1: \documentclass[12pt,preprint]{aastex}
2:
3: \newcommand{\ha}{H$\alpha$}
4: \newcommand{\oi}{[\ion{O}{1}]}
5: \newcommand{\oiw}{[\ion{O}{1}]~$\lambda$6300}
6: \newcommand{\hi}{\ion{H}{1}}
7: \newcommand{\kms}{km~s$^{-1}$}
8:
9: \shorttitle{\ha\ toward HD~93521 and the Lockman Window}
10: \shortauthors{Hausen et al.}
11:
12: \begin{document}
13: \title{Interstellar \ha\ Line Profiles toward HD~93521 and the Lockman Window}
14:
15: \author{N. R. Hausen, R. J. Reynolds, L. M. Haffner}
16: \affil{Department of Astronomy, University of Wisconsin--Madison}
17: \affil{475 North Charter Street, Madison, WI 53706}
18: \email{hausen@astro.wisc.edu, reynolds@astro.wisc.edu, haffner@astro.wisc.edu}
19: \and
20: \author{S. L. Tufte}
21: \affil{Department of Physics, Lewis \& Clark College}
22: \affil{0615 SW Palatine Hill Road, Portland, OR 97219}
23: \email{tufte@lclark.edu}
24:
25: \begin{abstract}
26:
27: We have used the Wisconsin \ha\ Mapper (WHAM) facility to measure
28: the interstellar \ha\ emission toward the high Galactic latitude O
29: star HD~93521 ($\ell~=~183\fdg1$, $b~=~+62\fdg2$). Three emission
30: components were detected along this line of sight. These components
31: have radial velocities of --10~\kms, --51~\kms, and --90~\kms\ with
32: respect to the local standard of rest (LSR) and have \ha\
33: intensities of 0.20~R, 0.15~R, and 0.023~R, respectively,
34: corresponding to emission measures of 0.55 cm$^{-6}$ pc, 0.42
35: cm$^{-6}$ pc, and 0.06 cm$^{-6}$ pc. We have also measured the \ha\
36: emission toward the direction $\ell~=~148\fdg5$, $b~=~+53\fdg0$,
37: which lies in the region of exceptionally low \hi\ column density
38: known as the Lockman Window. An emission component is detected in
39: this direction at --1~\kms\ (LSR) with an intensity of 0.20~R (0.55
40: cm$^{-6}$ pc). In addition, we studied the direction
41: $\ell~=~163\fdg5$, $b~=~+53\fdg5$. No Galactic emission was
42: detected along this line of sight, and upper limits on the possible
43: intensity of Galactic emission toward this direction are 0.11~R at
44: the LSR and 0.06~R at --50~\kms. As part of the process of
45: separating the interstellar from the terrestrial emission, we also
46: detected and characterized twelve faint ($\sim$0.03-0.15~R),
47: unidentified atmospheric lines present in WHAM \ha\ spectra. Lastly,
48: we have used WHAM to obtain \oiw\ spectra along the line of sight
49: toward HD~93521. We do not conclusively detect interstellar \oi\
50: emission toward the star, but place an upper limit of 0.060~R on the
51: \oi\ intensity of the intermediate velocity (--51~\kms) component.
52: If the temperature of the gas is 10,000 K, this limit implies that
53: within the \ha\ emitting region, the hydrogen ionization fraction
54: $n(\mathrm{H}^+)/n(\mathrm{H}_{\mathrm{total}}) > 0.6$.
55:
56: \end{abstract}
57:
58: \keywords{ISM: atoms --- ISM: clouds ---
59: ISM: general --- stars: individual (HD~93521)}
60:
61: \section{INTRODUCTION}
62: \label{sec:intro}
63:
64: The Wisconsin \ha\ Mapper (WHAM) is a dual-etalon, Fabry-Perot
65: spectrometer designed to measure optical emission from diffuse
66: interstellar sources. WHAM is approximately 100 times more sensitive than
67: the previous instrument of its kind, making it possible to study very low
68: levels of emission not accessible in the past. The WHAM \ha\ studies
69: presented in this paper were made along three lines of sight in which the
70: interstellar \ha\ emission is extremely faint. In particular, we have
71: measured the \ha\ emission line profile toward HD~93521
72: ($\ell~=~183\fdg1$, $b~=~+62\fdg2$), supplementing the work of
73: \cite{sf93}, who obtained ultraviolet \emph{HST} absorption line
74: measurements for numerous species toward this star. Secondly, we have
75: measured the \ha\ emission toward $\ell~=~148\fdg5$, $b~=~+53\fdg0$,
76: which lies in the region of sky known as the Lockman Window. This region
77: has the lowest \hi\ column density of any section of the northern sky, and
78: thorough 21~cm studies of the area have been made by \cite{jahoda90} and
79: \cite{lockman86}. The \ha\ observations provide information about the
80: amount of H~II in this unusual direction. Lastly, we have observed the
81: line of sight toward $\ell~=~163\fdg5$, $b~=~+53\fdg5$ (referred to below
82: as ``Off A''). This direction was selected for investigation based on
83: WHAM Northern Sky Survey data, which indicated that only minimal levels of
84: \ha\ emission could be present. Results for these three directions are
85: discussed in \S\ref{sec:ha_93521}, \S\ref{sec:lm}, and \S\ref{sec:offA},
86: respectively.
87:
88: Accompanying WHAM's ability to detect faint \emph{Galactic} emission is
89: its ability to detect faint \emph{terrestrial} emission lines as well.
90: In fact, WHAM \ha\ spectra centered near the local standard of rest (LSR)
91: show a total of twelve weak ($\sim$0.03 -- 0.15~R, 1~R~=~10$^{6}$/4$\pi$
92: photons cm$^{-2}$ s$^{-1}$ sr$^{-1}$) sky lines of unknown origin in
93: addition to the well known and much brighter geocoronal \ha\ line. These
94: weak lines are present at fixed positions with respect to the geocoronal
95: line, and their intensities vary slightly during the night. Of these
96: twelve lines, four on the blue side of the geocoronal line were first
97: noticed in WHAM spectra used to study high-velocity clouds and are
98: discussed by \cite{tufte98} and by \cite{tufte97}. One of the lines to
99: the red of the geocoronal line is discussed by \cite{haffner98}. Since
100: the atmospheric lines can significantly affect measurements of faint
101: Galactic \ha\ emission, it is important to understand and remove their
102: contributions to the data. In \S\ref{sec:atmlines}, we give a
103: characterization of the twelve faint atmospheric lines based on a study of
104: a large sample of WHAM \ha\ spectra.
105:
106: In addition to \ha\, other, fainter emission lines can provide information
107: about the physical state of the emitting gas. For example, measurements
108: of Galactic \oiw\ emission along a given line of sight, when combined with
109: the \ha\ intensity for that direction, yield information regarding the
110: ionization state of the emitting gas, a subject of interest and some
111: controversy for the HD 93521 sightline (e.g., \citealt{sf93, smh00}).
