular individual sight lines, both since the theoretical ISM mixing
528: timescale of 350 Myr suggests that variabilities are likely to have been
529: removed on these lengthscales in the same time as those within 100 pc of the Sun
530: \citep{dea02} and because empirical evidence suggests that mixing is efficient
531: in this region of the ISM \citep{car06}. Nevertheless, it is apparent that the
532: interstellar deuterium abundance also is subject to anomalies dependent on
533: heliocentric distance, and that the scale of the pattern is similar to that
534: evident in oxygen, copper, and other elements, particularly since the lengthy
535: Galactic sight lines of D/H studies generally would meet the definition of the
536: annulus group in this paper.
537:
538: The spatial extent of the abundance enhancement in heavy elements is made
539: obvious with the current krypton data because measurement of the gas-phase
540: abundance of this element is not subject to complication by depletion---UV
541: absorption profiles permit the {\it total} amount of neutral krypton along a
542: given sight line to be determined. For elements that play a role in
543: interstellar dust, spatial variations in abundance can be rivalled or exceeded
544: in magnitude by those related to density-dependent depletion. Using oxygen as
545: an example, the distinction between the derived warm and cold ISM depletion
546: levels is 0.10 dex, comparable to the location-based abundance enhancement
547: identified here in krypton. This complication made it easy to miss the trend in
548: the \citet{car04,car06} oxygen and copper data; specifically, to note that
549: beyond a distance of order 2.5 kpc, elemental abundances largely return to the
550: lower levels measured within 750 pc of the Sun. In order to confirm the reality
551: of this spatial abundance pattern, though, it is necessary to show that the
552: effect is not just distinguishable in the separate sight line samples for more
553: than one element, but that it is apparent using the same sight lines for each
554: element being considered.
555:
556: Since the distance effect has previously been observed in gas-phase O/H
557: abundance ratios and \ion{O}{1} $\lambda$1356 features are similar in optical
558: depth to \ion{Kr}{1} $\lambda$1236, given their respective interstellar
559: abundances, we measured the oxygen abundance for each sight line presented in
560: this paper. The oxygen measurements were made using the same techniques, with
561: the same underlying profile solutions, applied to the corresponding krypton
562: features and the results are presented in Table~\ref{OHdata}. Note that in
563: cases where the profile is very broad, oxygen abundances are only tallied over
564: radial velocities consistent with the krypton profile, so that O/Kr ratios are
565: representative of the reliable features in each spectrum. Also, the ionization
566: potential of neutral oxygen (13.618 eV) is very close to that of krypton and
567: just larger than the corresponding hydrogen value, ensuring that oxygen and
568: krypton will both be neutral in neutral gas. Care has been taken when plotting
569: the gas-phase O/H ratios in the top panel of Figure~\ref{okrdpplot} to avoid
570: paths with ${\langle}n_{\rm H}\rangle > 1.00 {\rm cm}^{-3}$, above which the
571: data are likely to be contaminated by enhanced oxygen depletion into grains
572: \citep{car04}. Using the same groupings adopted in the analysis of Kr/H
573: variation with pathlength, O/H and O/Kr abundance ratio means for each distance
574: range are given in Table~\ref{aratmeans}; the individual O/Kr abundance ratios
575: are plotted in the bottom panel of Figure~\ref{okrdpplot}. Notably, the same
576: distance pattern identified in Kr/H can easily be discerned in gas-phase O/H
577: ratios using the same sight lines. Numerically speaking, the two abundance
578: ratio increases are also similar: the Kr/H ratio is elevated by 0.11 dex
579: ($\sim4\sigma$, using the quadrature sum of the uncertainties in annulus and
580: non-annulus weighted means) for sight lines ending within the 550---2500 pc
581: annulus relative to the value both inside and outside, while the gas-phase O/H
582: ratio is elevated by 0.08 dex (2.5$\sigma$) in the same manner.
583:
584: Underlining this evidence for a nearly identical heliocentric distance
585: dependence for the krypton and oxygen abundances relative to hydrogen is an
586: extremely tight distribution when the ratio of oxygen to krypton is plotted
587: against pathlength in the bottom panel of Figure~\ref{okrdpplot}. Leaving out
588: sight lines subject to enhanced oxygen depletion, but including those paths
589: with significant krypton detections excluded from the Kr/H analysis for lack of
590: rigorously-determined hydrogen column densities (e.g., HD 94454), the overall
591: weighted mean O/Kr is $5.55\pm0.02$ with a scatter of 0.06 dex. It should be
592: noted that the breadth of this distribution is significantly smaller than what
593: would be expected given the typical abundance ratio uncertainty for these data
594: ($\sim$ 0.09 dex). Similar numbers were derived by \citet{car03} for the Kr/H
595: ratio, and used by \citet{car06} to demonstrate that the ISM within several
596: hundred parsecs of the Sun is very well mixed. Delving further into the details
597: of Table~\ref{aratmeans}, the mean gas-phase O/Kr abundance ratios for paths
598: both entirely within the annulus and those extending well beyond it agree
599: within their errors at a level somewhat larger than the mean for paths in the
600: annulus sample; however, the annulus and non-annulus values agree within the
601: quadrature sum of their 1$\sigma$ errors. Consequently, the abundance
602: enhancement observed in krypton for sight lines between 750 and 2500 pc in
603: length is also proven detectable in oxygen, and at a similar level.
604:
605: Combining these results, it appears that the spatial abundance enhancement is
606: nucleosynthetic in nature. While krypton, oxygen, and copper abundances all
607: rise by a similar amount among paths with a characteristic range of lengths,
608: nitrogen apparently does not. Specifically, we infer that if the O/H ratio is
609: elevated for the annulus sample by 0.08 dex, the N/H ratio for the same sight
610: lines must {\it fall} by 0.10 dex in order to reproduce the mean N/O abundance
611: ratios observed by \citet{kna06}. The scenario suggested by these properties
612: would be described as follows: paths shorter than 550 pc would probe a
613: well-mixed ISM; local ISM gas intermingles with material that has experienced
614: more recent nucleosynthetic activity at distances from the Sun beyond about
615: 500 pc, to the point where this more heavy-element-rich gas dominates in a ring
616: perhaps a few hundred parsecs wide; finally, more than 1000 pc away from the
617: Sun, the ISM appears to again resemble the local ISM in terms of elemental
618: abundance.
619:
620: Such a configuration would give rise to the observed abundance ratio
621: dependencies, given that the total hydrogen column densities for the entire
622: database of sight lines with krypton detections range only from log$_{10}N$(H)
623: = 20.46---21.77 and about 80\% of these are within a factor of 4 of one
624: another, ranging from log$_{10}N$(H) = 21.20---21.77 . Specifically, paths
625: shorter than 550 pc and some with lengths in the 550---750 pc range would be
626: dominated by the well-mixed local ISM. Other sight lines up to 750 pc long
627: would intersect gas dominated by more recently processed material. The paths in
628: this dataset with lengths of 750---2500 pc are exclusively of lower densities
629: (${\langle}n_{\rm H}\rangle < 1 {\rm cm}^{-3}$) and their elemental abundance
630: ratios might readily be dominated by processed gas concentrated in an annulus
631: centered near the Sun whose inner radius is about 500---600 pc. And beyond 2500
632: pc, sight lines with similar overall $N$(H) to shorter paths are on average
633: more rarified and less likely to be dominated by gas in the implied ring, and
634: so evince elemental abundance ratios apparently characteristic of both the
635: well-mixed local gas and more distantly situated material.
