Interstellar_medium - Wikiparency

# Interstellar medium  **Microsim — three.js (Wikitube framework):** *Out to the nearest star (Solar System explorer)* <div class="wt-sim" data-src="https://wikitube-3d-microsims.netlify.app/solar/Solar_System.html?view=neighborhood&embed=1" data-title="Out to the nearest star (Solar System explorer)"></div> *The Solar System explorer locked on this article's state (`?view=neighborhood`); every object and population of the [[PORTAL_Solar_System|Solar System portal]] has its own state in the same scene.*   *Try: find the heliopause tick at 121 AU at the inner end of the ruler, the point where Voyager 1 entered the interstellar medium, and follow the ruler out to Proxima Centauri at 4.25 light-years; set show to clouds and boundaries to leave the heliosphere shells and the Oort cloud points on their own; press l to hide the labels and scroll in on the shells to see how small the Sun's bubble is against the ruler.* The **interstellar medium** (ISM) is the gas, dust, cosmic rays and radiation that fill the space between the stars of a galaxy.[^ferriere2001] Its gas is found in ionized, atomic and molecular forms, and by number of atoms it is about 91 percent [[Hydrogen|hydrogen]], 8.9 percent [[Helium|helium]] and 0.1 percent heavier elements; by mass, about 99 percent is gas and 1 percent dust.[^ferriere2001][^boulanger2000] Densities range from about 10⁻⁴ particles per cubic centimetre in the hottest gas to about a million in the densest molecular clouds, all far below the best laboratory vacuum, yet collisions are frequent enough over interstellar distances that the medium behaves as a gas, and because it is at least slightly ionized, as a [[Plasma_(physics)|plasma]].[^ferriere2001][^spitzer1978] The medium links stars and galaxies. Stars form in its densest clouds and return matter and energy to it through winds, planetary nebulae and supernovae, so the medium sets how long a galaxy can go on making stars.[^herbst1995][^mckee1977] It also begins, from the Sun's point of view, at the [[Heliopause|heliopause]]. [[Voyager_1|Voyager 1]] crossed that boundary on 25 August 2012 and Voyager 2 on 5 November 2018, the first spacecraft to sample the medium directly.[^jpl-voyager] The explorer at the top of this page shows the start of the medium but not the medium itself. Its neighbourhood view marks the heliopause at 121 AU on a ruler that ends at Proxima Centauri, 4.25 light-years out; no interstellar gas is drawn, and the nearest cloud, the [[Local_Interstellar_Cloud|Local Interstellar Cloud]], is many times larger than the ruler. ## Interstellar matter Interstellar matter is not one substance at one temperature but a set of components that differ by factors of about a thousand in temperature and ten billion in density, from molecular gas colder than 20 K to coronal gas above a million kelvin. They are told apart by the state of their [[Hydrogen|hydrogen]], molecular, atomic or ionized, and each is seen through a different tracer, from the 21-cm line of atomic hydrogen to X-rays from the hottest gas.[^ferriere2001] | Component | Share of volume | Temperature (K) | Density (particles/cm³) | Hydrogen | Main tracer | |---|---|---|---|---|---| | Molecular clouds | under 1% | 10–20 | 10²–10⁶ | molecular | radio and infrared molecular lines | | Cold neutral medium | 1–5% | 50–100 | 20–50 | atomic | 21-cm absorption | | Warm neutral medium | 10–20% | 6,000–10,000 | 0.2–0.5 | atomic | 21-cm emission | | Warm ionized medium | 20–50% | about 8,000 | 0.2–0.5 | ionized | Hα emission, pulsar dispersion | | H II regions | under 1% | about 8,000 | 10²–10⁴ | ionized | Hα, radio recombination lines | | Hot ionized (coronal) gas | 30–70% | 10⁶–10⁷ | 10⁻⁴–10⁻² | ionized | X-rays, ultraviolet absorption | Components of the interstellar medium of the Milky Way, after Ferrière.[^ferriere2001] ### The three-phase model In 1969 Field, Goldsmith and Habing explained the ISM as two phases in pressure balance: cold, dense clouds below about 300 K and a warm intercloud medium near 10,000 K, the two temperatures at which heating and cooling can balance stably.[^field1969] McKee and Ostriker added in 1977 a third, hot phase near a million kelvin, heated by supernova shocks and filling most of the volume.