Cosmic Plasma


By Hannes Alfvén

Cosmic Plasma by Hannes Alfvén.

Publ. 1981. Astrophysics and Space Science Library, Vol. 82 (1981) Springer Verlag. ISBN 90-277-1151-8Buy


The aims of the present monograph may be summarized as follows:

  • (1) Attention will be given to the question of how much knowledge can be gained by a systematic comparison of different regions of plasma. We will consider plasmas with linear dimensions which vary from laboratory size up to the Hubble distance.
  • (2) The traditional magnetic field description of plasmas will be supplemented by an electric current description. It is demonstrated that many problems are easier to understand with a dualistic approach.
  • (3) A rather strict distinction will be made between plasma regions which are accessible to in situ diagnostics and those regions which are not. In the former, we can expect to understand the plasma phenomena reasonably well, and to be able to discriminate between different interpretations, whereas in the latter regions, conclusions must necessarily remain largely speculative.
  • (4) It is claimed in this monograph that in the speculative domain, those theories, which are based on what we know from laboratory experiments and from regions accessible to spacecraft measurements, deserve higher credibility than those which are based on theories which we know to be misleading when applied to regions that have been subject to high-quality diagnostics.
As it is impossible to treat an extensive and complicated field like cosmic plasma physics in a rigorous way in a monograph of the present size, we have followed the policy that – as Kadomtsev (1965) puts it – "when rigour appears to conflict with simplicity, simplicity is given preference".



Cosmology in the Plasma, Universe: An Introductory Exposition





Absfzzicf—Acceptance of the plasma universe model is now leading to drastically new views or the structure of the  universe.  The  basic aspects of cosmological importance are: a) The same basic laws of plasma physics hold everywhere; b) mapping of electric fields and cur- rents are necessary to understand cosmic plasma; c) space is filled with   a network of currents leading to the cellular and filamentary structure of  matter,  and d) double  layers, critical  velocity,  and  pinch effects tire  of decisive importance in how cosmic plasma evolves. This paper re- views a number of the outstanding questions of cosmology in the plasma universe.                               -






of  the  plasma  universe  model  is now


to drastically new views of the structure of the universe. Their main characteristics are given by Falthammar  [I 1]. Most important  are the following:

  • The same basic laws of plasma physics hold from laboratory and magnetospheric heliospheric plasmas  out to interstellar and  intergalactic
  • In order to understand the phenomena in a certain plasma region, it is necessary to map not only the mag- netic but also the electric field and the electric  
  • Space is filled with a network of currents which transfer energy and momentum over large or very large distances. The currents often pinch to filamentaiy or sur- face currents. The latter are likely to give space, as also interstellar and intergalactic space,  a  cellular  structure [1, ch.  10].
  • A number of plasma phenomena, like double layers, critical velocity, pinch effect, and the properties of elec- tric circuits, are of decisive importance. The phenomena mentioned have been known for decades (or even more than a century), but up to now they have almost system- atically been ignored in cosmic physics. If they are taken into account, not only interplanetary space but also inter- stellar and intergalactic space must have a cellular struc- ture [1, ch.  10].

In the present paper we shall analyze whether these drastic changes in our picture of the universe have cos- mological consequences.

When we discuss cosmic plasma we concentrate the at- tention  on the low-density  medium  which  fills space be-


Manuscript  received  April  19,  1989; revised  August  17, 1989.

The author is with the  Department  of  Plasma  Physics,  Royal  Institute of Technology,  S  100 44,  Stockhol in, Sweden.

IEEE  Log  Number 8933173.












strated by Birkeland in his terrella experiment. It can also be seen on space photographs from above. The figure shows a terrella experiment by Block (reference [7]). who further developed the terrella experiment technique introduced by Birkeland in the beginning of the century.




tween massive bodies like stars, etc. , that also consist of plasmas but of much higher   densities.



As plasma fills almost all the universe, it should be easy to observe. However, the Earth’s magnetic field prevents cosmic plasma from reaching most of the Earth’s surface  or upper atmosphere. Exceptions are the two auroral zones (cf. Fig. 1). As the Scandinavian third of the auroral zone (see Fig. 2) is the only one which has a tolerable climate, Scandinavia has for centuries been a privileged region for studies of the cosmic plasma. It was Anders Celsius who, 250 years ago, identified the aurora as an electromagnetic phenomenon by observing that a big compass needle placed on his desk in Uppsala changed its direction when the aurora appeared. By doing so he introduced a tradition of coordinating  laboratory  and space phenomena.