112: In
113: particular, the hydrogen ionization ratio
114: $n(\mathrm{H^+})$/$n(\mathrm{H^\circ})$ in the gas can be obtained via the
115: observed \oi/\ha\ line intensity ratio. The precise relation, as given by
116: \cite{reynolds98} and references therein, is
117: \begin{equation}
118: \frac{I_\mathrm{[O~I]}}{I_{\mathrm{H}\alpha}} =
119: 2.74 \times 10^4 \left(\frac{T_4^{1.854}}{1 + 0.605T_4^{1.105}}\right)
120: \mathrm{exp}\left(
121: -\frac{2.284}{T_4} \right)
122: \frac{n(\mathrm{O})}{n(\mathrm{H})}
123: \left[ \frac{1 + n(\mathrm{H^{\circ}})/n(\mathrm{H^+})}
124: {1 + (8/9)n(\mathrm{H^{+}})/n(\mathrm{H^\circ})} \right],
125: \end{equation}
126: where $I_\mathrm{[O~I]}/I_{\mathrm{H}\alpha}$ is the intensity ratio
127: measured in energy units and $T_4$ is the electron temperature in units of
128: 10$^4$~K. The term \emph{n}(O)/\emph{n}(H) is the gas-phase abundance of
129: oxygen, which we take to be $3.2~\times~10^{-4}$ \citep{mjc98}. The [O~I]
130: observations are discussed in \S\ref{sec:oi_93521} and \S\ref{sec:disc}.
131:
132:
133: \section{OBSERVATIONS}
134: \label{sec:obs}
135:
136: The Wisconsin \ha\ Mapper facility consists of a 15~cm aperture,
137: dual-etalon, multichannel Fabry-Perot spectrometer attached to a 0.6~m
138: siderostat. WHAM has a 1$\arcdeg$ beam on the sky and 12~\kms\ radial
139: velocity resolution over a 200~\kms\ spectral range. This 200~\kms\
140: interval can be centered on any wavelength between 4800 and 7300~\AA. A
141: more detailed description of this instrument and its operation is provided
142: by \cite{tufte97} and \cite{haffner99}. WHAM is a remotely-operable
143: facility located at Kitt Peak National Observatory, and all observations
144: for this paper were obtained remotely from Madison, WI.
145:
146:
147: \subsection{\ha}
148: \label{sec:ha_obs}
149:
150: An ``on-off'' technique similar to that described by \cite{tufte98} was
151: used to determine the interstellar \ha\ emission line profiles toward
152: HD~93521 and the Lockman Window. With this method, exposures toward the
153: direction of interest (the ``on'') are alternated with exposures toward an
154: ``off'' direction in which the Galactic emission is significantly weaker.
155: The ``off'' spectra are then subtracted from the ``on'' spectra, thereby
156: removing, or at least significantly reducing, all features common to the
157: two directions, including the geocoronal \ha\ line, other much fainter
158: terrestrial emission lines, and the weak \ha\ Fraunhofer absorption line
159: in the zodiacal light.
160:
161: A summary of \ha\ observations is given in Table~\ref{table:ha_obs}. Each
162: ``on'' spectrum and each ``off'' spectrum had an exposure time of 600~s,
163: except the Lockman Window spectra noted in Table~\ref{table:ha_obs}.
164: HD~93521 is the ``on'' direction in Data Sets 1a,b,c, 3, and 7. The
165: Lockman Window is the ``on'' direction in Data Sets 5 and 6. For the
166: HD~93521 observations that used the Lockman Window as the ``off,'' the
167: ``on-off'' technique was complicated by the presence of a significant
168: amount of Galactic emission toward the Lockman direction. Our correction
169: for this is discussed in \S\ref{sec:ha_93521}. When ``Off A'' was used
170: (Data Sets 5, 6, and 7), no correction for Galactic emission was made,
171: because the Galactic emission in this direction is below the detection
172: limit of these observations (see \S\ref{sec:offA}). In addition,
173: observations were repeated at different times of the year in order to
174: utilize the shifting positions of the Galactic emission relative to the
175: fixed positions of the atmospheric lines present in the spectra.
176: Consistency between results was used as a check that the detected emission
177: was Galactic in origin and not the result of contamination from the
178: atmospheric lines.
179:
180: \subsection{Faint Atmospheric Lines near \ha}
181: \label{atmlines_obs}
182:
183: Our characterization of the faint atmospheric lines is based primarily on
184: a study of more than eighty 60~s spectra toward the Lockman Window, taken
185: on roughly 20 nights spanning a 14-month period in 1997 and 1998. Of
186: these spectra, approximately half had the geocoronal line positioned on
187: the red side of WHAM's 200~\kms\ spectral range, while the other half had
188: the geocoronal line on the blue side of the spectrum. These data were
189: gathered routinely during observing sessions for the WHAM Northern Sky
190: Survey.
191: In addition, we made use of many of the Lockman Window spectra listed in
192: Table~\ref{table:ha_obs}, including the spectra (each 600~s exposures) in
193: Data Sets 2a,b and 4a,b.
194:
195: Obtaining Lockman Window spectra over many months was important to our
196: being able to separate the Galactic from the terrestrial emission, since
197: at any one time, the Galactic emission near the LSR blends with one or
198: more of the atmospheric lines, which have intensities comparable to that
199: of the interstellar \ha. Over the course of the year, the position of the
200: Galactic emission moves relative to the positions of the terrestrial
201: lines, making the detection of the Galactic emission possible. In the
202: process we were also able to characterize the atmospheric lines in the
203: spectrum.
204:
205: We also studied numerous high-latitude ``blocks'' taken from the WHAM
206: Survey data (\citealt{survey}; also see
207: \texttt{\url{http://www.astro.wisc.edu/wham/}}). Each block was the average
208: of about forty nine 1$\arcdeg$ pointings (each with 30~s of exposure
209: time) covering a roughly 6$\arcdeg$~$\times$~7$\arcdeg$ region of sky.
210: These blocks contained only low levels of Galactic emission, making it
211: possible to independently test the results from the Lockman Window
212: studies. This assured us that the atmospheric line parameters that we
213: had determined were not influenced by the Galactic emission toward a
214: particular direction.
215:
216:
217: \subsection{\oiw}
218: \label{sec:oi_obs}
219:
220: Two bright atmospheric lines, the \oi\ airglow line and an OH line at
221: 6297.9 \AA, are present in the \oiw\ spectra, making the detection of
222: faint Galactic \oi\ emission quite difficult. The \oiw\ studies for
223: HD~93521 were therefore also based on the ``on-off'' method. Furthermore,
224: for comparison purposes, observations were repeated at different times of
225: year as the position of the Galactic emission shifted with respect to the
226: terrestrial lines. Table~\ref{table:oi_obs} lists the observations. The
227: Lockman Window is the ``off'' direction for Data Sets 8a,b, 9a,b, and
228: 10a,b, whereas Data Set 11 uses Off A as the ``off.'' With the exception
229: of the Lockman Window spectra noted in Table~\ref{table:oi_obs}, all
230: individual spectra were 600~s exposures. Spectra in Data Set 11 were taken
231: in an ``on''/``off'' sequence like that described for the \ha\ Data Sets
232: (see \S\ref{sec:ha_obs}). Data in the remaining \oi\ Data Sets were
233: gathered in the repeated sequence ``on,'' ``off,'' ``on,'' yielding twice
234: as much integration time for the ``on'' direction as for the ``off.''