[^mckee1977] The balance works because pressure is proportional to density times temperature. With the table's values, the warm neutral gas at 0.3 per cubic centimetre and 8,000 K, cold gas at 30 per cubic centimetre and 70 K, and coronal gas at 0.003 per cubic centimetre and a million kelvin all give a product of roughly 2,000 to 3,000 K per cubic centimetre (derived). The relative shares of the phases remain uncertain.[^ferriere2001] [[Hydrogen]] sets the structure. Hot, thin gas radiates little and can stay hot for hundreds of millions of years; once gas cools toward 10⁵ K and becomes denser, electrons and protons recombine and radiate, and cooling runs away. Near hot O and B stars, photons above 13.6 eV, the [[Ionization_energy|ionization energy]] of hydrogen, keep the gas ionized at about 8,000 K out to the point where they are used up, the edge of an H II region. Lower-energy ultraviolet light passes on, ionizing [[Carbon|carbon]] and breaking apart molecules to form photodissociation regions. When massive stars form inside a dense cloud, their H II regions expand, and when the stars explode a few million years later, their blast waves heat the gas back to the coronal phase.[^ferriere2001][^lequeux2005] ### The ISM in different kinds of galaxy In a spiral galaxy such as the [[Milky_Way|Milky Way]] nearly all the interstellar mass lies in a thin disc, with a scale height of about 100 parsecs against a diameter of some 30,000 parsecs, and orbits the centre at about 200 km/s. Differential rotation shears clouds and magnetic field lines into long arcs, and spiral arms compress the gas and trigger star formation, which is why H II regions line the arms. Irregular galaxies have a similar but less ordered medium; elliptical galaxies hold mostly hot coronal gas and form few stars.[^lequeux2005][^ferriere2001] ### Structures The ISM is [[Turbulence|turbulent]], and its bulk motions are usually faster than the local [[Speed_of_sound|speed of sound]]. Colliding flows form shocks that compress and heat the gas, and the magnetic field adds Alfvén waves, described by [[Magnetohydrodynamics|magnetohydrodynamics]], that can be faster still. Winds and supernovae from young star clusters blow bubbles and superbubbles of hot gas that X-ray telescopes can see.[^mckee1977] Stars, once formed, are no longer held by gas pressure, so they drift through the medium. The Sun is now inside the Local Interstellar Cloud, a warm clump a few parsecs across, which sits inside the [[Local_Bubble|Local Bubble]], a cavity of hot, thin gas about 100 parsecs in radius.[^frisch2011] Both Voyagers have measured the electron density just beyond the heliopause rising with distance, a gradient that seems to be a large-scale feature ahead of the heliosphere.[^kurth2020] ### Interaction with interplanetary medium The [[Solar_wind|solar wind]] fills the [[Heliosphere|heliosphere]] with outflowing plasma until the pressure of the interstellar gas stops it. It becomes subsonic at the termination shock, about 90 to 100 AU from the Sun; Voyager 1 crossed that shock in December 2004 at 94 AU and entered the heliosheath, where solar and interstellar material interact.[^stone2005] The heliopause, crossed in 2012, is where the Sun's plasma ends and the interstellar plasma begins.[^jpl-voyager] ### Interstellar extinction Dust dims and reddens starlight because small grains scatter and absorb blue light more strongly than red; the effect weakens steadily toward longer wavelengths and is almost gone in the mid-infrared beyond about 5 μm. The dark rifts in the band of the Milky Way are dust clouds a few thousand light-years away.[^lequeux2005] With Gaia distances to millions of stars, the reddening of each star can be located along its line of sight, and by 2022 three-dimensional dust maps reached 3 kiloparsecs from the Sun.[^vergely2022] Neutral hydrogen absorbs far-ultraviolet light strongly at the Lyman-α line, 121.6 nm, and completely below the Lyman limit at 91.2 nm (13.6 eV), so stars more than a few hundred light-years away cannot be seen at those wavelengths; the medium becomes transparent again only in soft X-rays.