This tradition is still alive. One of its most prominent scholars was Birkeland, in Norway. The intimate collab- oration between  laboratory plasma experiments and  iono-




0093-3813/90/0200-0005$01.00       1990 IEEE


6                                                     IEEE  TRANSACTIONS    ON   PLASMA    SCIENCE.    VOL.    18.   NO.    1.   FEBRU A RY    1990





AN IIMAT TER     REGION          




Fig. 4. Leidenfrost layer. At the bou ndary between a region of matter and a region of antimatter there is proton-antiproton annihilation which pro- duces mesons that rapidly decay into 10’ eV electrons and positrons. These form a very hot and extremely thin boundary layer separating the matter and antimatter regions. The radiation from this layer is so smal I that it is difficult to detect (references [ 14] and [ 15]).




Fig. 2. The zone where cosmic plasma can reach the Earth consists polit- ically of three parts. Climatic conditions make it preferable to observe the aurora in the Scandinavian zone, which has given a preference to space research in that area, where auroral research has a long tradition.



Fig. 3. Astrophysics in a nutshell. Three revolutions have affected the de- velopment of plasma physics in space: 1) The Copernican revolution; 2) the plasma revolution; and 3) the Sputnik revolution.



spheric magnetospheric observations is now extended by the cooperative European space research in northern Scandinavia. The Swedish Viking mission  has continued in the Scandinavia tradition of studying the properties of the plasma which fills the universe from the small-scale phenomena in the laboratory to the large-scale phenomena in galactic and intergalactic space. Cosmological theories based only on mathematical calculations should be treated with  considerable  skepticism  (see Fig. 3).




As a consequence of Dirac’s theory, Klein [12], [13] suggested that the universe might be matter—antimatter symmetric. A quarter of a century ago the cosmological interest  was  focused  on  the  fight  between  the ‘Contin-

uous Creation” and what  was  later  called  the  ‘Big Bang.’ ’ To both of these cosmologies a matter—antimatter symmetric universe was a disturbing concept that was im- portant to get rid of. This was attempted by demonstrating that a homogeneous symmetric universe was out of the question, because it would be completely annihilated in a time of the order of millions  of  years.

As a starting point, Klein considered an essentially ho- mogeneous universe, but neither he nor anybody else claimed that the present universe should have any simi- larity to a homogeneous model. This did not help. It be- came a “generally accepted’ view that a matter—anti- matter symmetry was out of the question. A number of attempts to correct this conclusion and open a free and unbiased discussion have been in vain. All such attempts have been met with the answer that no one has demon- strated in an unquestionable way the cosmic existence of antimatter. This is correct. On the other hand, no coun- terproof exists to demonstrate that the universe is not symmetric (see [16)). Cosmical physics is now seeing a general drift towards inhomogeneous models, and it is re- alized that homogeneous models are often not useful, even as first approximations.

The plasma universe model introduces important new arguments  in this discussion.

From in situ space observations we know that there are current layers in space which separate space into regions with different magnetization, different temperatures and densities, and even different chemical compositions. Thus it has been found that space plasma ‘has a tendency to- wards a cellular structure. This tendency has been ob- served throughout the regions at present accessible to spacecraft. As it is impossible to claim that such a basic property of a plasma (its tendency to produce cellular structures) should be confined to the regions presently available to spacecraft, one must conclude that space in general has a cellular structure (see the general discussion in  [1, ch.  VI]).  The different  chemical  compositions  on




Fig. S. The plasma universe consists of a reliable diagnostic region in which laboratory experiments and in situ measurements make a sophisticated study  of a plasma possible.  Outside the reach of spacecraft (black line), investigations  must be based  on a sy mbiosis between observation and the know ledge of plasmas gained from the reliable diagnostic region.