235:
236:
237: \section{RESULTS}
238: \label{sec:results}
239:
240: \subsection{Faint Atmospheric Lines near \ha}
241: \label{sec:atmlines}
242:
243: We detected twelve weak atmospheric lines in the WHAM \ha\ spectra
244: centered near the local standard of rest. Figure~\ref{fig:mlines},
245: plotted with respect to the earth's velocity frame, shows the
246: atmospheric lines that are visible when the geocoronal line is on the
247: red side of the spectrum. The spectrum in Figure~\ref{fig:mlines} is
248: the sum of all Lockman Window spectra in Data Sets 1a,b,c, 3, 4a,b,
249: and 5, normalized to 600~s. In each of the Data Sets 1, 3, 4, and 5,
250: the Lockman Window Galactic emission near the LSR (discussed in \S3.3
251: below) appears at a different geocentric velocity. Hence, this
252: Galactic emission was removed from each of the constituent spectra
253: before summing. Figure~\ref{fig:elines}, also plotted with respect to
254: the earth's velocity frame, shows a different wavelength region, with
255: the geocoronal line located on the blue side of the spectrum. This
256: spectrum is the sum of all Lockman Window spectra in Data Sets 2a,b
257: and 6, normalized to 600~s. Again, the Galactic emission has been
258: removed from this spectrum.
259:
260: Parameters for the faint atmospheric lines are given in
261: Table~\ref{table:atmlines}. The low intensity and close spectral
262: proximity of the lines made it difficult to fit them using free position,
263: width, and intensity parameters. A fixed width of 10~\kms\ was therefore
264: adopted for all lines except those numbered 4 and 10. The width was fixed
265: at 15~\kms\ for Lines 4 and 10, which appear broader than the other lines.
266: Line 10 may in fact be a blend of two components, as some spectra show a
267: narrower-than-usual line with the position shifted by approximately
268: 0.05~\AA\ from the typical position of Line 10. It is possible that the
269: remaining lines are also blends that WHAM does not resolve. Since the
270: feature labeled in Figure~\ref{fig:elines} as Line 7a appeared to be
271: present in only a very limited number of spectra, we did not characterize
272: it in the same manner as the other atmospheric lines (see
273: Table~\ref{table:atmlines}). We have not identified the atomic or
274: molecular source of any of the weak atmospheric lines.
275:
276: Table~\ref{table:atmlines} gives the intensity of the lines as shown in
277: Figures~\ref{fig:mlines} and \ref{fig:elines}. The lines that are common
278: to both Figure~\ref{fig:mlines} and Figure~\ref{fig:elines} illustrate the
279: degree to which the relative intensities of the lines can vary. For
280: example, Line 6 is not detected in Figure~\ref{fig:elines}, and Line 7a is
281: not detected in Figure~\ref{fig:mlines}. We have found that, in general,
282: absolute intensities are quite stable ($< 0.05$ R) over periods of a half
283: hour or more. Furthermore, the relative intensities between lines are
284: consistent enough for us to use a set of ``adopted relative intensities'',
285: and these are given in Table~\ref{table:atmlines}. In a given spectrum,
286: the adopted intensity for each line is a percentage of the intensity of
287: Line 5 in that spectrum. The intensity of Line 5 itself is obtained from a
288: Gaussian fit to the spectrum.
289:
290:
291: \subsection{The ``Off A'' Direction}
292: \label{sec:offA}
293:
294: During reduction of the WHAM Northern Sky Survey, the Off A direction
295: ($\ell~=~163\fdg5$, $b~=~+53\fdg5$) was identified as having an
296: exceptionally low level of \ha\ emission. Subsequent long exposures
297: toward this direction have confirmed this (Data Set 5, in which the LSR is
298: maximally to the blue of the geocoronal line for this direction). Although
299: we detected no Galactic \ha\ emission toward this direction, we are
300: ultimately limited by the presence of the geocoronal line and by
301: uncertainties in the intensities of the faint atmospheric lines.
302: Table~\ref{table:ha_results} lists the resulting upper limits placed on
303: possible Galactic emission toward the Off~A direction from 0 to --50~\kms.
304: We derived these limits using Data Set 5 spectra and assuming that three
305: sources of uncertainty in our knowledge of the weak atmospheric lines (the
306: absolute intensity of Line 5, the exact intensity ratios in a given
307: spectrum between Line 5 and the other lines, and the uncertainty as to the
308: presence of Line 6) all add positively in the same direction.
309:
310:
311: \subsection{Interstellar \ha\ Toward the Lockman Window}
312: \label{sec:lm}
313:
314: The spectra in Data Set 5 were used to determine the interstellar \ha\
315: emission line
316: profile for the Lockman Window direction. Specifically, the six 600~s Off
317: A spectra in Data Set 5 were averaged and then subtracted from the average
318: of the six 600~s Lockman Window spectra in the same Data Set. The
319: resulting spectrum is shown in Figure~\ref{fig:main}. In this figure, the
320: subtraction residual of the geocoronal line (at +21~\kms) has been
321: removed. Results from fits of the LSR component in this spectrum are
322: listed in Table~\ref{table:ha_results}. The error bars in the table
323: result from baseline uncertainties and the blend of the Galactic emission
324: with the geocoronal line residual. For the intensity of the --1~\kms\
325: component, there is an additional systematic uncertainty (not included in
326: Table~\ref{table:ha_results}) of +0.11~R, resulting from the upper limit
327: on Galactic emission toward the Off A direction. We also derived the
328: Galactic \ha\ spectrum toward the Lockman Window by subtracting Lockman
329: Window spectra obtained in November (when the LSR is at --26~\kms\ wrt the
330: earth) from Lockman Window spectra obtained in April (when the LSR is at
331: +15~\kms\ wrt the earth). This difference spectrum greatly reduced the
332: atmospheric features and revealed
333: the Galactic emission as negative and positive spectral features near
334: --26~\kms\ and +15~\kms\, respectively, with respect to the earth. The
335: parameters for the Galactic emission derived from this technique were
336: consistent with those derived using the Off A direction, indicating
337: that the interstellar emission toward Off A is indeed less than 0.1~R.
338:
339: For comparison, Figure~\ref{fig:main} also shows the \hi\ profile toward
340: the Lockman Window. This spectrum is from the Leiden/Dwingeloo survey
341: \citep{hb97}, and is the average of their $0\fdg5$ pointings whose centers
342: fall within WHAM's 1$\arcdeg$ beam directed at $\ell = 148\fdg5$, $b =
343: +53\fdg0$. Since the \hi\ spectrum has an obvious component at --54~\kms,
344: an upper limit in the \ha\ at this velocity was determined. The limit,
345: given in Table~\ref{table:ha_results}, is based on fits to Lockman Window
346: spectra at various times of the year. For these fits, a component was
347: fixed in the spectra at --54~\kms\ with a width of 25~\kms. Different
348: possible values of atmospheric line intensities were then considered in
349: order to establish the Galactic emission upper limit listed in
350: Table~\ref{table:ha_results}.