[^ferriere2001] ## Heating and cooling Interstellar gas is far from thermodynamic equilibrium. Collisions give the particles a Maxwell–Boltzmann spread of speeds, so a kinetic temperature can be defined, but the radiation field is weak, roughly that of a star of about 10,000 K diluted enormously, so the energy levels of atoms and molecules are not populated as they would be in equilibrium.[^spitzer1978] The temperature of each region is set by the balance between the heating and cooling processes that dominate there.[^lequeux2005] ### Heating mechanisms The first heating source proposed was low-energy cosmic rays, which penetrate even dense clouds and deposit energy by ionizing and exciting atoms; the particles of a few MeV matter most because they are far more numerous than energetic ones.[^field1969] In the diffuse neutral medium the main source is the [[Photoelectric_effect|photoelectric effect]] on dust: ultraviolet [[Photon|photons]] knock electrons out of small grains, and the electrons share their leftover [[Kinetic_energy|kinetic energy]] with the gas. Because grain numbers rise steeply toward small sizes, following the Mathis–Rumpl–Nordsieck distribution n(r) ∝ r⁻³·⁵, the smallest grains supply most of the surface area and most of this heating.[^mathis1977][^weingartner2001] Other processes dominate elsewhere. Photoionization of hydrogen heats H II regions, and X-rays heat warm, low-density atomic gas. When two hydrogen atoms meet on a grain surface and form a hydrogen molecule, 4.48 eV is released, part of which ends up heating the gas. Deep in molecular clouds, gas and dust exchange heat by collision; Burke and Hollenbach measured the efficiency of that exchange as an accommodation coefficient of about 0.35.[^burke1983] On larger scales, gravitational collapse, supernovae, stellar winds, expanding H II regions and magnetohydrodynamic waves all feed energy into the gas.[^mckee1977] ### Cooling mechanisms Most interstellar gas cools by emitting in fine-structure lines. Collisions lift atoms and ions such as ionized [[Carbon|carbon]] and neutral [[Oxygen|oxygen]] in neutral gas, or ionized oxygen, [[Nitrogen|nitrogen]] and [[Neon|neon]] in H II regions, into levels just above their ground state; the atoms then decay by emitting photons that escape and carry the energy away. In warmer gas, collisions can also excite permitted lines such as hydrogen's Lyman-α. In molecular clouds, rotational lines of carbon monoxide take over.[^lequeux2005][^spitzer1978] ## Observations of the ISM Thin as it is, the ISM shines in almost every band of the electromagnetic spectrum, except, ironically, the optical band on which astronomers relied until the 20th century.[^lequeux2005] Ionized gas emits bremsstrahlung, seen in microwaves for gas near 10⁴ K and in soft X-rays for the coronal gas, together with many [[Spectral_line|spectral lines]]. The green glow of many nebulae, for example, is a forbidden transition of [[Oxygen|oxygen]] ions that have lost two electrons, and it was at first credited to a new element named "nebulium". Recombination lines of highly excited hydrogen reach into the radio, where dust does not absorb them, and so trace ionized gas across the whole Galactic disc.[^lequeux2005] Neutral hydrogen is mapped through the [[Hydrogen_line|21-cm line]] at 1,420 MHz, emitted mostly by the warm neutral medium; all-sky surveys such as HI4PI and the Arecibo GALFA-HI survey chart it at high resolution.[^benbekhti2016][^peek2017] Molecular clouds are seen mainly through rotational lines of carbon monoxide, above all the 115 GHz line, because molecular hydrogen itself hardly radiates at cloud temperatures. Some hydrogen is molecular yet shows no carbon monoxide; this "dark gas" was revealed by the Planck satellite as dust emission without matching line emission.[^planck2011] Dust re-emits absorbed starlight as nearly [[Black-body_radiation|black-body]] radiation in the far infrared at 20–100 K; polycyclic aromatic hydrocarbons emit bands in the mid-infrared, and tiny spinning grains are the likely source of anomalous microwave emission.[^lequeux2005] Cosmic rays striking gas produce gamma rays, and cosmic-ray electrons spiralling in the galactic magnetic field emit synchrotron radio waves.