Exploration of the solar corona by high resolution solar decametric observations

In this review, current state of knowledge of high resolution observations at decameter wavelengths of the quiet Sun, the slowly varying component (SVC), type I to V bursts and noise storms is summarized. These observations have been interpreted to yield important physical parameters of the solar corona and the dynamical processes around 2R from the photosphere where transition from closed to open field lines takes places and the solar wind builds up. The decametric noise bursts have been classified into (i) BF type bursts which show variation of intensity with frequency and time and (ii) decametric type III bursts. The angular sizes of noise storm sources taking into account refraction and scattering effects are discussed. An attempt has been made to give phenomenology of all the known varieties of decametric bursts in this review. Available polarization information of decametric continuum and bursts has been summarized. Recent simultaneous satellite and ground-based observations of decametric solar bursts show that their intensities are deeply modulated by scintillations in the Earth's ionosphere. Salient features of various models and theories of the metric and decametric noise storms proposed so far are examined and a more satisfactory model is suggested which explains the BF type bursts as well as conventional noise storm bursts at decametric wavelengths invoking induced scattering process for 1 t conversion. Some suggestions for further solar decametric studies from the ground-based and satellite-borne experiments have been made.     

Non-BBN Constraints on the Key Cosmological Parameters

Since the baryon-to-photon ratio 10 is in some doubt at present, we ignore the constraints on 10 from big bang nucleosynthesis (BBN) and fit the three key cosmological parameters (h, M, 10) to four other observational constraints: Hubble parameter (ho), age of the universe (to), cluster gas (baryon) fraction (fo fGh3/2), and effective shape parameter (o). We consider open and flat CDM models and flat CDM models, testing goodness of fit and drawing confidence regions by the 2 method. CDM models with M = 1 (SCDM models) are accepted only because we allow a large error on ho, permitting h < 0.5. Open CDM models are accepted only for _M 0.4. CDM models give similar results. In all of these models, large 10 ( 6) is favored strongly over small 10 ( 2), supporting reports of low deuterium abundances on some QSO lines of sight, and suggesting that observational determinations of primordial 4He may be contaminated by systematic errors. Only if we drop the crucial o constraint are much lower values of M and 10 permitted.     

Gamma-ray line production in astronomical objects

This review focuses on the conditions for -ray line production in the most interesting astronomical objects, in light of the planned experiments: Gamma-1, GRO, Sigma, GRASP, and others. Among these objects are the Sun, the galactic center region, molecular and dust clouds, accreting and exploding stars.     

Probing the solar wind acceleration region using spectroscopic techniques

Measurements of the intensities and profiles of UV and EUV spectral lines can provide a powerful tool for probing the physical conditions in the solar corona out to 8 R and beyond. We discuss here how measurements of spectral line radiation in conjunction with measurements of the white light K-corona can provide information on electron, proton and ion temperatures and velocity distribution functions; densities; chemical abundances and mass flow velocities. Because of the fundamental importance of such information, we provide a comprehensive review of the formation of coronal resonance line radiation, with particular emphasis on the H i L line, and discuss observational considerations such as requirements for rejection of stray light and effects of emission from the geocorona and interplanetary dust. Finally, we summarize some results of coronal H i L and white light observations acquired on sounding rocket flights.Paper presented at the IX-th Lindau Workshop The Source Region of the Solar Wind.     