636:
637: The suggested annulus of enriched interstellar gas around the Sun is difficult
638: to explain without considering a link to the Gould Belt. The Belt is a
639: disk-like concentration of young stars, OB associations, neutral gas, molecular
640: clouds and dust in a plane intersecting the Galactic disk near the Sun, that is
641: inclined by approximately 20$^\circ$ with respect to the stars in
642: the Galactic disk \citep{tor00}. Its dominant stellar feature is a ring of OB
643: associations bounding this disk; the inner and outer edges of this ring are
644: approximated in Figure~\ref{krradplot} by elliptical annuli whose dimensions
645: and location are set by the \citet{per03} fit with an annulus width of about
646: 80--100 pc. The Gould Belt's origins have generally been ascribed to a
647: combination of explosive events or the impact of a massive high velocity cloud
648: in the Solar neighborhood \citep{com94,pop97}. Regardless of the cause, the
649: Gould Belt includes OB associations formed in the last 30---60 Myr
650: \citep{tor00} at distances up to about 500 pc from the Sun, as well as gas
651: distributed throughout the region bounded by the OB associations in the
652: form of diffuse \ion{H}{1} and molecular clouds \citep{pop97,per03}. The
653: similarity between distances associated with the closer edge of the observed
654: annulus of abundance enhancements and those indicative of the outer reaches of
655: the Gould Belt is very suggestive, given that OB associations such as the ones
656: ringing Gould's Belt produce Type II supernovae. \ion{SN}{2} are important
657: production sites for krypton, oxygen, and copper that leave nitrogen
658: essentially unenhanced (e.g., \citealt{chi04}). In fact, the supernova rate for
659: the Gould Belt has been estimated at 3--5 times that of the local Galactic disk
660: \citep{gre00}. The observed outward flow of gas away from the Belt's central
661: regions \citep{per03} is consistent with both an explosive origin for the Belt
662: or the impact of a high velocity cloud \citep{com94}. The flow is also patchy,
663: which is consistent with the pattern of abundance enhancements identified in
664: this paper. Since the Belt was formed only a few tens of millions of years ago,
665: there has not been sufficient time for mixing to erase any abundance variations
666: introduced into interstellar gas by recent supernovae in the Belt.
667:
668: It is worthwhile to note that analyses of abundances based on weak UV
669: absorption lines can be influenced by the selection effect alluded to in the
670: last few paragraphs. If the databases are restricted only to spectra with
671: absorption features strong enough to be detected but weak enough to avoid
672: concerns of unresolved saturation, and also to observations of stars bright
673: enough for the exposure time to be relatively short, then the analyses will
674: tend to be limited in the range of total hydrogen column density being sampled.
675: Consequently, as longer sight lines are included, more rarefied gas is being
676: probed. This tends to restrict the study of depletion variations to the local
677: ISM. In order to fully explore current Galactic elemental abundances and
678: depletion levels with a revitalized STIS and new Cosmic Origins Spectrograph,
679: as well as search for features such as the spatial abundance pattern noted in
680: this paper or other effects at large distances, it will be important to make
681: long exposure time observations of distant stars an important part of the
682: analysis.
683:
684: \section{Conclusions}
685: This paper has combined new observations from an {\it HST} observing program
686: abbreviated by the STIS power failure with Archival detections and previously
687: published measurements of \ion{Kr}{1} $\lambda$1236 to compile a database of
688: 50 Kr/H abundance ratios for sight lines up to 5.9 kpc in length and with
689: average hydrogen densities from less than 0.10 cm$^{-3}$ to more than 10.0
690: cm$^{-3}$. Collectively, the data strengthen earlier evidence that krypton is
691: undepleted in the ISM and that the Kr/H ratio is remarkably uniform within 750
692: pc of the Sun. Furthermore, these two properties underline the significance of
693: an apparent enhancement of 0.10 dex in krypton abundance for sight lines of
694: length 750---2500 pc. This abundance elevation is matched by enhanced O/H
695: abundance ratios for the same sight lines, and similar distance dependencies in
696: both gas-phase oxygen and copper abundances among other sight lines. However,
697: N/O abundance ratios are reduced by a factor of two thirds in the same annulus.
698: These disparate pieces of evidence suggest a nucleosynthetic origin for the
699: abundance variations, while their spatial extent is comparable to the size of
700: the Gould Belt. It is possible that the noted elemental abundance variations
701: for Kr, O, Cu, and N arise out of star formation engendered as the Gould Belt
702: was formed. Given the spatial extent of D/H variations, the Gould Belt may be
703: at the bottom of this trend as well.
704:
705: \acknowledgments
706: We would like to thank Paul Cartledge for advice on statistical tests and the
707: anonymous referee for their comments, which have significantly improved the
708: quality of this paper. Support for this work was provided by the Space
709: Telescope Science Institute through grants to Louisiana State University and
710: Northwestern University (both GO9855 and AR10643). This research has made use
711: of the SIMBAD database, operated at CDS (Strasborg France).
712:
713: {\it Facilities:} \facility{FUSE (LWRS,MDRS,HIRS)}, \facility{HST (STIS)}.
714:
715: \begin{thebibliography}{}
716: %\bibitem[Anders \& Grevesse(1989)]{and89} Anders, E., \& Werner, M. W. 1989,
717: % Geochim. Cosmochim. Acta, 53, 197
718: \bibitem[Andr\'e et al.(2003)]{and03} Andr\'e, M. K., Oliveira, C. M., Howk, J.
719: C., Ferlet, R., D\'esert, J.-M., H\'ebrard, G., Lacour, S., Lecavelier des
720: \'Etangs, A., Vidal-Madjar, A., Moos, H. W. 2003, \apj, 591, 1000
721: \bibitem[Cardelli et al.(1991)]{car91} Cardelli, J. A., Savage, B.
722: D., \& Ebbets, D. C. 1991, \apjl, 393, L23
723: \bibitem[Cardelli \& Meyer(1997)]{car97} Cardelli, J. A., \& Meyer, D. M. 1997,
724: \apjl, 477, L57
725: \bibitem[Cartledge et al.(2004)]{car04} Cartledge, S. I. B., Lauroesch, J. T.,
726: Meyer, D. M., \& Sofia, U. J. 2004, \apj, 613, 1037
727: \bibitem[Cartledge et al.(2006)]{car06} Cartledge, S. I. B., Lauroesch, J. T.,
728: Meyer, D. M., \& Sofia, U. J. 2006, \apj, 641, 327
729: \bibitem[Cartledge et al.(2003)]{car03} Cartledge, S. I. B., Meyer, D. M., \&
730: Lauroesch, J. T. 2003, \apj, 597, 408
731: \bibitem[Cartledge et al.(2001)]{car01} Cartledge, S. I. B., Meyer, D. M.,
732: Lauroesch, J. T., \& Sofia, U. J. 2001, \apj, 562, 394
733: %\bibitem[Comer\'on \& Torra(1992)]{com92} Comer\'on, F., \& Torra, J. 1992,
734: % \aap, 261, 94
735: \bibitem[Chieffi \& Limongi(2004)]{chi04} Chieffi, A., \& Limongi, M. 2004,
736: \apj, 608, 405
737: \bibitem[Comer\'on \& Torra(1994)]{com94} Comer\'on, F., \& Torra, J. 1994,
738: \aap, 281, 35
739: %\bibitem[Comer\'on, Torra, \& G\'omez(1992)]{com92b} Comer\'on, F., Torra, J.,
740: % \& G\'omez, A. E. 1992, \apss, 187, 187
741: \bibitem[de Avillez \& Mac Low(2002)]{dea02} de Avillez, M. A., \& Mac Low,
742: M.-M. 2002, \apj, 581, 1047
743: \bibitem[Grenier(2000)]{gre00} Grenier, I. A. 2000, \aap, 364, L93
744: \bibitem[Howk \& Sembach(2000)]{how00} Howk, J. C., \& Sembach, K. R. 2000,
745: \aj, 119, 2481
746: \bibitem[Knauth et al.(2003)]{kna03} Knauth, D. C., Andersson, B.-G.,
747: McCandliss, S. R., \& Moos, H. W. 2003, \apjl, 596, L51
748: \bibitem[Knauth at al.(2006)]{kna06} Knauth, D. C., Meyer, D. M. \& Lauroesch,
749: J. T. 2006, \apjl, 647, L115
750: \bibitem[Lauroesch et al.(2008)]{lau08} Lauroesch, J. T., Cartledge, S. I. B.,
751: Clayton, G. C., Sofia, U. J., \& Meyer, D. M. 2008, \apjl, {\it submitted}
752: \bibitem[Linsky et al.(2006)]{lin06} Linsky, J. L., Draine, B. T., Moos, H. W.,
753: Jenkins, E. B., Wood, B. E., Oliveira, C. M., Blair, W. P., Friedman, S.
754: D., Gry, C., Knauth, D., Kruk, J. W., Lacour, S., Lehner, N., Redfield, S.,
755: Shull, J. M., Sonneborn, G., \& Williger, G. M. 2006, \apj, 647, 1106
756: \bibitem[Lodders(2003)]{lod03} Lodders, K. 2003, \apj, 591, 1220
757: \bibitem[Martin-Hern\'andez et al.(2002)]{mar02} Martin-Hern\'andez, N. L.,
758: Peeters, E., Morisset, C., Tielens, A. G. G. M., Cox, P., Roelfsema, P.
759: R., Baluteau, J. -P., Schaerer, D., Mathis, J. S., Damour, F., Churchwell,
760: E., \& Kessler, M. F. 2002, \aap, 381, 606
761: \bibitem[Meyer et al.(1997)]{mey97} Meyer, D. M., Cardelli, J. A., \& Sofia, U.
762: J. 1997, \apj, 490, L103
763: \bibitem[Meyer et al.(1998)]{mey98} Meyer, D. M., Jura, M., \& Cardelli, J. A.
764: 1998, \apj, 493, 222
765: \bibitem[Nittler(2005)]{nit05} Nittler, L. R. 2005, \apj, 618, 281
766: \bibitem[Oliveira et al.(2006)]{oli06} Oliveira, C. M., Moos, H. W., Chayer,
767: P., \& Kruk, J. W. 2006, \apj, 642, 283
768: \bibitem[Perrot \& Grenier(2003)]{per03} Perrot, C. A., \& Grenier, I. A. 2003,
769: \aap, 404, 519
770: \bibitem[P\"oppel(1997)]{pop97} P\"oppel, W. G. L. 1997, Fundamentals of Cosmic
771: Physics, 18, 1
772: \bibitem[Rachford et al.(2002)]{rac02} Rachford, B. L., Snow, T. P., Tumlinson,
773: J., Shull, J. M., Blair, W. P., Ferlet, R., Friedman, S. D., Gry, C.,
774: Jenkins, E. B., Morton, D. C., Savage, B. D., Sonnentrucker, P.,
775: Vidal-Madjar, A., Welty, D. E., \& York, D. G. 2002, \apj, 577, 221
776: \bibitem[Reddy et al.(2003)]{red03} Reddy, B. E., Tomkin, J., Lambert, D. L.,
777: \& Allende Prieto, C. 2003, \mnras, 340, 304
778: \bibitem[Rolleston et al.(2000)]{rol00} Rolleston, W. R. J., Smartt, S. J.,
779: Dufton, P. L., \& Ryans, R. S. I. 2000, \aap, 363, 537
780: \bibitem[Spitzer(1978)]{spi78} Spitzer, L. 1978, \underline{Physical Processes
781: in the Interstellar Medium} (New York: John Wikey \& Sons)
782: \bibitem[Torra et al.(2000)]{tor00} Torra, J., Fern\'andez, D., \& Figueras, F.
783: 2000, \aap, 359, 82
784: \bibitem[Vall\'ee(2002)]{val02} Vall\'ee, J. P. 2002, \apj, 566, 261
785: \bibitem[Wood et al.(2004)]{woo04} Wood, B. E., Linsky, J. L., H\'ebrard, G.,
786: Williger, G. M., Moos, H. W., \& Blair, W. P. 2004, \apj, 609, 838
787: \end{thebibliography}
788:
789: \clearpage
790:
791: \begin{figure}
792: \epsscale{1.0}
793: \plotone{f1.eps}
794: \caption{Density-dependent gas-phase interstellar krypton abundances. New
795: krypton-to-hydrogen abundance ratios for 26 sight lines overlap
796: previously-published GHRS and STIS measurements for 22 other paths, when
797: expressed as a function of mean hydrogen sight line density. The new data are
798: scattered more widely with respect to the ISM weighted mean of
799: log$_{10}$(Kr/H) = $-9.00\pm0.02$ than were the previous data to their mean;
800: the two weighted averages agree within their respective errors. The new and old
801: [log$_{10}$(Kr/H) = $-9.02\pm0.02$; \citealt{car03}] interstellar Kr/H
802: abundance ratios are indicated by the solid and dashed lines, respectively; the
803: dotted line is the \citet{lod03} Solar value [log$_{10}$(Kr/H) =
804: $-8.72\pm0.08$].
805: \label{krnhplot}
806: }
807: \end{figure}
808:
809: \begin{figure}
810: \epsscale{0.9}
811: \plotone{f2.eps}
812: \caption{Krypton abundance and molecular hydrogen fraction. The gas-phase Kr/H
813: abundance ratio is shown as a function of molecular hydrogen fraction in the
814: top panel and its base-ten logarithm in the bottom panel. It is clear in the
815: top panel that from low to high values of $f$(H$_2$), the data are consistent
816: with a uniform Kr/H ratio; the bottom panel emphasizes that this uniformity
817: extends across 5 orders of magnitude to very low values of $f$(H$_2$). These
818: plots confirm the hypothesis that krypton is undepleted in the diffuse ISM.
819: \label{krh2plot}
820: }
821: \end{figure}
822:
823: \begin{figure}
824: \epsscale{1.0}
825: \plotone{f3.eps}
826: \caption{Krypton abundance and Galactocentric radius. The gas-phase Kr/H
827: abundance ratio does not exhibit a strong variation as a function of the target
828: star's distance from the Galactic centre; the mean slope is $-0.013\pm0.014$
829: dex/kpc when a linear function is fit to the data (the dashed line). This value
830: suggests a radial abundance gradient of order $-$0.027 dex/kpc under the crude
831: assumption that on average the gas probed in each sight line is situated midway
832: along its pathlength. This trend is not very visually compelling, however,
833: particularly since the datapoints within 8.5 kpc of the Galactic center that
834: have the two largest Kr/H values are subject to the noted heliocentric distance
835: trend.
836: \label{krrgplot}
837: }
838: \end{figure}
839:
840: \begin{figure}
841: \epsscale{1.0}
842: \plotone{f4.eps}
843: \caption{Distance-dependent gas-phase interstellar krypton abundances. The
844: krypton-to-hydrogen abundance ratios for all 48 sight lines are plotted as a
845: function of the distance to the target star. The new data include many distant
846: paths, separating the sample into paths with low Kr/H ratios within 750 pc
847: [weighted mean log$_{10}$(Kr/H) = $-9.02\pm0.02$], those with high Kr/H ratios
848: and lengths between 550 and 2500 pc [weighted mean log$_{10}$(Kr/H) =
849: $-8.91\pm0.02$], and several longer than 2.5 kpc whose Kr/H ratios return to
850: values consistent with those in the local ISM [weighted mean log$_{10}$(Kr/H) =
851: $-9.03\pm0.04$]. The weighted mean Kr/H value in each circular region are
852: indicated by the dot-dashed lines.
853: \label{krdpplot}
854: }
855: \end{figure}
856:
857: \begin{figure}
858: \epsscale{1.0}
859: \plotone{f5.eps}
860: \caption{The Galactic plane distribution of gas-phase interstellar krypton
861: gas-phase abundances. The krypton sight line endpoints are identified by
862: symbols indicating the corresponding Kr/H abundance ratio relative to the
863: interstellar mean. The solid curves indicate spiral arm solutions \citep{val02}
864: and the dotted annuli specify heliocentric distance in kiloparsecs. In this
865: plot, the area near the Sun is dominated by low abundance ratios, surrounded by
866: a rough annulus of higher than average ratios. On average, sight lines longer
867: than 2.5 kpc return to lower abundances. The dot-dashed circle marks the outer
868: edge of the annulus of elevated Kr/H ratios we have noted; the solid ovals near
869: the center of the plot indicate the area bounded by projection of the Gould
870: Belt onto the Galactic plane, roughly identifying both the inner and outer
871: dimensions \citep{per03}. In the figure legend, $x$ refers to log$_{10}$(Kr/H).
872: \label{krradplot}
873: }
874: \end{figure}
875:
876: \begin{figure}
877: \epsscale{0.7}
878: \plotone{f6.eps}
879: \caption{Distance-dependent gas-phase interstellar oxygen abundances along
880: paths with ${\langle}n_{\rm H}\rangle < 1.0 {\rm cm}^{-3}$. Gas-phase oxygen
881: abundances for the current krypton sample are presented above as a function of
882: sight line pathlength. In the top panel, the data are displayed as O/H ratios;
883: an enhancement is clearly evident for paths between 750 and 2500 pc in length,
884: mirroring the trend in Kr/H abundance ratios from Figure~\ref{krdpplot}. This
885: trend is approximately equal in strength for krypton and oxygen, as indicated
886: by the extremely tight distribution of O/Kr abundance ratios in the bottom
887: panel. Note that only low-density sight lines [${\langle}n_{\rm H}\rangle
888: < 1.0 {\rm cm}^{-3}$] are used for these plots, to avoid confusing the clarity
889: of the effect by including paths affected by oxygen depletion \citep{car04}.
890: Numerical values for the ISM O/H means are found in Table~\ref{aratmeans}; the
891: Solar ratios are from \citet{lod03}.
892: \label{okrdpplot}
893: }
894: \end{figure}
895:
896: \clearpage
897:
898: \begin{deluxetable}{lccccccc}
899: %\tabletypesize{\scriptsize}
900: %\rotate
901: \tablecaption{Summary of Observations \label{datalist}}
902: \tablewidth{0pt}
903: \tablehead{
904: &\colhead{\it HST}&\colhead{\it STIS}&&\colhead{Time}&\colhead{{\it FUSE}}&
905: &\colhead{Time}\\
906: Sight Line&\colhead{Program}&\colhead{Data Set}&\colhead{Date}&\colhead{(s)}&
907: \colhead{Data Set}&\colhead{Date}&\colhead{(s)}\\
908: }
909: \startdata
910: HD 23478 & GO9505 & O6LJ01020 & 2002 Aug 20 & 2945 & A1200202000 & 2001 Jan 02 &
911: 1871 \\
912: HD 24190 & GO9505 & O6LJ02020 & 2003 Mar 28 & 2940 & A1200101000 & 2000 Sep 11 &
913: 4721 \\
914: %HD 36408& GO9819 & O8MM04020 & 2004 Mar 09 & 2738 & P1190404000 & 2001 Mar 06 &
915: % 4832 \\
916: % &\nodata & \nodata & \nodata &\nodata& P1190405000 & 2004 Oct 18 &
917: % 5596 \\
918: HD 40893 & GO9855 & O8NA02010 & 2004 Feb 20 & 2162 & P2160101000 & 2001 Oct 14 &
919: 7138 \\
920: & & O8NA02020 & 2004 Feb 20 & 7902 & \nodata & \nodata &
921: \nodata \\
922: HD 69106 & GO8484 & O5LH03010 & 2000 Jun 08 & 660 & P1022301000 & 2001 Apr 04 &
923: 58 \\
924: & & O5LH03020 & 2000 Jun 08 & 625 & P1022302000 & 2002 Dec 31 &
925: 404 \\
926: HD 94454 & GO9505 & O6LJ0H010 & 2002 Dec 17 & 1780 & A1200501000 & 2000 Apr 04 &
927: 7622 \\
928: HD 99857 & GO8043 & O54301010 & 1999 Feb 20 & 1307 & P1024501000 & 2000 Feb 05 &
929: 4267 \\
930: & & O54301020 & 1999 Feb 20 & 1307 & \nodata & \nodata &
931: \nodata \\
932: & & O54301030 & 1999 Feb 20 & 1307 & \nodata & \nodata &
933: \nodata \\
934: HD 99872 & GO9505 & O6LJ0I020 & 2002 May 06 & 1890 & A1200606000 & 2000 Apr 09 &
935: 3270 \\
936: HD 102065& GO7509 & O4O001010 & 1998 Mar 28 & 7679 & Q1010101000 & 2000 May 28 &
937: 6793 \\
938: HD 104705& GO7270 & O57R01010 & 1998 Dec 24 & 2400 & P1025701000 & 2000 Feb 05 &
939: 4482 \\
940: HD 108639& GO9505 & O6LJ0A020 & 2002 Oct 19 & 1860 & A1201301000 & 2000 Mar 20 &
941: 4980 \\
942: HD 112999& GO9505 & O6LJ0C010 & 2002 Oct 21 & 1620 & A1202001000 & 2000 May 15 &
943: 7544 \\
944: & & O6LJ0C020 & 2002 Oct 21 & 1650 & \nodata & \nodata &
945: \nodata \\
946: HD 114886& GO9505 & O6LJ0D010 & 2003 Apr 03 & 1560 & A1201801000 & 2000 Apr 06 &
947: 4450 \\
948: & & O6LJ0D020 & 2003 Apr 03 & 1590 & \nodata & \nodata &
949: \nodata \\
950: HD 115071& GO9505 & O6LJ0E010 & 2003 Mar 31 & 1890 & P1025901000 & 2000 May 14 &
951: 5796 \\
952: & & O6LJ0E020 & 2003 Mar 31 & 1890 & \nodata & \nodata &
953: \nodata \\
954: HD 116852& GO8241 & O5C01C010 & 2000 Jun 25 & 360 & P1013801000 & 2000 May 27 &
955: 7212 \\
956: & GO9855 & O8NA03010 & 2003 Aug 08 & 2522 & \nodata & \nodata &
957: \nodata \\
958: & & O8NA03020 & 2003 Aug 08 & 3200 & \nodata & \nodata &
959: \nodata \\
960: %HD 121968& GO7270 & O57R02010 & 1999 Feb 18 & 1558 & P1014501001 & 2001 Mar 09&
961: % 9258 \\
962: % & & O57R02020 & 1999 Feb 18 & 2899 & \nodata & \nodata &
963: %\nodata \\
964: HD 122879& GO8241 & O5C037010 & 2000 May 08 & 360 & B0710501000 & 2002 Mar 03 &
965: 1532 \\
966: & GO8484 & O5LH07010 & 2000 Aug 29 & 708 & \nodata & \nodata &
967: \nodata \\
968: & & O5LH07020 & 2000 Aug 29 & 708 & \nodata & \nodata &
969: \nodata \\
970: HD 124314& GO8043 & O54307010 & 1999 Apr 10 & 1466 & P1026201000 & 2000 Mar 22 &
971: 4382 \\
972: HD 137595& GO9505 & O6LJ03010 & 2002 Aug 21 & 1200 & A1201213000 & 2000 Jul 10 &
973: 5715 \\
974: HD 144965& GO9505 & O6LJ05010 & 2003 Feb 01 & 1510 & A1201601000 & 2000 Mar 21 &
975: 1195 \\
976: &\nodata & \nodata & \nodata &\nodata& A1201602000 & 2001 Aug 08 &
977: 7220 \\
978: HD 147165& GO5891 & Z34R0407T & 1996 Mar 23 & 108.8& \nodata & \nodata &
979: \nodata \\
980: HD 147683& GO9505 & O6LJ06020 & 2003 Mar 27 & 2940 & A1200909000 & 2000 Jul 10 &
981: 7367 \\
982: HD 151805& GO9434 & O6LZ63010 & 2003 Apr 03 & 1200 & P1026602001 & 2000 Jul 14 &
983: 2741 \\
984: HD 152590& GO8241 & O5C08P010 & 2000 Mar 03 & 1440 & B0710601000 & 2001 Jul 08 &
985: 1389 \\
986: & GO9855 & O8NA04010 & 2003 Sep 17 & 2240 & B0710602000 & 2001 Aug 12 &
987: 2773 \\
988: & & O8NA04020 & 2003 Sep 17 & 8076 & \nodata & \nodata &
989: \nodata \\
990: HD 165246& GO9855 & O8NA05010 & 2003 Aug 13 & 2158 & P1050301000 & 2000 Sep 01 &
991: 7782 \\
992: & & O8NA05020 & 2003 Aug 13 & 2626 & \nodata & \nodata &
993: \nodata \\
994: HD 177989& GO7270 & O57R03010 & 1999 May 28 & 1660 & P1017101000 & 2000 Aug 28 &
995: 10289 \\
996: & & O57R03020 & 1999 May 28 & 2897 & \nodata & \nodata &
997: \nodata \\
998: HD 203374& GO848 & O5LH08010 & 2000 Mar 23 & 1325 & B0300101000 & 2001 Aug 02 &
999: 10170 \\
1000: & & O5LH08020 & 2000 Mar 23 & 1325 & B0300102000 & 2001 Aug 02 &
1001: 8483 \\
1002: & & O5LH08030 & 2000 Mar 23 & 1325 & \nodata & \nodata &
1003: \nodata \\
1004: HD 206267& GO8484 & O5LH09010 & 2000 Mar 24 & 711 & B0460301000 & 2002 Jun 30 &
1005: 4869 \\
1006: & & O5LH09020 & 2000 Mar 24 & 711 & \nodata & \nodata &
1007: \nodata \\
1008: HD 208947& GO8484 & O5LH0A010 & 2000 Jul 14 & 714 & \nodata & \nodata &
1009: \nodata \\
1010: & & O5LH0A020 & 2000 Jul 14 & 714 & \nodata & \nodata &
1011: \nodata \\
1012: HD 209339& GO8484 & O5LH0B010 & 2000 May 18 & 708 & B0300501000 & 2001 Jul 21 &
1013: 2044 \\
1014: & & O5LH0B020 & 2000 May 18 & 708 & \nodata & \nodata &
1015: \nodata \\
1016: HD 224151& GO8043 & O54308010 & 1999 Feb 18 & 1496 & P1224101000 & 2000 Aug 11 &
1017: 6023 \\
1018: &\nodata & \nodata & \nodata &\nodata& S3040201000 & 1999 Dec 09 &
1019: 7092 \\
1020: &\nodata & \nodata & \nodata &\nodata& S3040202000 & 2000 Aug 13 &
1021: 9202 \\
1022: HD 303308& GO7301 & O4QX04010 & 1998 Mar 19 & 2220 & P1221601000 & 2000 May 25 &
1023: 6105 \\
1024: &\nodata & \nodata & \nodata &\nodata& P1221602000 & 2000 May 27 &
1025: 7692 \\
1026: %CPD $-$59 2603& GO7301 & O4QX03010 & 1999 Apr 19 & 5160 & P1221501000 & 2000 Mar
1027: %21 &5598\\
1028: % &\nodata & \nodata & \nodata &\nodata& S3040501000 & 2000 Jan 30 &
1029: %38804 \\
1030: % &\nodata & \nodata & \nodata &\nodata& S3040502000 & 2000 Apr 06 &
1031: %32282 \\
1032: \enddata
1033: %\tablenotetext{a}{All column densities are expressed in logarithmic form for
1034: %absolute values in cm$^{-2}$.}
1035: \end{deluxetable}
1036:
1037: \clearpage
1038:
1039: \begin{deluxetable}{lr@{ }lr@{ }lr@{ }lr@{ }lr@{ }lr@{ }llcr@{.}lc}
1040: \rotate
1041: \tabletypesize{\footnotesize}
1042: \tablecaption{New Krypton and Hydrogen Sight Line Properties \label{KrHdata}}
1043: \tablewidth{0pt}
1044: \tablehead{
1045: &\multicolumn{2}{c}{$W_{\lambda1235}$}&\multicolumn{2}{c}{log$_{10}$[$N$(Kr)]}&
1046: \multicolumn{2}{c}{log$_{10}$[$N$(H)]}&
1047: \multicolumn{2}{c}{log$_{10}$[$N$(\ion{H}{1})]}&
1048: \multicolumn{2}{c}{log$_{10}$[$N$(H$_2$)]}&
1049: \multicolumn{2}{c}{ }&\colhead{Spectral}&\colhead{$d_\ast$}&
1050: \multicolumn{2}{c}{log$_{10}{\langle}n_{\rm H}\rangle$}&\colhead{ }\\
1051: \colhead{Star}&\multicolumn{2}{c}{(m\AA)}&\multicolumn{2}{c}{[cm$^{-2}$]}&
1052: \multicolumn{2}{c}{[cm$^{-2}$]}&\multicolumn{2}{c}{[cm$^{-2}$]}&
1053: \multicolumn{2}{c}{[cm$^{-2}$]}&\multicolumn{2}{c}{log$_{10}$[Kr/H]}&
1054: \colhead{Type}&\colhead{(kpc)}&\multicolumn{2}{c}{[cm$^{-3}$]}&
1055: \colhead{log$_{10}f$(H$_2$)}\\
1056: }
1057: \startdata
1058: HD 23478\tablenotemark{a}&4.16&(0.33)&12.23&(0.05)&21.21&(0.04)&21.01&(0.06)&
1059: 20.48 & (0.07) & $-$8.98 &$^{+0.06}_{-0.07}$& B3IV &0.32& 0&22 & $-$0.43 \\
1060: HD 24190 & 5.87 & (0.72) & 12.35 & (0.06) & 21.30 & (0.05) & 21.18 &(0.06)&
1061: 20.38 & (0.07) & $-$8.95 &$^{+0.08}_{-0.09}$& B2V &0.32& 0&31 & $-$0.62 \\
1062: %HD 36408\tablenotemark{b}&4.51&(0.73)&12.25&(0.09)&\multicolumn{2}{c}{\nodata}&
1063: %\multicolumn{2}{c}{\nodata}&19.12&(0.10)&\multicolumn{2}{c}{\nodata}&
1064: %B7IIIe & 0.22 &\multicolumn{2}{c}{\nodata}&\nodata\\
1065: HD 40893 & 9.52 & (1.35) & 12.56 & (0.07) & 21.54 & (0.05) & 21.45 &(0.09)&
1066: 20.49 & (0.07) & $-$8.98 &$^{+0.08}_{-0.10}$& B0IV &2.80& $-$0&40 & $-$0.75 \\
1067: HD 69106 & 4.06 & (0.76) & 12.19 & (0.09) & 21.09 & (0.06) & 21.06 &(0.06)&
1068: 19.64 & (0.07) & $-$8.90 &$^{+0.11}_{-0.14}$& B0.5II &1.40& $-$0&55 & $-$1.15 \\
1069: HD 94454\tablenotemark{b}&5.37&(0.85)&12.36&(0.08)&\multicolumn{2}{c}{\nodata}&
1070: \multicolumn{2}{c}{\nodata}&20.70&(0.07)&\multicolumn{2}{c}{\nodata}&
1071: B8III & 0.33 &\multicolumn{2}{c}{\nodata}&\nodata\\
1072: HD 99857 & 5.47 & (0.94) & 12.34 & (0.08) & 21.36 & (0.05) & 21.28 &(0.06)&
1073: 20.29 & (0.07) & $-$9.02 &$^{+0.09}_{-0.12}$& B1Ib &3.06& $-$0&61 & $-$0.77 \\
1074: HD 99872\tablenotemark{b}&4.27&(0.47)&12.25&(0.07)&\multicolumn{2}{c}{\nodata}&
1075: \multicolumn{2}{c}{\nodata}&20.55&(0.11)&\multicolumn{2}{c}{\nodata}&
1076: B3V & 0.23 &\multicolumn{2}{c}{\nodata}&\nodata\\
1077: HD 102065\tablenotemark{b}&3.76&(0.52)&12.16&(0.11)&
1078: \multicolumn{2}{c}{\nodata}&\multicolumn{2}{c}{\nodata}&20.47&(0.08)&
1079: \multicolumn{2}{c}{\nodata}&B2V & 0.17 &\multicolumn{2}{c}{\nodata}&\nodata\\
1080: HD 104705 & 2.99 & (0.54) & 12.05 & (0.09) & 21.20 & (0.05) & 21.15 &(0.06)&
1081: 19.94 & (0.06) & $-$9.15 &$^{+0.10}_{-0.13}$&B0III/IV&3.90& $-$0&88 & $-$0.96 \\
1082: HD 108639 & 6.72 & (1.34) & 12.40 & (0.10) & 21.38 & (0.08) & 21.35 &(0.09)&
1083: 19.95 & (0.07) & $-$8.98 &$^{+0.12}_{-0.17}$& B1III &0.11& 0&84 & $-$1.13 \\
1084: HD 112999\tablenotemark{b}&4.50&(0.65)&12.27&(0.09)&\multicolumn{2}{c}{\nodata}&
1085: \multicolumn{2}{c}{\nodata}&19.99&(0.07)&\multicolumn{2}{c}{\nodata}&
1086: B6IIIn & 0.34 &\multicolumn{2}{c}{\nodata}&\nodata\\
1087: HD 114886 & 8.31 & (2.01) & 12.43 & (0.14) & 21.40 & (0.05) & 21.34 &(0.06)&
1088: 20.23 & (0.07) & $-$8.97 &$^{+0.15}_{-0.22}$& O9V &0.38& 0&32 & $-$0.87 \\
1089: HD 115071 &10.77 & (1.56) & 12.64 & (0.07) & 21.50 & (0.05) & 21.36 &(0.06)&
1090: 20.63 & (0.07) & $-$8.86 &$^{+0.08}_{-0.10}$& O9.5V &1.64& $-$0&20 & $-$0.57 \\
1091: HD 116852 & 3.02 & (0.38) & 12.06 & (0.06) & 21.03 & (0.06) & 20.98 &(0.06)&
1092: 19.79 & (0.11) & $-$8.97 &$^{+0.08}_{-0.10}$& O9III &4.80& $-$1&14 & $-$0.94 \\
1093: %HD 121968\tablenotemark{d}& 3.77 & (0.49) & 11.89 & (0.31) & 20.75 & (0.06) &
1094: %20.75 &(0.06)& 16.28 &(0.05)& $-$8.86 &$^{+0.31}_{unf}&3.62& $-$1&30 &$-$4.17\\
1095: HD 122879 & 9.27 & (1.26) & 12.60 & (0.13) & 21.35 & (0.10) & 21.26 &(0.12)&
1096: 20.31 & (0.08) & $-$8.75 &$^{+0.16}_{-0.25}$& B0Ia &2.10& $-$0&46 & $-$0.74 \\
1097: HD 124314 &12.72 & (2.50) & 12.70 & (0.09) & 21.47 & (0.05) & 21.39 &(0.06)&
1098: 20.40 & (0.07) & $-$8.77 &$^{+0.10}_{-0.13}$& O7 &1.15& $-$0&08 & $-$0.77 \\
1099: HD 137595\tablenotemark{a}&5.50&(0.63)&12.34&(0.07)&21.24&(0.04)&21.00&(0.06)&
1100: 20.56 & (0.06) & $-$8.90 &$^{+0.08}_{-0.10}$& B3Vn &0.83& $-$0&17 & $-$0.38 \\
1101: HD 144965\tablenotemark{a}&4.79&(0.70)&12.30&(0.08)&21.37&(0.05)&21.07&(0.06)&
1102: 20.77 & (0.07) & $-$9.07 &$^{+0.09}_{-0.12}$& B3Vne &0.68& 0&05 & $-$0.30 \\
1103: HD 147165 & 6.71 & (0.41) & 12.45 & (0.05) & 21.40 & (0.08) & 21.38 &(0.08)&
1104: 19.79 & (0.06) & $-$8.95 &$^{+0.09}_{-0.12}$& B1III &0.15& 0&75 & $-$1.31 \\
1105: HD 147683\tablenotemark{a}& 8.32 & (1.03) & 12.52 & (0.06) & 21.55 & (0.07) &
1106: 21.41 &(0.08)& 20.68 & (0.12) & $-$9.03 &$^{+0.09}_{-0.11}$& B4V & 0.59 &
1107: 0&29 & $-$0.57 \\
1108: HD 151805 & 6.54 & (1.30) & 12.41 & (0.08) & 21.41 & (0.05) & 21.32 &(0.06)&
1109: 20.36 & (0.07) & $-$9.00 &$^{+0.13}_{-0.18}$& B1Ib &5.91& $-$0&85 & $-$0.75 \\
1110: HD 152590 & 8.49 & (1.05) & 12.54 & (0.05) & 21.47 & (0.05) & 21.37 &(0.06)&
1111: 20.47 & (0.07) & $-$8.93 &$^{+0.07}_{-0.08}$& O7.5V &1.80& $-$0&30 & $-$0.70 \\
1112: HD 165246 & 7.95 & (0.78) & 12.53 & (0.06) & 21.46 & (0.06) & 21.41 &(0.07)&
1113: 20.15 & (0.08) & $-$8.93 &$^{+0.08}_{-0.10}$& O9III &1.59& $-$0&23 & $-$1.01 \\
1114: HD 177989 & 4.70 & (0.72) & 12.25 & (0.07) & 21.09 & (0.05) & 20.98 &(0.06)&
1115: 20.16 & (0.08) & $-$8.84 &$^{+0.08}_{-0.10}$& B2II &4.91& $-$1&09 & $-$0.63 \\
1116: HD 203374 & 9.36 & (0.95) & 12.56 & (0.06) & 21.38 & (0.06) & 21.20 &(0.08)&
1117: 20.60 & (0.09) & $-$8.82 &$^{+0.08}_{-0.10}$& B0IVpe &0.62& 0&10 & $-$0.48 \\
1118: HD 206267 &11.16 & (1.47) & 12.65 & (0.10) & 21.46 & (0.04) & 21.20 &(0.06)&
1119: 20.81 & (0.05) & $-$8.81 &$^{+0.11}_{-0.14}$& O6e &0.59& 0&20 & $-$0.35 \\
1120: HD 208947\tablenotemark{c}&2.53&(0.40)&11.99&(0.08)&21.10&(0.06)&21.10&(0.06)&
1121: \multicolumn{2}{c}{\nodata}& $-$9.11 &$^{+0.09}_{-0.12}$& B2V & 0.50 &
1122: $\geq-$0&08 & \nodata\\
1123: HD 209339 & 5.81 & (0.55) & 12.35 & (0.06) & 21.26 & (0.06) & 21.17 &(0.07)&
1124: 20.24 & (0.08) & $-$8.91 &$^{+0.08}_{-0.10}$& B0IV &1.00& $-$0&23 & $-$0.72 \\
1125: HD 224151 & 9.87 & (2.82) & 12.58 & (0.12) & 21.43 & (0.05) & 21.30 &(0.06)&
1126: 20.53 & (0.06) & $-$8.85 &$^{+0.13}_{-0.18}$& B0.5II &1.10& $-$0&10 & $-$0.60 \\
1127: HD 303308 & 5.36 & (0.96) & 12.32 & (0.09) & 21.46 & (0.06) & 21.41 &(0.06)&
1128: 20.18 & (0.12) & $-$9.14 &$^{+0.11}_{-0.14}$& O3V &2.67& $-$0&46 & $-$0.98 \\
1129: %CPD $-$59 2603&7.25 & (0.10) & 12.44 & (0.14) & 21.47 & (0.06) & 21.43 &(0.06)&
1130: %20.08& (0.06) & $-$9.03 &$^{+0.15}_{-0.23}$& O7V &2.67& $-$0&45 & $-$1.09 \\
1131: \enddata
1132: \tablenotetext{a}{The spectral type for the star associated with this sight
1133: line suggests that the interstellar Ly-$\alpha$ profile is potentially
1134: contaminated by stellar neutral hydrogen. These Kr/H ratios are used in the
1135: analysis with caution.}
1136: \tablenotetext{b}{The $N$(H), $N$(\ion{H}{1}), ${\langle}n_{\rm H}\rangle$, and
1137: $f$(H$_2$) values are undetermined, due to obvious contamination of the
1138: interstellar Ly-$\alpha$ line by stellar material. These sight lines are
1139: omitted from the Kr/H ratio analysis.}
1140: \tablenotetext{c}{No H$_2$ data are available; this sight line is omitted from
1141: the Kr/H ratio analysis.}
1142: \end{deluxetable}
1143:
1144: \clearpage
1145:
1146: \begin{deluxetable}{r@{$-$}lccccccccccc}
1147: \rotate
1148: \tabletypesize{\footnotesize}
1149: \tablecaption{Interstellar Gas-phase Abundance Weighted Means \label{aratmeans}}
1150: \tablewidth{0pt}
1151: \tablehead{
1152: \multicolumn{2}{c}{Distance}&&&&&&&&&&&\\
1153: \multicolumn{2}{c}{Range}&&\colhead{Error}&\colhead{Scatter}&\colhead{Sample}&
1154: &\colhead{Error}&\colhead{Scatter}&&\colhead{Error}&\colhead{Scatter}&
1155: \colhead{Sample}\\
1156: \multicolumn{2}{c}{(kpc)}&\colhead{log$_{10}$[Kr/H]$_{ISM}$}&\colhead{(dex)}&
1157: \colhead{(dex)}&\colhead{Size}&\colhead{log$_{10}$[O/H]$_{ISM}$}&
1158: \colhead{(dex)}&\colhead{(dex)}&\colhead{log$_{10}$[O/Kr]$_{ISM}$}&
1159: \colhead{(dex)}&\colhead{(dex)}&\colhead{Size}\\
1160: }
1161: \startdata
1162: 0 &0.75&$-9.02$&0.02&0.05&27&$-3.48$&0.03&0.04&5.59&0.04&0.07&6 \\
1163: 0.55&2.50&$-8.91$&0.02&0.08&16&$-3.40$&0.02&0.08&5.53&0.03&0.06&11\\
1164: 2.50&6.00&$-9.03$&0.04&0.10&7 &$-3.47$&0.03&0.10&5.56&0.04&0.04&7 \\
1165: \multicolumn{2}{c}{Near+Far}&$-9.02$&0.02&0.06&34&$-3.48$&0.02&0.08&5.57&0.03&0.05&13\\
1166: \enddata
1167: \end{deluxetable}
1168:
1169: \clearpage
1170:
1171: \begin{deluxetable}{lr@{ }lr@{ }lr@{ }lr@{ }lr@{ }lr@{ }lc}
1172: \rotate
1173: \tabletypesize{\footnotesize}
1174: \tablecaption{Oxygen Abundances \label{OHdata}}
1175: \tablewidth{0pt}
1176: \tablehead{
1177: &\multicolumn{2}{c}{$W_{\lambda1356}$}&\multicolumn{2}{c}{log$_{10}$[$N$(O)]}&
1178: \multicolumn{2}{c}{log$_{10}$[$N$(Kr)]}&
1179: \multicolumn{2}{c}{log$_{10}$[$N$(H)]}&
1180: \multicolumn{2}{c}{ }&\multicolumn{2}{c}{ }&\colhead{$d_\ast$}\\
1181: \colhead{Star}&\multicolumn{2}{c}{(m\AA)}&
1182: \multicolumn{2}{c}{[cm$^{-2}$]}&\multicolumn{2}{c}{[cm$^{-2}$]}&
1183: \multicolumn{2}{c}{[cm$^{-2}$]}&\multicolumn{2}{c}{log$_{10}$[O/Kr]}&
1184: \multicolumn{2}{c}{log$_{10}$[O/H]}&
1185: \colhead{(kpc)}\\
1186: }
1187: \startdata
1188: HD 23478\tablenotemark{a}& 9.58 & (0.57) & 17.79 & (0.09) & 12.23 & (0.05) &
1189: 21.21 & (0.04) & 5.56 &$^{+0.10}_{-0.13}$& $-$3.42 &$^{+0.10}_{-0.12}$&0.32\\
1190: HD 24190 & 13.51 & (0.96) & 17.91 & (0.05) & 12.35 & (0.06) &
1191: 21.30 & (0.05) & 5.56 &$^{+0.08}_{-0.09}$& $-$3.39 &$^{+0.07}_{-0.08}$&0.32\\
1192: HD 40893 & 19.94 & (1.06) & 18.07 & (0.03) & 12.56 & (0.07) &
1193: 21.54 & (0.05) & 5.51 &$^{+0.08}_{-0.09}$& $-$3.47 &$^{+0.06}_{-0.07}$&2.80\\
1194: HD 69106 & 8.34 & (0.67) & 17.69 & (0.05) & 12.19 & (0.09) &
1195: 21.09 & (0.06) & 5.50 &$^{+0.10}_{-0.13}$& $-$3.40 &$^{+0.08}_{-0.09}$&1.40\\
1196: HD 94454\tablenotemark{b}& 10.12 & (0.80) & 17.83 & (0.04) & 12.36 & (0.08) &
1197: \multicolumn{2}{c}{\nodata}& 5.47 &$^{+0.09}_{-0.11}$&
1198: \multicolumn{2}{c}{\nodata}& 0.33\\
1199: HD 99857 & 13.32 & (0.89) & 17.92 & (0.04) & 12.34 & (0.08) &
1200: 21.36 & (0.05) & 5.58 &$^{+0.09}_{-0.11}$& $-$3.44 &$^{+0.06}_{-0.07}$&3.06\\
1201: HD 99872\tablenotemark{b}& 9.10 & (0.72) & 17.78 & (0.05) & 12.25 & (0.07) &
1202: \multicolumn{2}{c}{\nodata}& 5.53 &$^{+0.08}_{-0.10}$&
1203: \multicolumn{2}{c}{\nodata}& 0.23\\
1204: HD 102065\tablenotemark{b}& 7.63 & (0.51) & 17.70 & (0.07) & 12.16 & (0.11) &
1205: \multicolumn{2}{c}{\nodata}& 5.54 &$^{+0.13}_{-0.18}$&
1206: \multicolumn{2}{c}{\nodata}& 0.17\\
1207: HD 104705\tablenotemark{d}& 7.82 & (0.47) & 17.67 & (0.06) & 12.05 & (0.09) &
1208: 21.20 & (0.05) & 5.62 &$^{+0.10}_{-0.12}$& $-$3.53 &$^{+0.06}_{-0.07}$&3.90\\
1209: HD 108639 & 20.81 & (1.79) & 18.09 & (0.06) & 12.40 & (0.10) &
1210: 21.38 & (0.08) & 5.69 &$^{+0.11}_{-0.15}$& $-$3.29 &$^{+0.10}_{-0.13}$&0.11\\
1211: HD 112999\tablenotemark{b}& 9.01 & (1.04) & 17.75 & (0.07) & 12.27 & (0.09) &
1212: \multicolumn{2}{c}{\nodata}& 5.48 &$^{+0.11}_{-0.15}$&
1213: \multicolumn{2}{c}{\nodata}& 0.34\\
1214: HD 114886 & 15.61 & (1.67) & 17.95 & (0.06) & 12.43 & (0.14) &
1215: 21.40 & (0.05) & 5.52 &$^{+0.15}_{-0.23}$& $-$3.45 &$^{+0.08}_{-0.09}$&0.38\\
1216: HD 115071 & 26.12 & (1.53) & 18.20 & (0.03) & 12.64 & (0.07) &
1217: 21.50 & (0.05) & 5.56 &$^{+0.08}_{-0.09}$& $-$3.30 &$^{+0.06}_{-0.07}$&1.64\\
1218: HD 116852 & 6.41 & (0.65) & 17.56 & (0.04) & 12.06 & (0.06) &
1219: 21.03 & (0.06) & 5.50 &$^{+0.07}_{-0.08}$& $-$3.47 &$^{+0.07}_{-0.08}$&4.80\\
1220: HD 122879 & 22.39 & (2.07) & 18.10 & (0.04) & 12.60 & (0.13) &
1221: 21.35 & (0.10) & 5.50 &$^{+0.13}_{-0.20}$& $-$3.25 &$^{+0.11}_{-0.14}$&2.10\\
1222: HD 124314 & 24.32 & (1.53) & 18.17 & (0.04) & 12.70 & (0.09) &
1223: 21.47 & (0.05) & 5.47 &$^{+0.10}_{-0.12}$& $-$3.30 &$^{+0.06}_{-0.07}$&1.15\\
1224: HD 137595\tablenotemark{a}& 9.15 & (0.78) & 17.76 & (0.04) & 12.34 & (0.07) &
1225: 21.24 & (0.04) & 5.42 &$^{+0.08}_{-0.10}$& $-$3.48 &$^{+0.06}_{-0.06}$&0.83\\
1226: HD 144965\tablenotemark{a}& 8.74 & (0.68) & 17.80 & (0.06) & 12.30 & (0.08) &
1227: 21.37 & (0.05) & 5.50 &$^{+0.08}_{-0.10}$& $-$3.57 &$^{+0.08}_{-0.09}$&0.68\\
1228: HD 147165 & 14.62 & (0.58) & 17.97 & (0.10) & 12.45 & (0.05) &
1229: 21.40 & (0.08) & 5.52 &$^{+0.11}_{-0.15}$& $-$3.43 &$^{+0.12}_{-0.17}$&0.15\\
1230: HD 147683\tablenotemark{a}&15.06 & (0.90) & 18.00 & (0.05) & 12.52 & (0.06) &
1231: 21.55 & (0.07) & 5.48 &$^{+0.08}_{-0.09}$& $-$3.55 &$^{+0.08}_{-0.10}$&0.59\\
1232: HD 151805\tablenotemark{d}&16.57 & (1.58) & 17.98 & (0.07) & 12.41 & (0.08) &
1233: 21.41 & (0.05) & 5.57 &$^{+0.13}_{-0.20}$& $-$3.43 &$^{+0.08}_{-0.10}$&5.91\\
1234: HD 152590 & 16.80 & (0.85) & 18.01 & (0.03) & 12.54 & (0.05) &
1235: 21.47 & (0.05) & 5.47 &$^{+0.06}_{-0.07}$& $-$3.46 &$^{+0.06}_{-0.07}$&1.80\\
1236: HD 165246 & 15.63 & (1.20) & 18.00 & (0.04) & 12.53 & (0.06) &
1237: 21.46 & (0.06) & 5.47 &$^{+0.07}_{-0.08}$& $-$3.46 &$^{+0.06}_{-0.07}$&1.59\\
1238: HD 177989 & 11.92 & (0.89) & 17.83 & (0.04) & 12.25 & (0.07) &
1239: 21.09 & (0.05) & 5.58 &$^{+0.08}_{-0.10}$& $-$3.26 &$^{+0.06}_{-0.07}$&4.91\\
1240: HD 203374 & 17.36 & (0.84) & 18.02 & (0.04) & 12.56 & (0.06) &
1241: 21.38 & (0.06) & 5.46 &$^{+0.07}_{-0.08}$& $-$3.36 &$^{+0.07}_{-0.08}$&0.62\\
1242: HD 206267 & 19.59 & (0.82) & 18.11 & (0.03) & 12.65 & (0.10) &
1243: 21.46 & (0.04) & 5.46 &$^{+0.10}_{-0.14}$& $-$3.35 &$^{+0.05}_{-0.06}$&0.59\\
1244: HD 208947\tablenotemark{c}&10.11 & (1.02) & 17.68 & (0.03) & 11.99 & (0.08) &
1245: 21.10 & (0.06) & 5.69 &$^{+0.08}_{-0.10}$& $-$3.42 &$^{+0.06}_{-0.07}$&0.50\\
1246: HD 209339 & 14.99 & (0.80) & 17.96 & (0.04) & 12.35 & (0.06) &
1247: 21.26 & (0.06) & 5.61 &$^{+0.07}_{-0.08}$& $-$3.30 &$^{+0.07}_{-0.08}$&1.00\\
1248: HD 224151 & 20.28 & (2.27) & 18.08 & (0.06) & 12.58 & (0.12) &
1249: 21.43 & (0.05) & 5.50 &$^{+0.13}_{-0.19}$& $-$3.35 &$^{+0.08}_{-0.09}$&1.10\\
1250: HD 303308\tablenotemark{d}&13.09 & (0.55) & 17.91 & (0.04) & 12.32 & (0.09) &
1251: 21.46 & (0.06) & 5.59 &$^{+0.10}_{-0.12}$& $-$3.55 &$^{+0.07}_{-0.08}$&2.67\\
1252: %CPD$ -$59 2603 & 20.87 & (0.03) & 18.09 & (0.03) & 12.44 & (0.14) &
1253: %21.47 & (0.06) & 5.65 &$^{+0.14}_{-0.21}$& $-$3.38 &$^{+0.07}_{-0.08}$&2.67\\
1254: \enddata
1255: \tablenotetext{a}{Potential Ly-$\alpha$ contamination; hydrogen abundance
1256: ratios used with caution.}
1257: \tablenotetext{b}{There is apparent Ly-$\alpha$ contamination; hydrogen
1258: abundance ratios are unavailable.}
1259: \tablenotetext{c}{No H$_2$ data are available; ratios involving hydrogen refer
1260: to \ion{H}{1} only.}
1261: \tablenotetext{d}{The given oxygen abundance is tallied over radial velocities
1262: consistent with the krypton profile; the measurement does not represent the
1263: total oxygen abundance for this sight line.}
1264: \end{deluxetable}
1265:
1266: \end{document}
1267: