Search for AU-Scale C I Structure in the Diffuse Interstellar Medium
Quick Summary (AKA AAS Abstract)
We present a multi-epoch, high resolution (R ~ 100,000) study of ultraviolet interstellar C I absorption line profiles taken by the Hubble Space Telescope. The 17 stars of this survey were chosen because each has high resolution spectra taken at least 10 years apart with the same instrument (STIS), grating (E140H), and aperture (0.2"x0.2") thus minimizing the instrumental differences between in the multi-epoch comparisons. Given the proper motions and distances of these stars, typically it was possible to observe variances in their C I line profiles which correspond to structure on scales of less that 200 AU. In 16 out of the 17 sightlines no significant differences in C I line profile between the two epochs were detected (75% of the 68 C I velocity components were measured to vary less than 20% at the 2 σ level). A measurement of ~ 5% of sightlines with variances is consistent with the fraction found recently by McEvoy et al. (2015) in their much larger survey of multi-epoch variance of Na I in optical spectra. However, the sightline toward HD210809 did show significant variance in its C I line profile.The C I absorption arising from both the J = 1 and J = 2 fine-structure states toward this star exhibits variations at an LSR velocity of -37 km/s indicative of C I structure on a scale less than 200 AU. Interestingly, the sky position of HD 210809 corresponds to the edge of an intervening H I supershell discovered by Suad et al. (2012) at the same LSR velocity. This connection is consistent with the optical survey of interstellar Na I by Meyer et al. (2015) who found that nearly all of their temporally-varying sightlines involved supernova remnants, H I supershells, or stellar bow shocks.
Back to top
Introduction: Structure in Cold Diffuse Interstellar Clouds
|
While many people imagine space as a mostly empty void, the space between stars is actually filled with the clouds of gas and dust that make up the interstellar medium. In fact, it is these clouds that contain the ingredients for new stars and are the site of star formation. Therefore, understanding the structure of these interstellar clouds can help lead to a better understanding of how stars are born. However, while it is easy to see small scale structure in bright objects of the ISM like emission and reflection nebulae, these features are not directly observable at optical wavelengths in cold diffuse clouds. Luckily, temporal variances in the absorption spectra of these clouds could reveal these unseen structures. The figure to the right depicts why this may be the case. We observe these cold diffuse clouds by looking at the background star's absorption spectra to see what light from the star the cloud blocks. Over time, the star moves a little bit relative to the Earth so the sightline points through a slightly different part of the cloud. If the cloud is very large, like the one on the left, you would not expect the sightline to be significantly different so the absorption spectra would not have any differences. However, sightlines change more significantly through thin structures than large ones making temporal variances in the spectra more likely. Other studies like those of Meyer et al. (2015, AAS, 141.23) and McEvoy et al. (2015, MNRAS, in press) have detected these temporal variations in optical Na I, but this study attempted to find similar variances in another trace neutral, C I, in ultraviolet data
Back to top
|
|
Methods: Measuring Absorption Spectra
|
For this survey, absorption spectra taken by the Hubble Space Telescope at two different times (epochs) for 17 stars were compared to look for temporal variations. These stars were chosen because they had high resolution data (R ~ 100,000) taken at least 10 years apart with the same instrument (STIS), grating (E140H), and aperture (0.2"x0.2") thus minimizing the instrumental differences in these comparisons. Line strengths at each epoch were measured using IRAF and compared for line profiles in the 1260-1350 angstrom wavelength region. The graph to the right shows some examples of lines from a sightline with no significant variation between the two epochs. This study focused on the trace neutral C I Multiplets 4 through 9. The first and second excited states of this element are labeled as C I * and C I ** in the graph respectively. CO and the trace neutrals Cl I and S I in this region were also compared. An example of a S I is also depicted to the right.
Back to top
|
|
Most Stars Had Little Absorption Variance
|
In 16 out of 17 sightlines of the survey, no significant C I absorption variance was seen. Additionally, there was no significant variance in the S I, Cl I, or CO line profiles in any of the 17 sightlines. Out of the 68 measured components, 75% were measured to vary less than 20% at the 2 σ level. The table at right shows how far each star moved between epochs given their proper motions and distances. This movement corresponds to an upper limit on the length scale of the intervening cloud structure of typically less than 200 AU. Note that many of the stars in the survey moved less than HD 210809 (highlighted in red) which did show significant variance. This could mean that the ISM in the direction of the other stars could have actually had small scale structure, but their sightlines did not change enough for variances to be measured.
Back to top
|
|
C I Variance Through a Hydrogen Shell
One sightline, towards HD 210809, did show significant variance in its C I profile. Its variances were seen in the fine structure J = 1 and J = 2 lines at an LSR velocity of -37 km/s, which can be seen in the two figures below. Note these plots contain not only data from the original set of two profiles (the Oct 1999 and Dec 2010 data) but also two additional profiles from April 2001 and Jan 2010 which had different resolutions. Thus, for comparison purposes, all data was smoothed to the lowest resolution. In both of the earlier spectra, the line strengths of the C I J = 1 and J = 2 high velocity lines are significantly larger than in later data, suggesting a change on the AU-scale in the density structure of the interstellar medium over this time frame.
|
Interestingly, the sky position of HD 210809 corresponds to an intervening H I shell at the same -37 km/s LSR velocity which was discovered by Suad et al. (2012, A&A, 538, A60) through their analysis of H 21 cm radiation and is shown in the figure on the right. Hydrogen shells occur when strong stellar winds or supernovae from massive stars at their center blow the surrounding gas and dust out into a thin shell structure. Since the sightline towards HD 210809 is near the edge of this shell it could explain temporal variations (see introduction figure).
Back to top
|
|
In our search for AU-scale structure in the cold, diffuse interstellar medium, only 1 out of 17 sightlines showed significant variation over time. This result of ~ 5% of sightlines with variation is consistent with the fraction found by McEvoy et al. (2015) in their much larger optical study. Our one case of variance in the C I lines corresponds to structure on a scale of less than 200 AU perhaps due to an intervening H I shell. This connection is consistent with an optical survey of interstellar Na I by Meyer et al. (2015) in which nearly all of their variances involved supernova remnants, H I shells, or stellar bow shocks.
Back to top
As of the time of completing this project, I am a rising junior at the University of Arkansas majoring in Physics with a minor in Mathematics. I absolutely love Astronomy; one of my favorite hobbies is sitting outside with a blanket and my little telescope. When I graduate I plan to go on to grad school with the hopes of getting a PhD in Astronomy or Astrophysics. As far as this project goes, it has been a lot of fun, and I'm quite sad that I will not be able to work on it all the time any more. Feel free to contact me at lmmarkwa (at) uark (dot) edu with any questions you may have, particularly about my project!
Back to top
Affiliations
Please contact me: lmmarkwa (at) uark (dot) edu