Solar microwave bursts — A review

We review the observational and theoretical results on the physics of microwave bursts that occur in the solar atmosphere. We particularly emphasize the advances made in burst physics over the last few years with the great improvement in spatial and time resolution especially with instruments like the NRAO three element interferometer, Westerbork Synthesis Radio Telescope and more recently the Very Large Array (VLA).We review the observations on pre-flare build-up of an active region at centimeter wavelengths. In particular we discuss the observations that in addition to the active region undergoing brightness and polarization changes on time scales of the order of an hour before a flare, there can be a change of the sense of polarization of a component of the relevant active region situated at the same location as the flare, implying the emergence of a flux of reverse polarity at coronal levels. The intensity distribution of cm- bursts is similar to that of soft X-ray and hard X-ray bursts. Indeed, it appears that the flaring behavior of the Sun at cm wavelengths is similar to that of some other cosmic transients such as flare stars and X-ray bursters.We discuss three distinct phases in the evolution of cm bursts, namely, impulsive phase, post-burst phase, and gradual rise and fall. The radiation mechanism for the impulsive phase of the microwave burst is gyrosynchrotron emission from mildly relativistic electrons that are accelerated near the energy release site and spiral in the strong magnetic field in the low corona. The details of the velocity distribution function of the energetic electrons and its time evolution are not known. We review the spectral characteristics for two kinds of velocity distribution, e.g., Maxwellian and Maxwellian with a power law tail for the energetic electrons. In the post-burst phase the energetic electrons are gradually thermalized. The thermal plasma released in the energy release region as well as the expanded parts of the overheated upper chromosphere may alter the emission mechanism. Thus, in the post-burst phase, depending on the average density and temperature of the thermal plasma, the emission mechanism may change from gyrosynchrotron to collisional bremsstrahlung from a thermal plasma. The gradual rise and fall (GFR) burst represents the heating of a flare plasma to temperatures of the order of 10⁶ K, in association with a flare or an X-ray transient following a filament disruption.We discuss the flux density spectra of centimeter bursts. The great majority of the bursts have a single spectral maximum, commonly around 6 cm- The U-shaped signature sometimes found in cm-dcm burst spectrum of large bursts is believed to a be a reflection of only the fact that there are two different sources of burst radiation, one for cm- and the other for dcm-, with different electron energy distributions and different magnetic fields.Observations of fine structures with temporal resolutionof 10–100 ms in the intensity profiles of cm- bursts are described. The existence of such fine time structures imply brightness temperatures in burst sources of order 10¹⁵ K; their interpretation in terms of gyrosynchrotron measuring or the coherent interaction of upper hybrid waves excited by percipitating electron beams in a flaring loop is discussed.High spatial resolution observations (a few seconds of arc to 1 arc) are discussed, with special reference to the one- and two-dimensional maps of cm burst sources. The dominance of one sense of circular polarization in some weak 6 cm bursts and its interpretation in terms of energetic electrons confined in an asymmetric magnetic loop is discussed. Two-dimensional snapshot maps obtained with the VLA show that multi-peak impulsive 6 cm burst phase radiation originates from several arcades of loops and that the burst source often occupies a substantial portion of the flaring loop, and is not confined strictly to the top of the loop. This phenomenon is interpreted in terms of the trapping of energetic electrons due to anomalous doppler resonance instability and the characteristic scale length of the magnetic field variation along the loop. The VLA observations also indicate that the onset of the impulsive phase of a 6 cm burst can be associated with the appearance of a new system of loops. The presence of two loop systems with opposite polarities or a quadrupole field configuration is reminiscent of flare models in which a current sheet develops in the interface between two closed loops.We provide an extensive review of the emission and absorption processes in thermal and non-thermal velocity distributions. Unlike the thermal plasma where absorption and emission are inter-related through Kirchoff's law, the radiation emitted from a small population of non-thermal electrons can be reabsorbed from the same electrons (self-absorption) or from the background (thermal) electrons through gyro-resonance absorption, and free-free absorption. We also suggest that the non-thermal electrons can be unstable and these instabilities can be the source of very high brightness temperature, fine structure ( 10 ms) pulsations.Finally in the last part of this review we present several microwave burst models-the magnetic trap model, the two-component model, thermal model and the flaring loop model and give a critical discussion of the strength and weakness of these models.     

Tearing instability of reconnecting current sheets in space plasmas

Magnetic reconnection provides an efficient conversion of the so-called free magnetic energy to kinetic and thermal energies of cosmic plasmas, hard electromagnetic radiation, and accelerated particles. This phenomenon was found in laboratory and space, but it is especially well studied in the solar atmosphere where it manifests itself as flares and flare-like events. We review the works devoted to the tearing instability — the inalienable part of the reconnection process — in current sheets which have, inside of them, a transverse (perpendicular to the sheet plain) component of the magnetic field and a longitudinal (parallel to the electric current) component of the field. Such non-neutral current sheets are well known as the energy sources for flare-like processes in the solar corona. In particular, quasi-steady high-temperature turbulent current sheets are the energy sources during the main or hot phase of solar flares. These sheets are stabilized with respect to the collisionless tearing instability by a small transverse component of magnetic fiel, normally existing in the reconnecting and reconnected magnetic fluxes. The collision tearing mode plays, however, an important and perhaps dominant role for non-neutral current sheets in solar flares. In the MHD approximation, the theory shows that the tearing instability can be completely stabilized by the transverse fieldB _n if its value satisfies the conditionB _n /BS ^–3/4 B is the reconnecting component of the magnetic field just near the current sheet,S is the magnetic Reynolds number for the sheet. In this case, stable current sheets become sources of temporal spatial oscillations and usual MHD waves. The application of the theory to the solar atmosphere shows that the effect of the transverse field explains high stability of high-temperature turbulent current sheets in the solar corona. The stable current sheets can be sources of radiation in the radio band. If the sheet is destabilized (atB _n /BS ^–3/4) the compressibility of plasma leads to the arizing of the tearing instability in a long wave region, in which for an incompressible plasma the instability is absent. When a longitudinal magnetic field exists in the current sheet, the compressibility-induces instability can be dumped by the longitudinal field. These effects are significant in destabilization of reconnecting current sheets in solar flares: in particular, the instability with respect to disturbances comparable with the width of the sheet is determined by the effect of compressibility.     

Concerning the interpretation of the ultraviolet radiation from B stars

Measurements of the shape of the ultraviolet spectrum from B stars are compared with the theoretical spectra predicted from a homogeneous series of eight model atmospheres which are known to be close to a state of radiative equilibrium and to give a good representation of the ordinarily observed spectral region. The broad-band photometer measurements of Byram, Chubb, and Friedman in the region 1314 indicate that the stars become brighter in the ultraviolet as their temperature increases. The theoretical spectra reproduce this trend. However, the theoretical spectra are about three times as bright at 1314 relative to their brightness at 5560 as is observed.The spectral observations at 50Å resolution of Stecher and Milligan of six absorption-line stars are compared in detail with theoretical spectra. The observed shape of the spectrum is reproduced well by the models from 2600 to longer wavelengths. At wavelengths shorter than 2600 Å, the observed fluxes from B stars are less than the predicted fluxes. At 2000 the deficiency is between a factor two and a factor four. The spectrum of Canis Majoris is observed to have a different shape from that found for four other early-type stars. In the case of Canis Majoris the deficiency at 2000 is about a factor 13.The proper manner in which to compare theory and observation is discussed and some astrophysical terminology is explained. Theoretical fluxes, , are given in Table 1 for eight early B type model atmospheres at wavelengths between the Lyman limit and 6251. These fluxes have been computed without consideration of the opacity due to line blanketing. It is shown that line blanketing can probably account for the differences noted between predicted and observed ultra-violet spectra of B stars. It is not necessary at present to invoke unusual sources of opacity in the stellar atmosphere or in the space between the star and the earth in order to explain the observations. Spectra of B stars in the 2000 region at sufficient resolution to show the line spectrum would clarify the problem.     

The emission mechanisms for solar radio bursts

Emission mechanisms for meter- solar radio bursts are reviewed with emphasis on fundamental plasma emission.The standard version of fundamental plasma emission is due to scattering of Langmuir waves into transverse waves by thermal ions. It may be treated semi-quantitatively by analogy with Thomson scattering provided induced scattering is unimportant. A physical interpretation of induced scattering is given and used to derive the transfer equation in a semi-quantitative way. Solutions of the transfer equation are presented and it is emphasized that standard fundamental emission with brightness temperatures 10⁹ K can be explained only under seemingly exceptional circumstances.Two alternative fundamental emission mechanisms are discussed: coalescence of Langmuir waves with low-frequency waves and direct conversion due to a density inhomogeneity. It is pointed out for the first time that the coalescence process (actually a related decay process) can lead to amplified transverse waves. The coalescence process saturates when the effective temperature T ^t of the transverse waves reaches the effective temperature T ^l of the Langmuir waves. This saturation occurs provided the energy density in the low-frequency waves exceeds a specific value which is about 10^-9 of the thermal energy density for emission from the corona at 100 MHz. It is suggested that direct emission has been dismissed as a possible alternative without adequate justification.Second harmonic plasma emission is discussed and compared with fundamental plasma emission. It also saturates at T ^t T ^l , and this saturation should occur in the corona roughly for T ^l 10¹⁵ K. If fundamental plasma emission is attributed to coalescence with low-frequency waves, then for T ^l 10¹⁵ K the brightness temperatures at the two harmonics should be equal and equal to T ^l . This offers a natural explanation for the approximate equality of the two brightness temperature often found in type II and type III bursts.Analytic treatments of gyro-synchrotron emission are reviewed. The application of the mechanism to moving type IV bursts is discussed in view of bursts with 10¹⁰ K at 43 MHz.     

An optical outburst from the periodic recurrent X-ray transient A0538-66

We present optical spectroscopy and photometry and IUE spectroscopy of the counterpart of the LMC recurrent X-ray transient A0538-66 during an outburst at the end of December 1980 which was consistent with the 16.6 day X-ray period (Skinner, 1980). The optical spectra show steadily increasing Balmer and HeI emission (indicative of a shell phase) superposed on a B2 IV spectrum with a substantial brightness increase of 2^m and the sharp turn-on of HeII 686 at the peak. Significant radial velocity changes have been detected but they show no correlation with the 16.6 day period. IUE spectra during a subsequent outburst show very strong and broad (5000 km s^–1) emission from C IV 1550 and HeII 1640. This behaviour is compared with other galactic transients and shell/Be stars.     

Collective radio-emission from plasmas

Collective radiation processes operating in laboratory and space plasmas are reviewed with an emphasis towards astrophysical applications. Particular stress is placed on the physics involved in the various processes rather than in the detailed derivation of the formulas. Radiation processes from stable non-thermal, weakly turbulent and strongly turbulent magnetized and unmagnetized plasmas are discussed. The general theoretical ideas involved in amplification processes such as stimulated scattering are presented along with their application to free electron and plasma lasers. Direct radio-emission of electromagnetic waves by linear instabilities driven by beams or velocity anisotropies are shown to be of relevance in space applications. Finally, as an example of the computational state of the art pertaining to plasma radiation, a study of the type III solar radio bursts is presented.

Frequently used Symbols

Latin Symbols teB ₀ ambient magnetic field - B ₁ perturbed magnetic field - c speed of light - E ₁ perturbed electric field - H Heaviside function - I unit dyadic - k wavevector of radiation fields - K _D inverse Debye length - m, M electron and ion mass - T _e , T _i electron and ion temperature - u relativistic velocity - V _e , V _i electron and ion thermal speeds - V _P , V _g wave phase and group velocities - W wave spectral energy density Greek Symbols relativistic factor - plasma dielectric tensor - _L , _T longitudinal and transverse components of in isotropic media (i.e., =kk _L /k ²+(l–kk/k ²) _T ) - index of refraction - angle between k and B ₀ - plasma dispersion tensor (i.e. =(c ²/ ²)(kk–k ² l)+) - determinant of - _D Debye length - _e electron cyclotron frequency - _u upper hybrid frequency - wave frequency - _e electron plasma frequency Proceedings of the NASA/JPL Workshop on the Physics of Planetary and Astrophysical Magnetospheres.National Research Council/Naval Research Laboratory Research Associate.     

Primordial Helium and ΔY/ΔZ from H II Regions and from Fine Structure in the Main Sequence Based on Hipparcos Parallaxes

The primordial helium abundance YP is important for cosmology and the ratio Y/Z of the changes relative to primordial abundances constrains models of stellar evolution. While the most accurate estimates of YP come from emission lines in extragalactic H II regions, they involve an extrapolation to zero metallicity which itself is closely tied up with the slope Y/Z. Recently certain systematic effects have come to light in this exercise which make it useful to have an independent estimate of Y/Z from fine structure in the main sequence of nearby stars. We derive such an estimate from Hipparcos data for stars with Z Z and find values between 2 and 3, which are consistent with stellar models, but still have a large uncertainty.     

Plasma waves at the dayside magnetopause

Experimental investigations of plasma waves at the magnetopause, including recent results from the AMPTE/IRM satellite, show that both E and B fluctuations typically have a featureless spectrum which monotonically decreases with frequency; integrated rms amplitudes are typically a few mV m^-1 for E and 10 nT for B, though in particular E can be as much as an order of magnitude larger in exceptional cases. Surveys show a lack of correlation between wave parameters and the magnetopause parameters. Under the assumption that crossing the diffusion region would give a pronounced signature in the waves, the survey data allow an upper limit to be placed on the latitudinal extent of the diffusion region, which is about 1000 km — implying that it is not surprising that the wave data surveys have so far failed to detect it. The observed wave turbulence levels have been used to estimate diffusion coefficients under different assumptions for the wave mode, but the resulting diffusion coefficient is always too small to explain either reconnection or boundary layer formation. Recent work of Galeev et al. (1986) indicates that the dominant diffusion process may be magnetic field migration, which is a macroscopic process involving the interaction of tearing mode islands. Assuming this mode to be present at the observed level of B, a particle diffusion coefficient of nearly 10⁹ m² s^-1 is obtained. Another macroscopic diffusive process which could occur at the magnetopause is stochastic E × B scattering, which also implies a diffusion coefficient the order of 10⁹ m² s^-1 if the observed E spectrum is assumed to be a turbulent cascade consisting of convective cells.     

Determination of the characteristics of the gamma-ray telescope Gamma-1

The Gamma-1 telescope has been developed through a collaboration of scientists in the USSR and France in order to conduct -ray astronomical observations within the energy range from 50 to 5000 MeV. The major characteristics of the telescope were established by Monte-Carlo simulations and calibrations made with the aid of electron and tagged -ray beams produced by an accelerator, and these have been found to be as follows: the effective area for photons coming along the instrument's axis varies from about 50 cm² at E = 50 MeV to approximately 230 cm² at E 300 MeV; the angular resolution (half opening of the cone embracing 68% events) is equal to 2.7° at E = 100 MeV, and 1.8° at E = 300 MeV; the energy resolution (FWHM) varies from 70% to 35% as the energy of the detected photons increases from 100 to 550 MeV; the telescope's field-of-view at the half-sensitivity level is 300–450 square degrees depending upon the spectrum of the detected radiation, and the event selection logic. Proceeding from the thus obtained characteristics it is demonstrated that a point source producing a photon flux J (E 100 MeV) = 3 × 10^-7 cm^-2 s^-1, can be detected with a 5 significance by observing it during 10⁶ s at the level of the Cygnus background, and a source having intensity J (E 100 MeV) = 10^-6 cm^-2 s^-1 can be detected to within a mean square positional accuracy of about 15.     

The solar flare plasma: Observation and interpretation

In the past several years, X-ray observations of the Sun made from rockets and satellites have demonstrated the existence of high temperature (20 × 10⁶ – 100 × 10⁶ K), low density plasmas associated with solar flare phenomena. In the hard X-ray range ( < 1 ), spectra of the flaring plasma have been obtained using proportional and scintillation counter detectors. It is possible from these data to determine the evolution of the hard X-ray flare spectrum as the burst progresses; and by assuming either a non-thermal or thermal (Maxwellian) electron distribution function, characteristic plasma parameters such as emission measure and temperature (for a thermal interpretation) can be determined. Thermal interpretations of hard X-ray data require temperatures of 100 × 10⁶ K.In contrast, the soft X-ray flare spectrum (1 <<30 ) exhibits line emission from hydrogen-like and helium-like ions, e.g. Ne, Mg, Al, Si,... Fe, that indicates electron energies more characteristic of temperatures of 20 × 10⁶ K. Furthermore, line intensity ratios obtained during the course of an event show that the flare plasma can only be described satisfactorily by assuming a source composed of several different temperature regions; and that the emission measures and temperatures of these regions appear to change as the flare evolves. Temperatures are determined from line ratios of hydrogen-like to helium-like ions for a number of different elements, e.g., S, Si, and Mg, and from the slope of the X-ray continuum which is assumed to be due to free-free and free-bound emission. There is no obvious indication in soft X-ray flare spectra of non-thermal processes, although accurate continuum measurements are difficult with the data obtained to date because of higher order diffraction effects due to the use of crystal spectrometers.Soft X-ray flare spectra also show satellite lines of the hydrogen-like and helium-like ions, notably the 1s ²2s ² S-1s2s2p ² P transition of the lithium-like ion, and support the contention that in low density plasmas these lines are formed by dielectronic recombination to the helium-like ion. Also, series of allowed transitions of hydrogen-like and helium-like ions are strong, e.g., the Lyman series of S up to Lyman-, and ratios of the higher member lines to the Lyman- line can be compared with theoretical calculations of the relative line strengths obtained by assuming various processes of line formation.This review will discuss the X-ray spectrum of solar flares from 250 keV to 0.4 keV, but will be primarily concerned with the soft X-ray spectrum and the interpretation of emission lines and continuum features that lie in this spectral range.     

Tests for coalescing binary neutron stars as cosmological origin of gamma-ray bursts

Theoretical logN-logS distributions and (V/V _max) tests of gamma-ray bursts in the model of coalescence of neutron star (NS+NS) and/or NS+black hole (NS+BH) binaries are calculated for a flat Universe (=1) with different values of the cosmological constant and under various assumptions about the star formation history. The observed logN-logS distribution and value of (V/V _max)=0.33 for 411 bursts with knownC _max/C _lim from the 2d BATSE catalogue are best fitted with a model for which = 0.2 and primary star formation occurs at redshiftsz5–6.     