both sides of the magnetopause may have a counterpart in interstellar and intergalactic space, where there may be a difference in the kind of matter: Ordinary matter (koino- matter) on one side, and antimatter on the other. This con- clusion should be combined with the theory of Leiden- frost layers as analyzed by Lchnert [14], [15, fig. 4]. It is important to note that such layers—if static—may emit a negligible  amount of radiation.  They should be   depicted



    ...        .. ..   ......      ...  ..       ..      ......-....           .. ......      . ..'   ....         ...    .--          ——'-'—.  .     


as being thin, very hot layers of almost complete vacuum.                                                                                          The result of such discussions  is that we have little  rea-

son to question that in the present state of development  of


our concept of interstellar and intergalactic space the plasma universe could very well be matter-antimatter symmetric  (see  [1, ch. VI]).



The result of extensive discussions  about  the structure of the plasma universe is depicted in Fig. S. We can dis- tinguish between two categories of regions: One is the “reliable diagnostic region,” comprising laboratory plas- mas and those regions of the magnetospheres and the helio- sphere that are accessible to spacecraft. The other part of the universe comprises all other regions, i.e. , those out- side of the outer planets and the Sun, where our knowl- edge of the properties of plasmas depends on extrapola- tion from results obtained in the reliable diagnostic regions (see Table I and Fig.   5).




How little we know about the universe outside the re- liable diagnostic regions can be demonstrated by asking whether one of our closest stars, say n Centauri,    consists



NoX-moor/-nys               Eéu:£ p%c1es’Noi’  ego:r

,grdahyinconoecdoo éJb pmdurtoo








of matter or antimatter. Of course, “of matter’ ’ is the ob- vious reply, but how can this be proven? An antimatter  star should emit the same spectrum. Certainly the sign of the rotational Faraday effect is different for electrons and positrons, but as we do not know the sign of the magnetic field, this does not  help.

It may be suggested that if the solar wind (from our  Sun) approaches  o Centauri and that this star also emits  a


was   "r RANSACTIONS   ON   PLASMA   SCIENCE,    VOL.    18,   NO.    1.   FEBRU A RY     l9»0


7 I M E

Fig. 6. Evolution of the metagalaxy in Klein’s model (see reference [1, fig. VI. 3]). The time scale before the turning should be enlarged.




solar wind, collisions between these two winds would give rise to violent annihilation with detectable gamma-ray emission. However, there is plenty of room for a large number of Leidenfrost layers between our Sun and n Cen- tauri.

The absence of large quantities of antiparticles in the Cosmic  Radiation  (CR)  is another  argument.  If  the sign

of primary  cosmic  rays were  known  for energies  > 1014

eV, this would be a strong argument.  However,  the sign  of primary CR is known only up to  less than  10'' eV. Such particles have Larmor radii, which are very small, and all the CR in this energy range could very well be generated  inside an extended heliosphere.

So once again we do not know enough to exclude a Dirac—Klein symmetry. On the other hand, there is not decisive observational argument in favor  of antimatter. (In [1, ch. VI] this problem is discussed in some detail.) Hence it must be considered legitimate to study the con- sequences of both hypotheses: Is the plasma universe symmetric, or does it consist of exclusively ordinary mat- ter? As the second alternative is treated in a gigantic lit- erature, it is appropriate that we here concentrate on the first alternative. If this is correct, an unprecedented change

in astrophysics  would occur.

At the American Geophysical Union 1988 Fall Meeting in San Francisco, attention was drawn to the enormous difference it would make to astrophysics  in general if the  n Centauri consists of matter or of antimatter. A symbolic award was proposed to be given to the scientist who could prove  which  of these alternatives  is correct.


Klein [12], [13] makes the natural assumption that the Dirac matter—antimatter symmetry is valid also in the uni- verse. (Klein used the old term “metagalaxy’ because he does not take for granted that the region of space we ob- serve at present is the whole universe.) He assumes  that our metagalaxy “initially’ was in the form of a gigantic homogeneous cloud of ambiplasma —koinomatter (from greek kionos, meaning ordinary)—and antimatter mixed homogeneously (Fig. 6). Its density is so extremely small that annihilation is negligible. This sphere contracts under the action of gravitation. When it has reached a size of perhaps 10fip ( fip = Hubble distances), annihilation be- comes important and produces a force opposite to the gravitation, which slows down the contraction. Annihi- lation increases with increasing density, and eventually it is large enough to convert the contraction into expansion. After the turning which may take place at, say, 0. 1 Rg , the sphere expands again. This expansion is identical with the Hubble expansion. Some 10 percent is annihilated at the turning and converted into kinetic energy of the   Hub-

ble  expansion  and  different  kinds  of radiation—among           

them cosmic microwave background  radiation.