351:
352: Lastly, two noticeable features in the Lockman Window \ha\ spectrum should
353: be mentioned: the elevated data points near +60~\kms\ and the rise in the
354: data at the negative edge of the spectrum. These features are within the
355: baseline uncertainty of the spectrum and have no apparent correspondence
356: in \hi. It is nevertheless possible that one or both represent very
357: low-intensity Galactic \ha\ emission. If the true baseline is near zero in
358: the Figure~\ref{fig:main} spectrum, then the positive-velocity feature is
359: centered at about +56~\kms\ with an intensity of 0.04~R. The spectra in
360: Data Set 6 were obtained in an attempt to verify whether this is a real
361: Galactic emission component. Unfortunately, because the atmospheric lines
362: near the region of interest showed marked intensity fluctuations between
363: temporally-adjacent 600~s spectra in Data Set 6, we were unable to verify
364: the reality of this weak feature.
365:
366: No additional data were obtained to test whether the feature at the
367: negative edge of the spectrum is indeed a high-velocity Galactic emission
368: component. However, while using the Lockman Window as an ``off''
369: direction during a study of emission from high-velocity clouds,
370: \cite{tufte97} found evidence for the presence of a very weak emission
371: component toward the Lockman Window near --130~\kms. This lends support
372: to the idea that the feature in Figure~\ref{fig:main} is Galactic in
373: origin, but clearly further ``on'' minus ``off'' observations centered at
374: a high, negative velocity would be needed to provide a definitive result.
375:
376:
377:
378: \subsection{Interstellar \ha\ Toward HD 93521}
379: \label{sec:ha_93521}
380:
381: ``On-off'' studies of the spectra in Data Sets 1a,b,c and 3 reveal three
382: interstellar \ha\ emission components toward HD~93521. There is a
383: ``slow'' component at --10~\kms, an ``intermediate velocity'' component at
384: --51~\kms, and a fainter, ``high velocity'' component at --90~\kms. Full
385: results are given in Table~\ref{table:ha_results}. The error bars given
386: in Table~\ref{table:ha_results} for HD~93521 incorporate uncertainties
387: associated with the baseline level and the blend of the slow component
388: with the residual of the incompletely-subtracted geocoronal line. They
389: also include the uncertainty generated by the error bars of the --1~\kms\
390: Lockman emission. The results in Table~\ref{table:ha_results} do not
391: include a systematic uncertainty of +0.11~R at the LSR, due to the upper
392: limit on the Galactic emission toward Off A, or a systematic uncertainty
393: of +0.06~R at --50~\kms, due to the upper limit on the Galactic emission
394: at --50~\kms\ toward the Lockman Window.
395:
396: The analysis conducted to obtain the results presented in
397: Table~\ref{table:ha_results} proceeded as follows: The five 600~s Lockman
398: Window spectra in Data Sets 1a,b,c were averaged together, and the
399: Galactic emission at --1~\kms\ (parameters given in
400: Table~\ref{table:ha_results}) was removed from the average spectrum. It
401: was assumed that there was no other Galactic emission in the Lockman
402: Window spectra. The corrected average ``off'' spectrum was then
403: subtracted from the average of the five 600~s HD~93521 spectra in Data
404: Sets 1a,b,c. A similar procedure was followed for the data in Data Set 3,
405: with the additional step of normalizing the average Lockman Window
406: spectrum to 600~s of exposure time. Both ``on-off'' spectra were then
407: studied, although parameters for the slow emission component were
408: determined from the data in Data Sets 1a,b,c only, as the slow
409: component was farthest from the incompletely-subtracted geocoronal line in
410: this case. During the Data Set 3 analysis, the parameters of the slow
411: component were fixed at the best-fit values derived from Data Sets 1a,b,c.
412:
413: As a check on these results, the spectra in Data Set 7 (for which Off A is
414: the ``off'' direction) were also analyzed using the ``on-off'' method.
415: This work was carried out under the assumption that there was no Galactic
416: emission present in the Off A spectra. Also, because the Galactic
417: emission near the LSR was blended with the incompletely-subtracted
418: geocoronal line, parameters for the slow emission component were fixed
419: during the analysis at the best-fit values given for this component in
420: Table~\ref{table:ha_results}. The intermediate velocity emission
421: component was then detected at --53~\kms, with a width of 31~\kms\ and an
422: intensity of 0.13~R. The high velocity emission component was detected at
423: --86~\kms, with a width of 20~\kms\ and an intensity of 0.022~R. These
424: results are generally consistent with the uncertainties listed in
425: Table~\ref{table:ha_results}, except for the best-fit width of the
426: intermediate velocity component, which is slightly outside the error bars
427: of the value derived from the study that used the Lockman Window as the
428: ``off.'' It is possible that this discrepancy is due to very low levels of
429: undetected Galactic emission in the Lockman Window and/or Off A
430: directions. Such emission may produce effects of this magnitude in the
431: ``on-off'' line profile of an ``on'' direction in which the Galactic
432: emission is faint.
433:
434: Figure~\ref{fig:main} shows the \ha\ profile toward HD~93521. This
435: profile was obtained from the spectra in Data Sets 1a,b,c and 3.
436: Specifically, the residuals from the incompletely-subtracted geocoronal
437: line were removed from the two ``on-off'' spectra generated from these
438: Data Sets. (These residuals were located at +26~\kms\ in the spectrum
439: from Data Sets 1a,b,c and at +14~\kms\ in the Data Set 3 spectrum.) Then
440: the two resulting spectra were averaged, yielding the spectrum given in
441: Figure~\ref{fig:main}. Due to the LSR velocity off-set between 1997
442: November (Data Sets 1a,b,c) and 1999 January (Data Set 3), several data
443: points at the positive velocity extreme in the November data and at the
444: negative velocity extreme in the January data were unmatched with points
445: in the other spectrum. These points were excluded from the average
446: spectrum, and hence the velocity scale of the HD~93521 \ha\ spectrum in
447: Figure~\ref{fig:main} covers only a 188~\kms\ (rather than 200~\kms)
448: interval.
449:
450: Figure~\ref{fig:main} shows the \hi\ spectrum toward HD~93521 as well. The
451: \hi\ spectrum is the average of the Leiden/Dwingeloo pointings whose
452: centers fall within WHAM's 1$\arcdeg$ beam directed at $\ell = 183\fdg1$,
453: $b = +62\fdg2$. The \hi\ spectrum has the same 188 \kms\ velocity scale
454: as the HD~93521 \ha\ spectrum.
