Physical structure of the local interstellar medium
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Advances in Space Research 34 (2004) 41–45
www.elsevier.com/locate/asr
Physical structure of the local interstellar medium
S. Redfield *,1, B.E. Wood, J.L. Linsky
JILA, University of Colorado and NIST, 440 UCB, Boulder, CO 80309, USA
Received 19 October 2002; received in revised form 27 January 2003; accepted 13 February 2003
Abstract
The physical structure and morphology of the interstellar medium that surrounds our solar system directly effects the heliosphere
and the interplanetary environment. High resolution ultraviolet absorption spectra of nearby stars and the intervening interstellar
medium, observed by the Hubble Space Telescope, provide important information about the chemical abundance, ionization,
temperature, kinematics, density, morphology, and turbulent structures of the local interstellar medium. Fortunately, nearly all
observations of nearby stars contain useful local interstellar medium absorption lines. The number of useful observations is large
enough that we can start analyzing the local interstellar medium as a three-dimensional object, as opposed to focusing on individual
sightlines. We present the results of high resolution observations of nearby gas obtained by the Hubble Space Telescope. Our focus
will be on the kinematic, temperature, and turbulent velocity structures in the Local Interstellar Cloud and other nearby clouds.
Understanding the physical characteristics of these structures is necessary if we are to discuss the morphology of the local interstellar
medium, its evolution, origin, and impact on the heliosphere and our solar system.
Ó 2004 COSPAR. Published by Elsevier Ltd. All rights reserved.
Keywords: Local interstellar medium; Heliosphere; Turbulent velocity structure; Interplanetary environment
1. Introduction its environs and find ourselves in the ISM of the Milky
Way galaxy.
The Earth and the other planets in our solar system The LISM exists just beyond the boundary of our
reside within the heliosphere, a protective umbrella Solar System. In fact, the Voyager spacecraft are near-
supported by the solar wind. The boundary of the he- ing the boundary of the heliosphere on their way into
liosphere is determined by the location where the pres- the local interstellar environment (Miner, 1995). The
sure of the expanding solar wind precisely balances the proximity of the LISM is illustrated, not only by our
ambient pressure of the local interstellar medium ability to send spacecraft there, but also because the
(LISM). Just beyond the heliosphere lies warm partially LISM can effect our most immediate environment: the
ionized interstellar gas, with a temperature of 7000 K, Earth’s atmosphere. Many authors have speculated that
which we call the Local Interstellar Cloud (LIC). The the motion of the Sun through different interstellar en-
LIC is just one of a number of such clouds that exist in vironments might be the cause of quasi-periodic climate
the LISM. This collection of warm absorbers, in a catastrophes (i.e., a ‘‘snowball’’ Earth) and periodic
substrate of hot gas, terminates at the edge of what is mass extinctions of animals on Earth (Shapley, 1921;
called the Local Bubble, a low density cavity with a Begelman and Rees, 1976; Zank and Frisch, 1999).
radius of 100–200 pc, beyond which, cool, high-density Although scores of lines of sight through the LISM
interstellar material dominates. At this point, we have have been analyzed, typically with a few intervening
traveled beyond the provincial influence of the Sun and absorbers, it is difficult to identify coherent cloud
structures. So far, only a few coherent cloud structures
* have been identified (Lallement and Bertin, 1992; Lall-
Corresponding author.
E-mail address: sredfield@astro.as.utexas.edu (S. Redfield).
ement et al., 1995; Frisch et al., 2002). One is the LIC,
1
Currently a Harlan J. Smith Postdoctoral Fellow at McDonald the cloud that surrounds the Sun and solar system.