When discussing how to approach the origin and evo- lution of the sDlar system, G. Arrhenius (private com- munication),  who  is  a  geologist,  pointed  out  that when




the geological history of the Earth is studied, the  actu- alistic approach is very valuable. This principle says that' the present is the key to the past. In  other  words,  we should not approach a historical problem in science by making a guess about how the conditions were in a certain region several billion years ago,  because  the probability that such a guess is correct is very close to zero. Instead,  we should start from the present  conditions.

During the ages innumerable prophetic guesses have been made. They have survived to our times only in cases when the guesses have been claimed to derive from divine inspiration. This means that the guesses must have been made by great religious prophets. Hence we find such guesses included as important parts of holy religious scriptures.

Hence, there are two different ways of approaching the prehistory of the present state of the plasma universe, or part of it.                                  ’

  1. The Prophetic  Approach

A guess is made about the state very long ago, and this is made credible by prophetic authority. This approach often assumes that there was a “creation” at  a  certain time, and it is often claimed that we know more about this event than about more recent  times.

  1. The Actualistic Approach

We start from the observed present-state and try to ex- trapolate backwards in time to even more ancient states. From this follows that the further backwards we go, the larger is the uncertainty about the state. This approach  does not necessarily lead to a “creation” at a certain time, nor does it exclude this possibility. In principle, it is also reconcilable with a universe which is ‘ungenerated and indestructible,     as Aristotle  expressed it.



The Hubble diagram is usually plotted on a logarithmic scale. Taking account of the great uncertainties, it is rec- oncilable with a picture of a universe in which the expan- sion derives from a “Big Bang’ at a singular point. How- ever, this does not mean that the Hubble diagram proves this.

First of all, the observed red shifts do not necessarily derive from a longitudinal Doppler effect. But even if we assume that they are caused by a Doppler effect, the re- construction of the orbits of the individual galaxies leads  to the diagram which merely shows that once they were much closer together. In fact, it seems legitimate to con- clude that the metagalaxy (the  “universe’  according  to the “Big Bang” hypothesis) once had dimensions of  about

  1. 1 ?tp, but it seems to be not legitimate to conclude that it was even smaller. For a discussion of this, see [1, ch. VI], and also the discussions  by Bonnevier  [8].


  1. A BIGGER  “BIG  BANG’ ’

Let us for a moment eliminate a number of   ‘prophetic”

ad hoc hypotheses;  i.e. : a) That there are some orders   of


magnitude of more mass in the universe than is really ob- served (“missing mass’ ’); b) that the Hubble expansion was caused by unknown effects at a singular point; c) that the present universe does not contain an appreciable amount of antimatter; and d) that cosmology can be treated by homogeneous models.

We try to construct a universe such that: i) It is essen- tially matter—antimatter symmetric; ii) the Hubble expan- sion is caused by well-known processes (among them, en- ergy release by annihilation) in a region of 10’ light years (a bigger “Big Bang”); iii) it does not contain large quan- tities of missing mass; and iv) it is highly inhomogeneous and has a cellular  structure.

What is said above does not lead to the conclusion that we accept the Klein cosmology  as it was  presented.



Klein [12], [13] bases his analysis on the  assumption that very long ago the metagalaxy consisted in an ex- tremely large sphere of matter and antimatter. This clas- sifies his theory as “prophetic.” However, the picture he gives of the evolution after the turning (the bigger .“Big Bang’ ’) can probably serve as a guideline to an evolution- ary actualistic approach. Whether this is correct or not can only be found if the observed present state of the universe is used as a basis for a reconstruction of increasingly old states. It is reasonable to use well-established laws of na- ture as a first approximation. The enormous mass of ob- servations should be subject to an unbiased application of modern plasma physics as derived‘ from extrapolation of studies of the reliable  diagnostic  regions,  etc.