455:
456:
457: \subsection{A Search for Interstellar \oiw\ Emission Toward HD 93521}
458: \label{sec:oi_93521}
459:
460: The strength of the \oi\ emission relative to \ha\ is a measure of the
461: hydrogen ionization fraction in the emitting gas and thus has the
462: potential to test the conclusion of \cite{sf93} that the warm ionized gas
463: toward HD 93521 is mixed with the H~I, forming partially ionized (30\%)
464: clouds. Unfortunately, because it is so much weaker than the \ha\, we
465: were unable to make a clear detection of Galactic \oi\ emission toward
466: HD~93521, despite using an ``on-off'' technique and repeating observations
467: at different times of year. The incomplete subtraction of the very bright
468: \oi\ airglow line in each ``on-off'' spectrum prevented the detection of
469: interstellar \oi\ emission near the LSR. Although the intermediate
470: velocity component near --50~\kms\ was adequately separated from both the
471: airglow line and the prominent OH feature 115~\kms\ to the blue of the
472: airglow line, the residual atmospheric line contamination in the
473: ``on-off'' spectra led to baseline uncertainties that were comparable to
474: or greater than the intensity of any Galactic emission present near
475: --50~\kms.
476:
477: Figure~\ref{fig:oi} shows three ``on-off'' \oi\ spectra. The top panel is
478: the 600~s average of the ``off'' (the Lockman Window) spectra in Data Sets
479: 8a and 8b subtracted from the 600~s average of the HD~93521 spectra in the
480: same Data Sets. The middle and bottom panels show spectra obtained
481: similarly for Data Sets 9a,b and 11, respectively. An upper limit on the
482: possible interstellar \oi\ intensity is given in
483: Table~\ref{table:oi_results}. This limit is based on fits to ``on-off''
484: spectra using data from Data Sets 8a through 11. For these fits, it was
485: assumed that the Galactic \oi\ emission was at --51~\kms\ (the velocity of
486: the \ha\ emission) and had a width of 34~\kms. This width was
487: calculated from the width of the \ha\ component at --51~\kms\, assuming
488: that the \ha\ and \oi-emitting gas is well-mixed and at a temperature of
489: 8000~K. The Gaussian component shown in each of the spectra in
490: Figure~\ref{fig:oi} illustrates one possible fit to the data. This
491: component has a radial velocity fixed at --51~\kms (LSR), a width fixed at
492: 34~\kms\ (FWHM), and an intensity slightly less than half the upper limit.
493: For completeness, Table~\ref{table:oi_results} includes \oi/\ha\ intensity
494: ratios for both the upper limit and sample fit cases. These intensity
495: ratios incorporate a correction for the slight differences in instrument
496: response (optical transmittance and detector quantum efficiency) between
497: 6300 \AA\ and 6563 \AA.
498:
499:
500: \section{DISCUSSION AND CONCLUSIONS}
501: \label{sec:disc}
502:
503: Interstellar \ha\ emission has been detected and characterized toward two
504: high Galactic latitude, low H~I column density sightlines, the Lockman
505: Window ($\ell~=~148\fdg5$, $b~=~+53\fdg0$) and HD 93521
506: ($\ell~=~183\fdg1$, $b~=~+62\fdg2$). The relatively bright geocoronal
507: line and the presence of many lower intensity, unidentified atmospheric
508: emission lines make it difficult to measure accurately the interstellar
509: \ha\ emission in these faintest parts of the sky. However, from careful,
510: long integration WHAM spectra spread over many
511: months, the interstellar and terrestrial emissions were successfully
512: separated. The results provide new constraints on the nature of the
513: interstellar clouds along these two sightlines and their environment.
514:
515: We have found that the Lockman Window, in addition to having the
516: lowest H~I column density in the sky ($5 \times 10^{19}$ cm$^{-2}$),
517: also has unusually weak interstellar \ha\ emission (0.20~R), well
518: below the 0.8~R average for this Galactic latitude (from the WHAM sky
519: survey; \citealt{survey}). This region of the sky thus appears to be
520: a true low column density window through the Galactic disk, depleted
521: both in H~I and in H~II. Only one \ha\ velocity component is detected
522: toward the Lockman Window, the component near the LSR, even though in
523: the H~I spectrum there is also a prominent emission component near
524: --50~\kms\ (see Fig.~\ref{fig:main}). An \ha\ intensity of 0.20~R for
525: the Lockman Window implies an emission measure of 0.55 cm$^{-6}$ pc
526: (at 10$^4$ K), which corresponds to a column density
527: N$_{\mathrm{H~II}} \approx 2 \times 10^{19}$ cm$^{-2}$, if the mean
528: density within the ionized regions is about 0.08 cm$^{-3}$ (see
529: \citealt{reynolds91}). This is approximately 40\% the H~I column
530: density and implies that the total hydrogen column density in this
531: direction is about $7 \times 10^{19}$ cm$^{-2}$.
532:
533: Toward HD 93521, $21\arcdeg$ away, the slow and intermediate velocity
534: components detected in \ha\ clearly correspond to the two principal
535: emission components present in the 21 cm spectrum. The intensity of the
536: intermediate velocity (--50~\kms) component relative to the slow
537: (--10~\kms) component is significantly weaker in the \ha\ spectrum than in
538: the 21~cm spectrum, suggesting that perhaps an intermediate velocity \ha\
539: component toward the Lockman Window is just below the detection limit of
540: the observation. For HD 93521 there is also strong evidence for a high
541: velocity \ha\ emission component at --90~\kms, but no corresponding
542: feature in the H~I spectrum. This sightline is very near the High
543: Velocity H~I Cloud Complex M, parts of which have been detected in \ha\ at
544: radial velocities ranging from --109~\kms\ to --61~\kms\
545: \citep{tufte98}.
546: Therefore, the high velocity \ha\ component may be associated with a fully
547: ionized portion of this high velocity cloud complex. The total \ha\
548: intensity (0.37~R) for the HD 93521 sightline corresponds to an emission
549: measure of 1.0 cm$^{-6}$ pc, and thus N$_{\mathrm{H~II}} \approx 4 \times
550: 10^{19}$ cm$^{-2}$ (about 30\% of the H~I column density of $1.25 \times
551: 10^{20}$ cm$^{-2}$).
552:
553: Although there is good correspondence between the H~II and H~I for the low
554: velocity and (for HD 93521) the intermediate velocity components, the
555: difference in the general appearance of the \ha\ and 21 cm line profiles
556: strongly suggests that the H~II and the H~I are not mixed together in the
557: form of partially ionized clouds. In particular, the \ha\ components
558: appear systematically broader. For example, the width (FWHM) of the
559: intermediate velocity H~I component toward HD 93521 is 20 \kms, only about
560: half that of the corresponding \ha\ emission. As a result, there is
561: significant blending between the slow and intermediate velocity components
562: in the \ha\ spectrum, whereas these components are well resolved in the
563: H~I spectrum. This is not a result of the different spectral resolutions
564: of the \ha\ and 21~cm observations. Therefore, rather than being
565: primarily neutral clouds that are 30\%--40\% ionized, as proposed by
566: \cite{sf93} and \cite{sciama98}, each component seems to consist of
567: separate regions of H~I and H~II at nearly the same radial velocity.