Observatory, University of Texas, Austin, TX 78712-0259, USA. Another is our closest cloud neighbor, in the Galactic
0273-1177/$30 Ó 2004 COSPAR. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.asr.2003.02.05342 S. Redfield et al. / Advances in Space Research 34 (2004) 41–45
Center direction, called the Galactic (G) Cloud. A three- ISM control star formation (Evans, 1999), the dynamics
dimensional model of the LIC indicates that the G of the ISM provides information on the stellar winds for
Cloud is less than 0.2 pc away (Redfield and Linsky, both early and late-type stars (Linsky and Wood, 1996;
2000). Our current knowledge of the G Cloud, the cloud Frisch, 1995), the ionization of the ISM provides in-
that will surround the Sun and our solar system in less formation on the interstellar radiation field (Vallerga,
than 7000 years, is limited to only two lines of sight: a 1998), and the chemical abundances and enrichment of
Cen (Linsky and Wood, 1996); and 36 Oph (Wood et the ISM provides information about the death of stars
al., 2000). The analysis of these lines of sight indicates and supernovae (McCray and Snow, 1979). Theoretical
that the physical properties of the G Cloud are different studies of the phases of the ISM have produced classic
than those of the LIC, with a significantly lower tem- papers in the astrophysical literature (e.g., Field et al.,
perature of T ¼ 5900 500 K, compared with 1969; McKee and Ostriker, 1977), the ideas of which are
T ¼ 7000 500 K for the LIC. The metal depletions are still being analyzed and discussed today (Cox, 1995;
also significantly different. The physical extent and Heiles, 2001).
spatial properties of the G Cloud remain unknown.
More observations through the G Cloud are needed to
satisfactorily characterize its morphology. 2. The 100 parsec database
Determining the three-dimensional structure of warm
partially ionized clouds in the LISM is a tractable The ultraviolet is the ideal wavelength region to study
problem. With enough lines of sight, the morphology of the LISM because many resonance lines are present,
an interstellar cloud can be deduced. The most com- including HI, DI,CII,NI, OI,MgI, MgII, SiII, Si III,
monly observed LISM cloud is, of course, the LIC, A1II, and FeII. The analysis of LISM absorption of
which directly surrounds the solar system. Using the various ions provides the opportunity not only to mea-
available observations of nearby stars though the LIC, sure the velocity structure along the line of sight, but to
we were able to determine the morphology of this cloud also measure the temperature, turbulent velocity, ioni-
(Redfield and Linsky, 2000). The extent of the LIC zation, abundances, and depletions. Because the kine-
varies from 5–8 pc and its mass is approximately 0.3M . matic structure along a line of sight that isS. Redfield et al. / Advances in Space Research 34 (2004) 41–45 43
chose to first analyze those absorption lines that provide
the best opportunity to accurately measure the kinematic
structure along a given line of sight: FeII, MgII, and CaII
(Redfield and Linsky, 2002). The velocity structure ob-
served in these lines can be applied to absorption features
of ions with more blended LISM absorption profiles.
Fig. 2 shows an example of LISM absorption in Mg II.
Because Mg II is a strong absorber, the relatively weak
absorption at )43 km s1 is easily detected.
The three-dimensional velocity structure of the LISM
has been studied by many researchers (Crutcher, 1982;
Lallement et al., 1995; Frisch et al., 2002). The initial
approach involves fitting a single velocity vector to
projected velocity measurements of the same cloud.
Fig. 1. Locations of all stars with high resolution LISM absorption
Implicit in this procedure is the assumption that indi-
observations within 100 pc in Galactic coordinates. The size of the
symbol is inversely proportional to the distance of the star, so closer vidual cloud structures in the LISM move at a unique
objects appear larger and more distant objects appear smaller. The single bulk velocity. This assumption was necessary
open circles indicate UV observations of MgII and/or FeII, and the when few velocity measurements were available, but
filled circles indicate optical Call observations. Sight lines described in now that our sky coverage is growing, there are indi-
Figs. 2 and 3 are labeled.
cations that a simple velocity vector may not adequately
describe the motion of clouds in the LISM. Redfield and
which on average corresponds to 1.6 absorbers per sight
Linsky (2001) analyzed the MgII absorption of a col-
line. A collection of all LISM observations is required to
lection of Hyades stars, which are located almost di-
understand the three-dimensional physical and mor-
rectly downwind of the LIC flow. The observed LIC
phological structure of the LISM.
absorption is consistently 3 km s1 slower than the
predicted LIC velocity. The G Cloud, in the upwind
direction of the LIC flow, is consistently faster than the
3. Velocity structure
predicted LIC velocity. Instead of individual clouds with
unique velocity vectors, these data suggest that the LIC
Many factors complicate an accurate measurement of
is being accelerated and decelerated, and a tensor, rather
the velocity structure of LISM absorption along a given
than a vector, may be required to adequately describe
sight line: (1) The intrinsic line width of the lightest ele-
the macroscopic motions of the gas in the LIC.
ments, such as HI, DI, and CII have broad thermal
widths, which leads to absorption line blending; (2) Weak
absorption lines make the continuum placement difficult, 4. Temperature and turbulent velocity structure
and the line fit is not well centered; (3) Observations with
low spectral resolution cannot disentangle a blended A comparison of the absorption line widths of ions
absorption line profile. Due to these complications, we with different atomic weights leads to a measurement of
Fig. 2. MgII h and k profiles of 70 Ophiuchi, only 5.1 pc away. The light solid line indicates the stellar continuum, the dashed lines show the three
velocity components, and the thick solid line is the final fit. A high velocity absorption component is observed at )43 km s1 . The strong absorption
at )26 km s1 is consistent with the predicted G Cloud velocity.44 S. Redfield et al. / Advances in Space Research 34 (2004) 41–45
Fig. 3. Comparison of line widths of various ions with different atomic weights. The shape of each curve is based on Eq. (1). The solid line is the best
fit line width, and the dashed lines indicate the l r error bars. Where the curves intersect is the best value for the temperature and turbulent velocity,
which is printed in the upper left corner. Only DI (the vertical line) and either MgII or FeII (the horizontal line) are shown for clarity. The projected
velocity of all these absorbers is consistent with the predicted LIC velocity. This demonstrates that there are temperature variations within the LIC.
the temperature and turbulent velocity of the absorbing In Fig. 3, all the projected cloud velocities agree with
cloud. The line width, or Doppler parameter ðbÞ. is re- the predicted LIC velocity. Therefore, they are mea-
lated to the temperature ðT Þ and turbulent velocity ðnÞ suring the same cloud structure along individual sight
of the cloud, in the following way: lines. The temperatures all agree fairly well with the
2kT T canonical LIC temperature of 7000 K, but the temper-
b2 ¼ þ n2 ¼ 0:016629 þ n2 ; ð1Þ atures vary from 5330 to 7400 K, indicating that tem-
m A
perature gradients may exist in the LIC. Likewise, the
where k is Boltzmann’s constant, m is the mass of the turbulent velocity varies from 1.3 to 2.1 km s1 . A full
ion observed, and A is the atomic weight of the element analysis of all sight lines is required to understand the
in atomic mass units. Often, we observe more than two distribution of these parameters. A full discussion of
ions, and therefore, we have an overconstrained system these measurements will be the subject of a future paper.
of equations. Fig. 3 shows how we incorporate all the
observed line widths to measure temperature and tur-
bulent velocity. In most cases, it is the lightest element 5. Conclusions
that can be reliably measured (DI) and the heaviest el-
ement, (Fe II or Mg II) that determine the two param- By analyzing and compiling the complete database of
eters. Intermediate mass ions typically confirm the high resolution LISM observations within 100 pc, we
measurements, but do not increase their accuracy. The are able to study the macroscopic structure of important
line width of hydrogen is not used because it is difficult physical parameters, such as kinematics, temperature,
to measure owing to its saturated line profile and po- and turbulent velocity. Understanding the full three-
tential contamination by heliospheric and astrospheric dimensional distribution of these quantities will answer
absorption. Eq. (1) assumes that all ions have the same fundamental questions regarding the morphology, ori-
kinetic temperature. Although this is a fair assumption gins, and evolution of clouds in our local environment.
for gas in the LISM, it does not hold for plasma in the We have demonstrated that variations in the tempera-
interplanetary medium (IPM). Because of the accelera- ture and turbulent velocity exist within the same cloud
tion and expansion of the solar wind, and wave-particle structure, specifically the LIC, which directly surrounds
interactions, IPM ions of different mass can have very our solar system. We have also shown that a single ve-
different kinetic temperatures. locity vector may no longer fully describe the motion ofS. Redfield et al. / Advances in Space Research 34 (2004) 41–45 45
the gas in the LIC. In fact, we observe systematic devi- Lallement, R., Bertin, P. Northern-hemisphere observations of nearby
ations from the velocity vector predictions in different interstellar gas: Possible detection of the local cloud. Astron.
Astrophys. 266, 479–485, 1992.
galactic directions, specifically in the upwind and Lallement, R., Ferlet, R., Lagrange, A.M., et al. Local cloud structure
downwind directions. The implications of these varia- from HST-GHRS. Astron. Astrophys. 304, 461–474, 1995.
tions could have a direct impact on the details of the Linsky, J.L., Wood, B.E. The alpha Centauri line of sight: D/H ratio,
interaction between our solar system and the LISM, physical properties of local interstellar gas, and measurement of
through the heliospheric interface. heated hydrogen (the ‘‘Hydrogen Wall’’) near the heliopause.
Astrophys. J. 463, 254–270, 1996.
McCray, R., Snow, T.P. The violent interstellar medium. Ann. Rev.
Astron. Astrophys. 17, 213–240, 1979.
Acknowledgements McKee, C.F. The Local Bubble and Beyond: Summary, in: Bre-
itschwerdt, D., Freyberg, M.J., Tr€ umper, J. (Eds.), The Local
This research is supported by NASA Grants NGT- Bubble and Beyond. Proceedings of IAU Colloquia, 166. Springer-
50242 and 5-56500-D and Space Telescope Science In- Verlag, Berlin, pp. 565–580, 1998.
McKee, C.F., Ostriker, J.P. A theory of the interstellar
stitute grant AR-09525 to the University of Colorado at
medium – Three components regulated by supernova explo-
Boulder. sions in an inhomogeneous substrate. Astrophys. J. 218,
148–169, 1977.
Miner, E.D. What’s new, Voyager: The discoveries continue. Planetary
References Report 15, 4–9, 1995.
Redfield, S., Linsky, J.L. The three-dimensional structure of the warm
Begelman, M.C., Rees, M.J. Can cosmic clouds cause climatic local interstellar medium. II. The Colorado Model of the Local
catastrophes? Nature 261, 298–299, 1976. Interstellar Cloud. Astrophys. J. 534, 825–837, 2000.
Cox, D. Interstellar medium – Crumbling canons. Nature 375, 185– Redfield, S., Linsky, J.L. Microstructure of the Local Interstellar
186, 1995. Cloud and the identification of the Hyades Cloud. Astrophys. J.
Crutcher, R.M. The local interstellar medium. Astrophys. J. 254, 82– 551, 413–428, 2001.
87, 1982. Redfield, S., Linsky, J.L. The structure of the local interstellar
Evans, N.J. Physical conditions in regions of star formation. Ann. medium. I. High-resolution observations of FeII, MgII, and CaII
Rev. Astron. Astrophys. 37, 311–362, 1999. toward stars within 100 parsecs. Astrophys. J. Suppl. Ser. 139, 439–
Field, G.B., Goldsmith, D.W., Habirig, H.J. Cosmic-ray heating of the 465, 2002.
interstellar gas. Astrophys. J. 155, L149–L154, 1969. Shapley, H. Note on a possible factor in changes of geological climate.
Frisch, P.C. Characteristics of nearby interstellar matter. Space Sci. J. Geology 29, 502–504, 1921.
Rev. 72, 499–592, 1995. Vallerga, J. The stellar extreme-ultraviolet radiation field. Astrophys.
Frisch, P.C., Grodnicki, L., Welty, D.E. The velocity distribution of J. 497, 921–927, 1998.
the nearest interstellar gas. Astrophys. J. 574, 834–846, 2002. Wood, B.E., Linsky, J.L., Zank, G.P. Heliospheric, astrospheric, and
Heiles, C. The McKee/Ostriker model: Paradigm?Woodward, C.E., interstellar Ly-alpha; Absorption toward 36 Ophiuchi. Astrophys.
Bicay, M.D., Shull, J.M. (Eds.), Tetons 4: Galactic Structure, J. 537, 304–311, 2000.
Stars, and the Inter stellar Medium. ASP Conference Series, 231. Zank, G.P., Frisch, P.C. Consequences of a change in the Galactic
ASP, San Francisco, pp. 294–316, 2001. environment of the Sun. Astrophys. J. 518, 965–973, 1999.You can also read