The primary aim should be to try to reconstruct evolu- tionary history back to the ‘turning’ (Fig. 6). If this at- tempt runs into difficulties, the time is ripe for drawing conclusions  about missing  mass, etc.



Space research has resulted in a model of the ‘plasma universe” (see Fig. 5). In [9] and [10] it is shown how in situ observations in the Earth’s magnetosphere can pro- vide a key to the understanding of astrophysical and cos- mological phenomena  in the plasma  universe.

In the plasma universe not only the  present  state but also its prehistory is of importance. As discussed in Sec- tion VII above, there are two different types of ap- proaches, called ‘prophetic’  and  “actualistic.”  It  has  been proved that in cosmogony the actualist approach is preferable [2]—[4], and a number of prophetic approaches are now falling down. Because the galactic problems are similar to the ionospheric—magnetospheric problems, this approach is likely to be also preferable in this field. How- ever, this does not mean that we necessarily should accept it in the case of  cosmology.

We have given above a brief summary of the Klein cos- mology. The conventional ‘Big Bang” is too well-known to need a recapitulation. Let us first state the aspects where there is agreement  between  the two approaches.  Both at-


10                                                                                                        IEEE  TRANSACTIONS  ON  PLASMA  SCIENCE,  VOL.  18.  NO.  I .  FEBRUARY 1990



tribute the Hubble expansion to a ‘Big Bang,”  and both are prophetic theories. Further, as Fig. 6 shows, they both give a similar Hubble from between the present time and back to about 0. 1 Tp (Tg —— the Hubble time). However, this does not mean that the properties of the expanding plasma are the same. These properties are derived from  the general properties of the early-time plasma of the two prophetic theories.

As the cosmological problems are outside the “reliable diagnostic region” in Fig. 5,  it  is appropriate  to  derive the properties of the expanding plasma after 0. 1 Tp from the plasma universe model. Hence we should apply an actualistic approach to this part of the Hubble expansion and leave the discussion of what happened before 0. 1 Tg for later discussion. If this extrapolation seems reason- able, i.e. , if we succeed in a reconstruction  of the state  at

  1. 1 Tp, we could use this as a basis for a discussion of earlier periods, but this is outside the aim of this   paper.



What has been said above means that we should try to adopt the “Big Bang” cosmology to the plasma universe model in two different ways. Both may run into difficul- ties, and only a free and unbiased discussion  can clarify  on  how  to  deal  with these.

The hrst one is that we start from the traditional “Big Bang,” which necessarily leads to a number of difficul- ties. It is basically a ‘prophetic approach.” The second approach starts from the present state of the plasma uni- verse and attempts to reconstruct  earlier states.  Hence  it is essentially an ‘actualistic approach.” This certainly contains  a  number  of uncertain  and doubtful  points. An

essential point is that the Hubble expansion  was caused  by annihilation in a large region (1  9 light years). We  call

this a bigger  “Big Bang.’’  We leave the early-time  part  of the Klein cosmology—which is prophetic—outside the discussion.



This paper—as well as most of my papers—is due to an intimate collaboration  with  Prof. N.  Herlofson.



[ I]  H.  Alfvén,  Cosmic Plasma  (monograph).    Dordrecht, The Nether- lands:  Reidel,  1981.

[2] H. Alfvén, ‘‘The plasma universe,’ ’ Phys.  Today,  vol.  39,  pp.  22- 27,  1986.

[3]  H.  Alfvén,   ‘‘Model  of  the  plasma  universe,”   /€66  Trans. Plasma

Sci. , vol. PS-14, pp. 629—638,  1986.

[4]  H. Alfvén,  ‘ ‘Double layers and circuits in astrophysics,”  IEEE Trans.

Plasma Sci. , vol . PS- 14, pp. 779—793, 1986.

[5] H. Alfvén, ‘ ‘Plasma universe,’ ’ in Proc.  1985—8 7 MIT Ssmp.  Phj s.  of Space Plasmas, T. Chang, J . Belcher, J.R. Jasperse, and  G.B. Crew, Eds.    Cambridge,  MA: Scientific,  1987, pp. 5—  17.

[6] H. Alfvén, “Three revolutions in cosmical science: From the tele- scope to the Sputnik,’ ’ presented at the l4th Texas Symp. on Relative Astrophy s. , 1989, pp. I - 17.

[7] L.P. Block,  “Model experiment  on aurorae  and  magnetic  storms,”

Tellys, vol. 7, pp. 65—86,  1955.

[8] B. Bonnevier, “The expansion of the metagalaxy,” Astroph ys. Space Sci. , voI. 79, pp. 509—513. 198 I .

[9] C.-G. Falthammar, ‘ ‘Magnetosphere—ionosphere interactions—near- earth manifestations of the plasma u niverse,’ ’ IEEE Trans. Plasmu Sci. , vol. PS- 14, pp. 6 16—628, 1986.

[10) C.-G. Falthammar, “‘Astrophysical significance of observations on sxperiments in the earth’s magnetosphere,’ ’ ESA SP-285. vol. I , pp. 9- 19.

[I 1] C.-G. Falthammar, ‘ ‘Electromagnetics of cosmical plasmas—some basic aspects of cosmological importance,’ ’ f£££ Trans. Plasma Sci. , this issue, pp. 1 l— 17.

[ 12] O. Klein, “Instead of cosmology,’ ’ Nature, vol. 2 I I , pp. l337— 134 I, 1966.

[ 13) 0. Klein, ‘ “Arguments concerning relativity and cosmology," ’ Sci- ence, vol.  l7 1, pp.  339-345,  197 I .

[14] B. Lehneri, “Problems of matter—antimatter boundary layers,’ ’ A.s - trophys. Space Sci. , voi. 46, pp. $1 —7 I , 1977.

[ 15] B. Lehnert , ‘ ‘Matter-antimatter bou ndary layers and a magnetic neu- tral sheet,’ ’ Astrophys.  Space  Sci. ,  voI. 53, pp. 459-465,  1978.

( 16) W.B. Thompson, ‘ ‘The cosmological separation of matter and anti- matter,’ ’ Astroph ys.  Space Sci. , vo I. 55, pp. 15—23, 1978.

[ 17] A.C. Williams, ‘ ‘Double layers in astrophysics: Highlights of the 1986 MSFC symposium,’ ’ USA Trans. Plasma Sci. , vo I. PS- 14, pp. 838—84 I, 1986.




Hannes Olof Giista Alfvén (SM’55—F’60—LF’83) was born in Norrkorping, Sweden, on May 30, 1908. He receiv‹(d the Ph. D. degree from the Uni- versity of Uppsala  in  1934.

He served as a Docent, and was Professor of Theoretical Electrody nannies at the Royal Insti- tute of Te nology, Stockholm ( l940— 1945); Pro- “fessor of El ronics ( 1945— 1964); Professor of Plasma Physics    964— 1973); and has been a Pro-

,     ,       fessor at the University of California, San Diego,

since 1967.

He is usually regarded as the father of the modern discipline of classical physics known as hydromagnetism or magnetohydrody nannies. He has pub- lished over a hundred papers on plasma physics, magnetohy drody nannies, astrophy sics, and cosmology. He is the author of Cosmical Electrod›'- rnm‹cs (Oxford, 1950); Cosmical Electrodynumics (2nd ed.) (with C.-G. Falthammar; Oxford, 1963); Evolution of the Solar Sj'stem (with G. Arrhenius; NASA, 1976); Cosmic Plasma (Reidel, 198 I ); and the popular science books Worlds-Anti worlds (Freeman, 1966), The Tale of the Big Computer (Coward—McCann, 1968), Atom, Man, and the Universe (Free- man, 1969), and Living on the rhird Planet (with K. Alfvén; Freeman, 1972).

Professor Alfvén is a member of the Royal Swedish Academy of Sci- ences, the Akademia NAUK (USSR), the Yugoslavian Academy of Sci- ences, the American Academy of Arts and Sciences, the National Academy of Sciences, and the Royal Society . He was awarded the  Gold Medal of  the Royal Astronomical Society (1967), the Nobel Prize in Physics ( 1970), the Gold Medal of the Franklin Institute ( 197 1), the Lomonosov Medal of the USSR Academy of Sciences ( 197 I ), and the Bowie Medal of the Amer- ican Geophysical Union ( 1988).