Gamma Rays From Molecular Clouds

High energy -rays from individual giant molecular clouds contain unique information about the hidden sites of acceleration of galactic cosmic rays, and provide a feasible method for study of propagation of cosmic rays in the galactic disk on scales 100 pc. I discuss the spectral features of ⁰-decay -radiation from clouds/targets located in proximity of relatively young proton accelerators, and speculate that such `accelerator+target systems in our Galaxy can be responsible for a subset of unidentified EGRET sources. Also, I argue that the recent observations of high energy -rays from the Orion complex contain evidence that the level of the `sea of galactic cosmic rays may differ significantly from the flux and the spectrum of local (directly detected) particles.     

Observed Densities in the Universe

Baryons observed in Ly absorbers contribute to the density parameter 0 by bar 0.06 in close agreement with the value of 0.06 from primordial nucleosynthesis (H0=55 km s-1 Mpc-1, = 0 assumed throughout). A number of methods are known to measure 0 from density fluctuations; bound structures tend to yield lower values (m 0.2-0.4), field galaxies over large scales higher, but still undercritical values (m 0.6 ± 0.2). The best compromise value is 0 0.5, but the present methods are blind to diffusely distributed, exotic matter which still could make 0 = 1. A satisfactory solution of 0 (and ) will only come from a fundamental cosmological test (e.g. the Hubble diagram of [evolution-corrected] supernovae type Ia) in combination with the CMB fluctuation spectrum.     

Physical mechanisms for turbulent dissipation in collisionless shock waves

Temporal and spectral characteristics of solar hard X-ray bursts are briefly reviewed. The merits of non-thermal and thermal flare models are discussed. The validity of these models may be checked by future measurements of X-ray polarization. Finally, some important results of recent satellite experiments are described providing information on the spatial distribution of hard X-ray sources: the multi-spacecraft observation of X-ray bursts and the imaging of X-ray sources by means of the HXIS instrument.Paper presented at the IX-th Lindau Workshop The Source Region of the Solar Wind.     

Magnetic fields in the ionosphere of Venus

This review surveys the observations of the ionospheric magnetic fields of Venus as observed on the Pioneer Venus Orbiter and the models that have been developed to describe them over the last decade. The models for the large-scale ionospheric field have developed to the advanced stage of one-dimensional, self-consistent, multi-fluid MHD models which provide a detailed picture of the field in the subsolar region for specific upper boundary conditions. In contrast, the models for the small-scale fields and the nightside fields have only reached a rudimentary stage. Much challenging work remains to be done on the origin of the ionospheric flux ropes and nightside ionospheric hole fields. On the whole, the subject of the ionospheric fields would greatly benefit from 3-dimensional global MHD models with self-consistent treatments of the ionosphere.     

The magnetogram inversion technique: Applications to the problem of magnetospheric substorms

The magnetogram inversion technique (MIT) is based upon recordings of geomagnetic variations at the worldwide network of ground-based magnetometers. MIT ensures a calculation of a global spatial distribution of the electric field, currents and Joule heating in the ionosphere. Variant MIT-2 provides, additionally, continuous monitoring of the following parameters: Poynting vector flux from the solar wind into the magnetosphere (); power, both dissipated and accumulated in the magnetosphere; magnetic flux in the open tail; and the magnetotail length (l _T) (distance between the dayside and nightside neutral points in the Dungey model). Using MIT-2 and data of direct measurements in the solar wind, an analysis is made of a number of substorms, and a new scenario of substorms is suggested. The scenario includes the convection model, the model with a neutral line and the model of magnetosphere-ionosphere coupling (outside the current sheet), i.e., the three known models. A brief review is given of these and some other substorms models. A new element in the scenario is the strong positive feedback in the primary generator circuit, which ensures growth of the ratio = / _Aby an order of magnitude or more during the substorms. Here _Ais the Pointing vector flux in the Akasofu-Perrault approximation, i.e., without the feedback taken into account. The growth of during the substorm is caused only by the feedback effect. It is assumed that the feedback arises due to an elongation of the magnetotail, i.e., a growth of l _Tby a factor of (23) during the substorm.In the active phase of substorm, a part (the first active phase) has been identified, where the principal role in the energetics is played by the feedback mechanism and the external energy source (although the internal source plus reconnection inside the plasma sheet make a marked contribution). In the second active phase (expansion) the external generator (solar wind) is switched off, and the main role is now played by the internal energy source (the tail magnetic field and ionospheric wind energy).Models of DP-2 DP-1 transitions are also considered, as well as the magnetospheric substorm-solar flare analogy.