[^ferriere2001] ## Radiowave propagation The ISM is a plasma, and a plasma cannot carry waves below its plasma frequency, f_p ≈ 8.98 kHz × √n_e, with n_e the electron density per cubic centimetre. For the density Voyager 1 measured beyond the heliopause, about 0.08 per cubic centimetre, f_p ≈ 2.5 kHz (derived), close to the 2–3 kHz oscillations its plasma wave instrument recorded.[^gurnett2013] The longest radio waves astronomers have recorded from beyond Earth, around 1 km, can travel only 10 to 50 parsecs through the Local Bubble before free–free absorption stops them.[^novaco1978] Above the plasma frequency, the refractive index depends on frequency and on the density of free electrons, with three useful consequences. Pulses from pulsars and fast radio bursts arrive later at low frequencies, by 4.15 ms × DM × (f/1 GHz)⁻², where the dispersion measure DM is the column of electrons in parsecs per cubic centimetre; a pulsar with a DM of 30 is delayed by about 0.8 s at 400 MHz relative to high frequencies (derived). Pulsar dispersion therefore maps the ionized gas and gives distances.[^lorimer2005] Random density variations make radio sources scintillate and broaden their images, more strongly at low frequencies. And the plane of linearly polarized waves, such as synchrotron emission, rotates by an amount set by the electron density and the magnetic field (Faraday rotation), which makes it the main probe of the interstellar field.[^ferriere2001] Otherwise the medium is transparent to radio waves, which is why radio astronomy can see across the Galaxy. The exceptions are the strongest lines, such as carbon monoxide at millimetre wavelengths and the 21-cm line in cold gas, and dense ionized regions, which become opaque to their own bremsstrahlung at long wavelengths; at metre wavelengths H II regions appear as dark patches against the synchrotron background.[^novaco1978][^lequeux2005] ## History of knowledge of interstellar space Francis Bacon used the word "interstellar" in the 1620s, when the stars were still imagined on a sphere; later in the 17th century, with stars scattered through infinite space, writers such as Robert Boyle debated whether that space was empty or filled with a fluid, and in the 19th century the luminiferous aether was invoked to carry light between the stars.[^bacon1627][^boyle1674] Evidence began with [[Spectroscopy|spectroscopy]]. In 1864 William Huggins found that the spectrum of a nebula consisted of bright lines, showing it to be gas rather than unresolved stars.[^skytel2014] From 1889 Edward Barnard photographed dark "holes" in the Milky Way, and by 1899 he argued that they must be clouds of matter in front of the stars; they are now known as dark nebulae.[^barnard1899] The first direct detection of diffuse interstellar gas came in 1904. Johannes Hartmann, studying the spectroscopic binary Delta Orionis at Potsdam, saw that the [[Calcium|calcium]] K line at 393.4 nm did not shift back and forth with the star's orbit like the other lines, and concluded that it came from a cloud of gas somewhere along the line of sight.[^hartmann1904] Mary Lea Heger found stationary [[Sodium|sodium]] D lines in 1919, and Beals showed in 1936 that interstellar lines are often double or asymmetric: each is a blend of lines from several clouds moving at different Doppler velocities, the first evidence that the medium is clumpy.[^heger1919][^beals1936] Victor Hess's discovery of cosmic rays in 1912 led Kristian Birkeland to propose that space was full of electrons and ions thrown off by stars.[^birkeland1913] Recent work has turned to interstellar chemistry. In 2012 laboratory experiments showed that polycyclic aromatic hydrocarbons in ice under interstellar conditions are converted into more complex organic molecules.[^gudipati2012] Hubble spectra confirmed ionized buckminsterfullerene, C₆₀⁺, in interstellar space in 2019, and in 2020 evidence was presented for water ice mixed with silicate dust grains in the diffuse medium.[^cordiner2019][^potapov2021] ## See also - [[Local_Interstellar_Cloud]] · [[Local_Bubble]] - [[Heliosphere]] - [[Milky_Way]] - [[Interplanetary_dust_cloud]] - Molecular cloud · Diffuse interstellar band · Photodissociation region · List of interstellar and circumstellar molecules ## References ### Citations [^ferriere2001]: Ferrière, K. M. (2001). "The interstellar environment of our galaxy". *Reviews of Modern Physics* 73: 1031–1066. https://doi.org/10.1103/RevModPhys.73.1031 [^boulanger2000]: Boulanger, F.; Cox, P.; Jones, A. P. (2000). "Course 7: Dust in the interstellar medium". In Casoli, F.; Lequeux, J.; David, F. (eds.), *Infrared Space Astronomy, Today and Tomorrow*. Les Houches Summer School 70, p. 251. Bibcode 2000isat.conf..251B. [^spitzer1978]: Spitzer, L. (1978). *Physical Processes in the Interstellar Medium*. Wiley. ISBN 978-0-471-29335-4. [^herbst1995]: Herbst, E. (1995). "Chemistry in the interstellar medium". *Annual Review of Physical Chemistry* 46: 27–54. https://doi.org/10.1146/annurev.pc.46.100195.000331 [^mckee1977]: McKee, C. F.; Ostriker, J. P. (1977). "A theory of the interstellar medium: three components regulated by supernova explosions in an inhomogeneous substrate". *The Astrophysical Journal* 218: 148–169. https://doi.org/10.1086/155667 [^jpl-voyager]: Jet Propulsion Laboratory. "Voyager – Interstellar mission". https://voyager.jpl.nasa.gov/mission/interstellar-mission/ [^field1969]: Field, G. B.; Goldsmith, D. W.; Habing, H. J. (1969). "Cosmic-ray heating of the interstellar gas". *The Astrophysical Journal* 155: L149. https://doi.org/10.1086/180324 [^lequeux2005]: Lequeux, J. (2005). *The Interstellar Medium*. Astronomy and Astrophysics Library. Springer. ISBN 978-3-540-21326-0. https://doi.org/10.1007/b137959 [^frisch2011]: Frisch, P. C.; Redfield, S.; Slavin, J. D. (2011). "The interstellar medium surrounding the Sun". *Annual Review of Astronomy and Astrophysics* 49: 237–279. https://doi.org/10.1146/annurev-astro-081710-102613 [^kurth2020]: Kurth, W. S.; Gurnett, D. A. (2020). "Observations of a radial density gradient in the very local interstellar medium by Voyager 2". *The Astrophysical Journal Letters* 900: L1. https://doi.org/10.3847/2041-8213/abae58 [^stone2005]: Stone, E. C.; Cummings, A. C.; McDonald, F. B.; Heikkila, B. C.; Lal, N.; Webber, W. R. (2005). "Voyager 1 explores the termination shock region and the heliosheath beyond". *Science* 309: 2017–2020. https://doi.org/10.1126/science.1117684 [^vergely2022]: Vergely, J. L.; Lallement, R.; Cox, N. L. J. (2022). "Three-dimensional extinction maps: inverting inter-calibrated extinction catalogues". *Astronomy & Astrophysics* 664: A174. https://doi.org/10.1051/0004-6361/202243319 [^mathis1977]: Mathis, J. S.; Rumpl, W.; Nordsieck, K. H. (1977). "The size distribution of interstellar grains". *The Astrophysical Journal* 217: 425–433. https://doi.org/10.1086/155591 [^weingartner2001]: Weingartner, J. C.; Draine, B. T. (2001). "Photoelectric emission from interstellar dust: grain charging and gas heating". *The Astrophysical Journal Supplement Series* 134: 263–281. https://doi.org/10.1086/320852 [^burke1983]: Burke, J. R.; Hollenbach, D. J. (1983). "The gas-grain interaction in the interstellar medium: thermal accommodation and trapping". *The Astrophysical Journal* 265: 223–234. https://doi.org/10.1086/160667 [^benbekhti2016]: HI4PI Collaboration; Ben Bekhti, N.; Flöer, L.; et al. (2016). "HI4PI: a full-sky H I survey based on EBHIS and GASS". *Astronomy & Astrophysics* 594: A116. https://doi.org/10.1051/0004-6361/201629178 [^peek2017]: Peek, J. E. G.; Babler, B. L.; Zheng, Y.; et al. (2018). "The GALFA-H I survey data release 2". *The Astrophysical Journal Supplement Series* 234: 2. https://doi.org/10.3847/1538-4365/aa91d3 [^planck2011]: Planck Collaboration; Ade, P. A. R.; Aghanim, N.; et al. (2011). "Planck early results. XIX. All-sky temperature and dust optical depth from Planck and IRAS: constraints on the 'dark gas' in our Galaxy". *Astronomy & Astrophysics* 536: A19. https://doi.org/10.1051/0004-6361/201116479 [^gurnett2013]: Gurnett, D. A.; Kurth, W. S.; Burlaga, L. F.; Ness, N. F. (2013). "In situ observations of interstellar plasma with Voyager 1". *Science* 341: 1489–1492. https://doi.org/10.1126/science.1241681 [^novaco1978]: Novaco, J. C.; Brown, L. W. (1978). "Nonthermal galactic emission below 10 megahertz". *The Astrophysical Journal* 221: 114–123. https://doi.org/10.1086/156009 [^lorimer2005]: Lorimer, D. R.; Kramer, M. (2005). *Handbook of Pulsar Astronomy*. Cambridge University Press, ch. 4. ISBN 978-0-521-82823-9. [^bacon1627]: Bacon, F. (1627). *Sylva Sylvarum, or A Natural History in Ten Centuries*. W. Lee, §§ 354–455. [^boyle1674]: Boyle, R. (1674). *The Excellency of Theology Compar'd with Natural Philosophy*. Henry Herringman, p. 178. https://quod.lib.umich.edu/e/eebo/A28966.0001.001/1:1?rgn=div1;view=toc [^skytel2014]: *Sky & Telescope* (14 August 2014). "The first planetary nebula spectrum". https://www.skyandtelescope.com/astronomy-news/observing-news/first-planetary-nebula-spectrum-08142014/ [^barnard1899]: Barnard, E. E. (1899). "Photographs of comets and of the Milky Way". *Monthly Notices of the Royal Astronomical Society* 59: 361–370. [^hartmann1904]: Hartmann, J. (1904). "Investigations on the spectrum and orbit of delta Orionis". *The Astrophysical Journal* 19: 268–286. https://doi.org/10.1086/141112 [^heger1919]: Heger, M. L. (1919). "Stationary sodium lines in spectroscopic binaries". *Publications of the Astronomical Society of the Pacific* 31: 304–305. https://doi.org/10.1086/122890 [^beals1936]: Beals, C. S. (1936). "On the interpretation of interstellar lines". *Monthly Notices of the Royal Astronomical Society* 96: 661–678. https://doi.org/10.1093/mnras/96.7.661 [^birkeland1913]: Birkeland, K. (1913). "Polar magnetic phenomena and terrella experiments". In *The Norwegian Aurora Polaris Expedition, 1902–1903*, vol. 1, section 2. H. Aschehoug & Co., p. 720. https://archive.org/details/norwegianaurorap01chririch [^gudipati2012]: Gudipati, M. S.; Yang, R. (2012). "In-situ probing of radiation-induced processing of organics in astrophysical ice analogs: novel laser desorption laser ionization time-of-flight mass spectroscopic studies". *The Astrophysical Journal Letters* 756: L24. https://doi.org/10.1088/2041-8205/756/1/L24 [^cordiner2019]: Cordiner, M. A.; Linnartz, H.; Cox, N. L. J.; et al. (2019). "Confirming interstellar C₆₀⁺ using the Hubble Space Telescope". *The Astrophysical Journal Letters* 875: L28. https://doi.org/10.3847/2041-8213/ab14e5 [^potapov2021]: Potapov, A.; Bouwman, J.; Jäger, C.; Henning, T. (2021). "Dust/ice mixing in cold regions and solid-state water in the diffuse interstellar medium". *Nature Astronomy* 5: 78–85. https://doi.org/10.1038/s41550-020-01214-x ### Sources Survey data: the Wisconsin Hα Mapper northern sky survey (Haffner, L. M.; et al. 2003, *The Astrophysical Journal Supplement Series* 149: 405) and the all-sky 21-cm surveys cited above. ## External links - NASA Science, "Interstellar Boundary Explorer (IBEX)": https://science.nasa.gov/mission/ibex/ - NASA/JPL Voyager interstellar mission: https://voyager.jpl.nasa.gov/mission/interstellar-mission/ - Ferrière, K. M. (2001), review of the Galactic interstellar environment (open access): https://arxiv.org/abs/astro-ph/0106359 *Article prose: Solar System portal child articles, wave 1, 2026-09-18 · row SOL-091 · strict pair pinned at revision [1373528401](https://en.wikipedia.org/w/index.php?oldid=1373528401). The sections below this block are the page's earlier material, kept in place.*   ## Microsims — three.js ### Interstellar medium (three.js) <div class="microsim-player"> <iframe src="https://wikitube-3d-microsims.netlify.app/Interstellar_medium.html" width="100%" height="620" frameborder="0" loading="lazy" sandbox="allow-scripts allow-same-origin" title="Interstellar medium — three.js microsim"></iframe> </div> **Open it full-screen:** [Interstellar_medium.html](https://wikitube-3d-microsims.netlify.app/Interstellar_medium.html) · library `threejs` · route `microsim/threejs/` ### Related microsims Live sims on neighbouring articles: - [[Hydronium]] - [[Haber_process]] - [[Electrolysis_of_water]] - [[Fuel_cell]] *Sim hosted off-article; the article owns the reference, not the runtime (WIKI_RULES §10.4). Placed by `g08_place_microsims.py`.*  ## Overview The gas between the stars is neither empty nor uniform, and it is overwhelmingly hydrogen -- about 91 per cent by number, 9 per cent helium, the rest at the 0.1 per cent level. What it is not is one substance at one temperature. It is sorted into distinct thermal phases spanning eight orders of magnitude in temperature and in density, from 10 K molecular clumps at 10^6 particles per cubic centimetre to million-kelvin cavities at 10^-3. The reason to look closely is one inversion. Rank the phases by how much of the *room* they take and by how much of the *matter* they hold, and the two orders come out nearly opposite: molecular gas is a large share of the mass in a tiny share of the volume, the hot ionised medium the reverse -- nearly all of the room, almost none of the matter. Flipping that weighting is what the sim exists to show. ## The physics Five phases, with representative values: phase T (K) n (cm^-3) volume mass Molecular 10-20 1e2-1e6 ~1% ~20% CNM (H I) 50-100 20-50 ~1-4% ~30% WNM (H I) 6000-10000 0.2-0.5 ~30-40% ~30% WIM (H II) ~8000 0.2-0.5 ~20-25% ~15% HIM ~1e6 1e-3-1e-2 ~30-50% ~1% They coexist because they sit at roughly the same thermal pressure. Density and temperature trade off along a line of constant nT, near 3000 to 4000 K cm^-3 in the solar neighbourhood, so a 15 K molecular clump and a million-kelvin supernova cavity can share a wall and neither wins. Field, Goldsmith and Habing (*ApJ* **155**, L149, 1969) showed why two are stable. Gas heated at a rate roughly independent of density and cooled as density squared has an S-shaped equilibrium curve: over a range of pressures *two* stable solutions exist at one pressure, a cold dense one and a warm rarefied one, separated by an unstable branch where dP/dn is negative. That is the CNM and the WNM. The sim solves it live rather than tabulating it, on the Koyama and Inutsuka (2002) fit to photoelectric heating against C II and Lyman-alpha cooling, Lambda(T)/Gamma = 1e7 exp(-1.184e5/(T+1000)) + 1.4e-2 sqrt(T) exp(-92/T) with equilibrium at n_eq(T) = 1/[bracket]. The pressure slider picks nT and the code finds every root of n_eq(T) T = nT. Push the pressure outside the two-phase window -- computed live, about P/k = 1760 to 5510 K cm^-3 for this cooling curve -- and one branch genuinely disappears, from the box and from the readout. McKee and Ostriker (*ApJ* **218**, 148, 1977) added the third phase and the engine that drives it. Supernovae blow hot, low-density cavities that overlap and percolate, and the hot filling factor is set by how large a remnant gets before it merges back: Sedov-Taylor while adiabatic, R = 1.15 (E t^2 / rho)^(1/5); a radiative snowplough after t_rad, R ~ t^(2/7); merged when the shock speed falls to the ambient sound speed. That threshold depends on the pressure, so a higher-pressure medium stops remnants sooner and smaller -- the feedback loop that sets the hot filling factor in the first place. The sim measures the result rather than quoting it: 3600 random points thrown into the cube each frame, counted against the cavities. **No single observation sees the whole medium.** The 21 cm hyperfine line ([[Hydrogen_line]]) traces *both* neutral phases -- the CNM in absorption against background continuum, the WNM in emission -- and nothing else. It is blind to ionised hydrogen, which has no bound electron to flip, and blind to H2. Molecular hydrogen is homonuclear: no permanent electric dipole moment, so no dipole rotational spectrum. Its lowest allowed transition is the electric-quadrupole 0-0 S(0) line at 28.221 um, upper level 510 K above ground, and in a 15 K cloud that level is empty. The bulk of molecular gas in the universe is invisible. CO stands in for it: roughly 10^-4 as abundant, with a small permanent dipole and a J = 1 -> 0 line whose upper level is only 5.5 K up, so it is excited even in cold gas. The conversion is **X_CO = 2 x 10^20 cm^-2 (K km/s)^-1, carrying about 30 per cent uncertainty in the Milky Way disc** and much worse at low metallicity and in galaxy centres (Bolatto, Wolfire and Leroy, *ARA&A* **51**, 207, 2013). It is a calibration, not a constant of nature, and a sixty-page review exists because of that. The sim's observability control dims what a survey misses; it computes no conversion. **How certain are the numbers.** Not very. Temperatures and densities are good to a factor of a few. Volume filling factors are much worse: they depend on the model, the sight line, and how much of the disc thickness is averaged over, and reviews disagree by tens of per cent -- the hot phase is quoted anywhere from **20 to 70 per cent**. Worse, filling factors and mass fractions are measured over *different volumes* -- the molecular layer is about 75 pc thick, the warm layer nearer 1 kpc -- so f x n does not reproduce the mass fraction and cannot be expected to. Every figure in the table above is representative, not measured. ## Controls -> what each maps to | Control | Maps to | Range / values | Physical meaning | |---|---|---|---| | WEIGHT BY | the fraction each phase is drawn to occupy | VOLUME (filling factor) / MASS | THE control: a phase's drawn volume is its share of the cube, so the ranking inverts as a change of size | | highlight | which phase the readout describes | the five phases | T, n, nT and total P/k, or a note that the phase has no solution at this pressure | | show | per-phase visibility | five checkboxes | Hides a phase; the totals renormalise over the rest and still sum to 100 per cent | | see with | observability | all / 21 cm H I only / CO only | 21 cm cannot see H2 or ionised gas; CO sees only the molecular phase | | P/k | total thermal pressure | 800 to 10000 K cm^-3, default 3400 | Moves every phase along its own n = P/(k T), and switches the two-phase medium on and off | | box | box edge length | 250 to 550 pc, default 400 | Changes the volume averaged over -- what makes filling factors ambiguous | | phase diagram | -- | on / off | log n against log T, with the equilibrium locus and the isobar | | supernova | -- | button | Injects one: it carves a cavity, sweeps a shell and stalls | | running | -- | on / off | Pauses the clock, which runs at 0.4 Myr per wall-clock second | | reset | -- | button | Rebuilds cube, pressure and remnants from a fixed seed | The HUD's headline quantity is the pair of columns -- volume per cent against mass per cent, both summing to 100, the hot share measured rather than tabulated. ## Learning objective After playing, a learner can name the five phases with their rough temperatures and densities, explain why they coexist at one thermal pressure, state the volume/mass inversion and why f x n does not recover it, and say which phases 21 cm can and cannot see. ## Limits and connections The parcels are a sizing device, not hydrodynamics. Their positions never change; only their radii do, set so a phase's total drawn volume equals its share of the cube. There is no flow, no gravity, no magnetic field, no chemistry and no collapse; nothing evolves except the supernova remnants. What the picture is honest about is geometry: the hot phase percolates in three dimensions in a way no flat chart shows, and its filling factor is measured rather than asserted. The uncertainty quoted on that measured fraction is the binomial standard error over 3600 points -- sampling error, not the astrophysical one, which is far larger. Only two temperatures are solved. The CNM and WNM sit on roots of the cooling curve, so both n and T fall out of the solve and move with the pressure slider. Molecular gas is pinned at 15 K, the WIM at 8000 K and the HIM at 10^6 K, because their thermostats are different physics: CO and dust cooling against cosmic rays, a fixed energy per photoionisation, and shock heating with cooling far too slow to have reached equilibrium. Pressure balance uses *total particle* pressure, with a particle count per hydrogen nucleus for each phase -- helium adds about 0.1, free electrons roughly double it in ionised gas -- which is the only reason the ionised phases sit off the neutral n_H T line. The remnant fits for t_rad and R_rad differ between authors by tens of per cent; these are the commonly quoted ones. The Stromgren radius is a readout, not a constraint on the geometry. This is where hydrogen stops being a laboratory subject. The line that maps two of these phases is [[Hydrogen_line]]; the phase that holds most of the mass and emits nothing is the H2 whose invisibility forces the CO calibration above. ## Wikipedia : Wikitube **Strict pair:** [Wikipedia](https://en.wikipedia.org/wiki/Interstellar_medium) : [Wikitube](https://en.wikitube.io/wiki/Interstellar_medium) ## Previous hub tags Tree parent: [[Hydrogen]]. Legacy hubs: `HYDROGEN`. --- *Created 2026-08-05 - append-only - authored to WIKI_REPOPULATION_PROTOCOL v1.0 section 5 - portal-microsim-pass (PORTAL_Hydrogen batch 2) - 0 deletions*