568: This could be the result of either a close physical association
569: between the H~I and H~II \citep{mo77} or large scale organized
570: motions of the interstellar medium along this sightline \citep{wc01}.
571:
572: Because the \oiw/\ha\ intensity ratio is a probe of the hydrogen
573: ionization fraction within the emitting gas (see \S\ref{sec:intro}),
574: observations of \oi\ provide a potentially rigorous test of these
575: ``mixed'' versus ``separated'' models for the H~I and H~II in these
576: clouds. Unfortunately, the \oi\ spectrum, which only allows an upper
577: limit on the intermediate velocity (--50~\kms) component
578: (Table~\ref{table:oi_results}; see also \S\ref{sec:oi_93521}), does not
579: provide an unambiguous result. If the temperature is as low as 6000 K, as
580: adopted by \cite{sf93} based on the widths of lines generally associated
581: with H~I clouds, then the \oi\ observations are consistent with their
582: model of primarily neutral clouds containing a 30\% mixture of H$^+$. On
583: the other hand, if the temperature is closer to 10,000 K, as recent
584: observations of [S~II] $\lambda$6717 and [N~II] $\lambda$6584 suggest
585: (Pifer et al, in preparation), then these \oi\ results imply that the
586: emitting regions are predominantly ionized (i.e.,
587: $n(\mathrm{H}^+)/n(\mathrm{H}_{\mathrm{total}}) > 0.6$). The \oi\ limit
588: is also just consistent with recent predictions of a supernova remnant
589: ionization model for the HD 93521 sightline by \cite{smh00}. Given the
590: severe atmospheric line contamination of this region (worse than at \ha ),
591: it will be extremely difficult with ground-based observations to push the
592: sensitivity limit of the \oi\ observations down to the level of 0.03 R or
593: lower needed to test these models more definitively.
594:
595: If the clouds in these two sightlines are in fact ionized by an external
596: flux of ionizing radiation, then the \ha\ intensity of each H~I cloud is a
597: measure of the incident ionizing flux. Specifically, the one-sided
598: incident flux is $2.2 \times 10^6$ $I_{\mathrm{H}\alpha}$ photons
599: cm$^{-2}$ s$^{-1}$, for $I_{\mathrm{H}\alpha}$ in Rayleighs and a
600: temperature of 10$^4$ K (e.g., \citealt{reynolds95, tufte98}). These
601: observations therefore imply a variation in the intensity of the ambient
602: ionizing flux, from less than $1.3 \times 10^5$ photons cm$^{-2}$ s$^{-1}$
603: at the intermediate velocity (--50~\kms) H~I cloud in the Lockman Window
604: to about $4.4 \times 10^5$ photons cm$^{-2}$ s$^{-1}$ for the low velocity
605: clouds. The \ha\ intensity of the high velocity (--90~\kms) component
606: toward HD 93521 gives a flux of $5 \times 10^4$ photons cm$^{-2}$
607: s$^{-1}$; however, because there is no associated H~I detected at this
608: velocity, the H~II is probably density bounded, and thus the derived value
609: for the ionizing flux must be considered as a lower limit. If the higher
610: velocity clouds are farther from the Galactic midplane than the lower
611: velocity clouds \citep{kuntz96}, these results indicate a decrease in
612: the ionizing flux with distance above the plane.
613:
614: \acknowledgements
615:
616: We thank Mark Quigley for his help with the observations and Trudy
617: Tilleman for her on-site monitoring of Kitt Peak sky conditions during
618: these observations. N. R. H. gratefully acknowledges support through a
619: Barry M. Goldwater Scholarship from the Excellence in Education
620: Foundation. This work, including the operation of WHAM, was funded by
621: the National Science Foundation through grant AST96-19424.
622:
623: \clearpage
624:
625: \begin{thebibliography}{ }
626:
627: \bibitem[Haffner(1999)]{haffner99} Haffner, L. M. 1999, Ph.D. Thesis,
628: University of Wisconsin--Madison
629:
630: \bibitem[Haffner, Reynolds, \& Tufte(1998)]{haffner98} Haffner, L. M.,
631: Reynolds, R. J., \& Tufte, S. L. 1998, \apj, 501, L83
632:
633: \bibitem[Haffner et al.(2002)]{survey} Haffner, L. M., Reynolds, R.
634: J., Tufte, S. L., Jaehnig, K. P., \& Percival, J. W. 2002, in
635: preparation
636:
637: \bibitem[Hartmann \& Burton(1997)]{hb97} Hartmann, D., \& Burton, W. B.
638: 1997, Atlas of Galactic Neutral Hydrogen (Cambridge: Cambridge Univ. Press)
639:
640: \bibitem[Jahoda, Lockman, \& McCammon(1990)]{jahoda90} Jahoda, K.,
641: Lockman, F. J., \& McCammon, D. 1990, \apj, 354, 184
642:
643: \bibitem[Kuntz \& Danly(1996)]{kuntz96} Kuntz, K. D., \& Danly, L. 1996,
644: \apj, 457, 703
645:
646: \bibitem[Lockman, Jahoda, \& McCammon(1986)]{lockman86} Lockman, F. J.,
647: Jahoda, K., \& McCammon, D. 1986, \apj, 302, 432
648:
649: \bibitem[McKee \& Ostriker(1977)]{mo77} McKee, C. F., \& Ostriker, J. P.
650: 1977, \apj, 218, 148
651:
652: \bibitem[Meyer, Jura, \& Cardelli(1998)]{mjc98} Meyer, D. M.,
653: Jura, M., \& Cardelli, J. A. 1998, \apj, 493, 222
654:
655: \bibitem[Reynolds(1991)]{reynolds91} Reynolds, R. J. 1991, \apj,
656: 372, L17
657:
658: \bibitem[Reynolds et al.(1998)]{reynolds98} Reynolds, R. J.,
659: Hausen, N. R., Tufte, S. L., \& Haffner, L. M. 1998, \apj, 494, L99
660:
661: \bibitem[Reynolds et al.(1995)]{reynolds95} Reynolds, R. J., Tufte, S. L.,
662: Kung, D. T., McCullough, P. M., \& Heiles, C. 1995, \apj, 448, 715
663:
664: \bibitem[Sciama(1998)]{sciama98} Sciama, D. W. 1998, \apj, 505, L35
665:
666: \bibitem[Slavin, McKee, \& Hollenbach(2000)]{smh00} Slavin, J. D.,
667: McKee, C. F., \& Hollenbach, D. J. 2000, \apj, 541, 218
668:
669: \bibitem[Spitzer \& Fitzpatrick(1993)]{sf93} Spitzer, L., Jr., \&
670: Fitzpatrick, E. L. 1993, \apj, 409, 299
671:
672: \bibitem[Tufte(1997)]{tufte97} Tufte, S. L. 1997, Ph.D. Thesis,
673: University of Wisconsin--Madison
674:
675: \bibitem[Tufte et al.(1998)]{tufte98} Tufte, S. L., Reynolds, R. J.,
676: \& Haffner, L. M. 1998, \apj, 504, 773
677:
678: \bibitem[Walters \& Cox(2001)]{wc01} Walters, M. A. \& Cox, D. P. 2001,
679: \apj, 549, 353
680:
681: \end{thebibliography}
682:
683:
684: \clearpage
685:
686: \begin{figure}[htbp]
687: \begin{center}
688: \plotone{f1.eps}
689: \caption{A spectrum of the faint atmospheric lines near \ha,
690: with the geocoronal line positioned on the red side of the
691: spectrum. The interstellar emission has been subtracted from
692: this spectrum. The weak atmospheric lines are fit with solid
693: lines and are numbered according to the line identifications in
694: Table~\ref{table:atmlines}. The bright geocoronal line at
695: --2~\kms\ with respect to the earth is also fit (the shift from
696: zero is a fine structure effect, resulting from the fact that
697: this line is not a recombination line). The solid line through
698: the data points shows the composite fit. A 15 \kms\ wide line
699: with an intensity of 0.1~R would have a peak height above
700: background of about 92 arbitrary data units (ADU) on the
701: intensity scale. The vertical scale has been expanded to show
702: the weak atmospheric lines. The peak of the geocoronal line is
703: at 4250~ADU.}
704: \label{fig:mlines}
705: \end{center}
706: \end{figure}
707:
708: \begin{figure}[htbp]
709: \begin{center}
710: \plotone{f2.eps}
711: \caption{A spectrum of faint atmospheric lines near \ha\
712: similar to that in Figure~\ref{fig:mlines}, except the
713: geocoronal line is positioned on the blue side of spectrum and
714: the observations were taken at a different time of the year. The
715: vertical scale matches that of Figure~\ref{fig:mlines}. The peak
716: of the geocoronal line is at 3800~ADU.}
717: \label{fig:elines}
718: \end{center}
719: \end{figure}
720:
721: \begin{figure}[htbp]
722: \begin{center}
723: \plotone{f3.eps}
724: \caption{\ha\ and \hi\ spectra toward the Lockman Window and
725: HD~93521. The \ha\ data are ``on-off'' subtraction spectra with
726: the residuals from the incompletely-subtracted geocoronal line
727: removed (see text). The intensity scale for the \ha\ spectra is
728: in milli-rayleighs (mR) per km~s$^{-1}$. A constant was
729: subtracted from the HD~93521 \ha\ spectrum in order to bring the
730: background level to zero. Arrows mark the velocity of the
731: detected emission components. The \hi\ spectra are from the
732: Leiden/Dwingeloo survey \citep{hb97}.}
733: \label{fig:main}
734: \end{center}
735: \end{figure}
736:
737: \begin{figure}[htbp]
738: \begin{center}
739: \plotone{f4.eps}
740: \caption{\oiw\ spectra toward HD~93521. The spectra are
741: ``on-off'' subtractions made from data gathered at three
742: different times of year (details given in text). The Gaussian
743: fit (solid line) through the data points in each panel is
744: centered at the radial velocity of the corresponding \ha\
745: emission (marked with an arrow). This Gaussian has a width of
746: 34~\kms\, and an intensity of 0.026~R (see text). A 34 \kms\
747: wide line with an intensity of 0.1~R would have a peak height
748: above the baseline of about 30 ADUs on this intensity scale. In
749: all three panels, the vertical scale has been expanded. In the
750: top panel, the incompletely-subtracted \oi\ airglow line, at
751: +28~\kms (LSR), has a peak height of 1450~ADU. In the middle
752: panel, the airglow line residual is at +14~\kms\ and peaks at
753: 3240~ADU. In the bottom panel, the airglow line residual is at
754: +~6~\kms\ and peaks at 825~ADU. The feature near --100~\kms\ in
755: the three spectra is the incompletely-subtracted atmospheric OH
756: line.}
757: \label{fig:oi}
758: \end{center}
759: \end{figure}
760:
761: \clearpage
762:
763: \begin{deluxetable}{ccccc}
764: \tablecolumns{5}
765: \tabletypesize{\scriptsize}
766: \tablewidth{0pt}
767: \tablecaption{List of \ha\ Observations\label{table:ha_obs}}
768: \tablehead{
769: \colhead{} & \colhead{} & \multicolumn{3}{c}{Total Exposure Time}\\
770: \colhead{} & \colhead{} & \multicolumn{3}{c}{(s)}\\
771: \cline{3-5}
772: \colhead{} & \colhead{} & \colhead{HD 93521} &
773: \colhead{Lockman Window} &
774: \colhead{Off A}\\
775: \colhead{Data Set} & \colhead{Observation Date} &
776: \colhead{($\ell = 183\fdg1$, $b = +62\fdg2$)} &
777: \colhead{($\ell = 148\fdg5$, $b = +53\fdg0$)} &
778: \colhead{($\ell = 163\fdg5$, $b = +53\fdg5$)}}
779: \startdata
780: 1a & 1997 Nov 5 & 1200 & 1200 & \nodata\\
781: 1b & 1997 Nov 7 & 1200 & 1200 & \nodata\\
782: 1c & 1997 Nov 8 & 600 & 600 & \nodata\\
783: 2a & 1998 Apr 22\tablenotemark{a} & \nodata & 1800 & \nodata\\
784: 2b & 1998 Apr 23\tablenotemark{a} & \nodata & 1800 & \nodata\\
785: 3 & 1999 Jan 18 & 3000 & 6000\tablenotemark{b} & \nodata\\
786: 4a & 1999 Apr 16 & \nodata & 600 & \nodata\\
787: 4b & 1999 Apr 20 & \nodata & 600 & \nodata\\
788: 5 & 1999 Dec 12 & \nodata & 3600 & 3600\\
789: 6 & 2000 Feb 2\tablenotemark{a} & \nodata & 3600 & 3600\\
790: 7 & 2000 Feb 4 & 3600 & \nodata & 3600\\
791: \enddata
792: \tablenotetext{a}{Spectra from this date have the geocoronal line
793: positioned on the blue side of the 200~\kms\ spectral range.}
794: \tablenotetext{b}{Individual spectra are 1200~s exposures.}
795: \end{deluxetable}
796:
797: \clearpage
798:
799: \begin{deluxetable}{ccccc}
800: \tablecolumns{5}
801: \tabletypesize{\scriptsize}
802: \tablewidth{0pt}
803: \tablecaption{List of \oiw\ Observations\label{table:oi_obs}}
804: \tablehead{
805: \colhead{} & \colhead{} &
806: \multicolumn{3}{c}{Total Exposure Time}\\
807: \colhead{} & \colhead{} & \multicolumn{3}{c}{(s)}\\
808: \cline{3-5}
809: \colhead{} & \colhead{} & \colhead{HD 93521} &
810: \colhead{Lockman Window} &
811: \colhead{Off A}\\
812: \colhead{Data Set} & \colhead{Observation Date} &
813: \colhead{($\ell = 183\fdg1$, $b = +62\fdg2$)} &
814: \colhead{($\ell = 148\fdg5$, $b = +53\fdg0$)} &
815: \colhead{($\ell = 163\fdg5$, $b = +53\fdg5$)}
816: }
817: \startdata
818: 8a & 1998 Nov 18 & 2400 & 1200 & \nodata\\
819: 8b & 1998 Nov 21 & 2400 & 1200 & \nodata\\
820: 9a & 1999 Jan 18 & 2400 & 2400\tablenotemark{a} & \nodata\\
821: 9b & 1999 Jan 22 & 2400 & 1200 & \nodata\\
822: 10a & 1999 Apr 16 & 1200 & 600 & \nodata\\
823: 10b & 1999 Apr 20 & 3600 & 1800 & \nodata\\
824: 11 & 2000 Feb 4 & 2400 & \nodata & 2400\\
825: \enddata
826: \tablenotetext{a}{Individual spectra are 1200~s exposures.}
827: \end{deluxetable}
828:
829:
830: \clearpage
831:
832: \begin{deluxetable}{cccccc}
833: \tablecolumns{6}
834: \tabletypesize{\small}
835: \tablewidth{0pt}
836: \tablecaption{Faint Atmospheric Lines near \ha\label{table:atmlines}}
837: \tablehead{
838: \colhead{} & \colhead{Wavelength} & \colhead{Width\tablenotemark{b}}
839: & \colhead{Adopted Intensity}
840: & \multicolumn{2}{c}{Measured Intensity\tablenotemark{c}}\\
841: \cline{5-6}
842: \colhead{Line\tablenotemark{a}} & \colhead{(\AA)}
843: & \colhead{(\kms)} & \colhead{(\% of Line 5 Intensity)}
844: & \multicolumn{2}{c}{(R)}
845: }
846: \startdata
847: 1 & 6559.55 & 10 & 54 & 0.070\tablenotemark{e} & \nodata\\
848: 2 & 6559.92 & 10 & 24 & 0.052 & \nodata\\
849: 3 & 6560.86 & 10 & 20 & 0.041 & \nodata\\
850: 4 & 6561.45 & 15 & 33 & 0.057 & 0.027\\
851: 5 & 6561.88 & 10 & 100 & 0.13 & 0.11\\
852: 6 & 6562.25 & 10 & 30 & 0.033 & \nodata\\
853: 7 & 6563.51 & 10 & 30 & 0.033 & 0.021\\
854: 7a & 6563.27\tablenotemark{d} & 10 & \nodata & \nodata & 0.038\\
855: 8 & 6563.69 & 10 & 65 & 0.088 & 0.045\\
856: 9 & 6563.98 & 10 & 39 & 0.048 & 0.058\\
857: 10 & 6564.39 & 15 & 160 & \nodata & 0.17\\
858: 11 & 6564.90 & 10 & 36 & \nodata & 0.032\\
859: 12 & 6565.44 & 10 & 29 & \nodata & 0.027\\
860: \enddata
861: \tablenotetext{a}{Line numbers correspond to labels in
862: Figure~\ref{fig:mlines} and Figure~\ref{fig:elines}.}
863: \tablenotetext{b}{Fixed widths were adopted. See discussion in
864: \S\ref{sec:atmlines}.}
865: \tablenotetext{c}{Left column refers to line intensities in
866: Figure~\ref{fig:mlines}, and right column refers to line intensities
867: in Figure~\ref{fig:elines}.}
868: \tablenotetext{d}{Wavelength was determined from the spectrum in
869: Figure~\ref{fig:elines} only.}
870: \tablenotetext{e}{Intensity in the fit was fixed at 54\% of Line 5
871: intensity.}
872: \end{deluxetable}
873:
874:
875: \clearpage
876:
877: \begin{deluxetable}{cccc}
878: \tablecolumns{4}
879: \tablewidth{0pt}
880: \tablecaption{\ha\ Results\label{table:ha_results}}
881: \tablehead{
882: \colhead{} & \colhead{V$_{\mathrm{LSR}}$} & \colhead{FWHM} &
883: \colhead{\emph{I}}\\
884: \colhead{Direction} & \colhead{(\kms)} & \colhead{(\kms)} & \colhead{(R)}}
885: \startdata
886: HD 93521 & --10 $\pm$ 4
887: & 22 $\pm$ 6 & 0.20 $\pm$ 0.07\\
888: & --51 $\pm$ 3 & 39 $\pm$ 7 & 0.15 $\pm$ 0.03\\
889: & --90 $\pm$ 5 & 25\tablenotemark{a} $\pm$ 5 & 0.023 $\pm$ 0.009\\
890: Lockman Window & --1 $\pm$ 4
891: & 30 $\pm$ 5 & 0.20 $\pm$ 0.05\\
892: & (--54)\tablenotemark{b,c} & (25) & $<$ 0.06\\
893: Off A & (0) & (25) & $<$ 0.11\\
894: & (--50) & (25) & $<$ 0.06 \\
895: \enddata
896: \tablenotetext{a}{Fixed at this value during the fitting.}
897: \tablenotetext{b}{Parentheses denote no \ha\ component detected.}
898: \tablenotetext{c}{Velocity of the \hi\ emission in the Lockman Window
899: spectrum shown in Figure~\ref{fig:main}.}
900: \end{deluxetable}
901:
902: \clearpage
903:
904: \begin{deluxetable}{ccccccc}
905: \tablecolumns{7}
906: \tablewidth{0pt}
907: \tablecaption{\oiw\ Results\label{table:oi_results}}
908: \tablehead{
909: \colhead{} & \colhead{V$_{\mathrm{LSR}}$} & \colhead{FWHM} &
910: \colhead{\emph{I}} &
911: \colhead{\emph{I}(\oi)/\emph{I}(\ha)} &
912: \multicolumn{2}{c}{\emph{n}(H$^+$)/\emph{n}(H$^{\circ}$)}\\
913: \cline{6-7}
914: \colhead{Direction} & \colhead{(\kms)} & \colhead{(\kms)} & \colhead{(R)} &
915: \colhead{(energy units)} & \colhead{6000 K} & \colhead{10,000 K}}
916: \startdata
917: HD 93521 & (--51)\tablenotemark{a,b} & (34)\tablenotemark{c} &
918: $<$0.060 & $<$0.43 &
919: $>$0.13 & $>$1.4\\
920: & (--51) & (34) & 0.026\tablenotemark{d} & 0.18 & 0.32 & 3.3\\
921: \enddata
922: \tablenotetext{a}{Parentheses denote no component detected.}
923: \tablenotetext{b}{Velocity of the \ha\ emission component.}
924: \tablenotetext{c}{Calculated based on the width of the \ha\ component
925: and a gas temperature of 8000 K.}
926: \tablenotetext{d}{Intensity of the gaussian fit in Figure~\ref{fig:oi}.}
927: \end{deluxetable}
928:
929:
930: \end{document}
931: % LocalWords: Lockman Hausen LSR etalon Fabry WHAM's sr sightline smh mjc wrt
932: % LocalWords: siderostat Fraunhofer airglow hb FWHM sightlines wc Pifer kuntz
933: % LocalWords: Quigley Tilleman AST eps ADU rayleighs mR ADUs
934: