ALBERT

Description: We recently proposed that a magnetohydrodynamic (MHD) turbulent cascade produces the bulk energization of electrons to approximately 25 keV in the impulsive phase of solar flares (LaRosa & Moore 1993). In that scenario, (1) the cascading MHD turbulence is fed by shear-unstable Alfvenic outflows from sites of strongly driven reconnection in the low corona, and (2) the electrons are energized by absorbing the energy that flows down through the cascade. We did not specify the physical mechanism by which the cascading energy is ultimately transferred to the electrons. Here we propose that Fermi acceleration is this mechanism, the process by which the electrons are energized and by which the cascading MHD turbulence is dissipated. We point out that in the expected cascade MHD fluctuations of scale 1 km can Fermi-accelerate electrons from 0.1 keV to approximately 25 keV on the subsecond timescales observed in impulsive flares, provided there is sufficient trapping and scattering of electrons in the MHD turbulence. We show that these same fluctuations provide the required trapping; they confine the electrons within the turbulent region until the turbulence eis dissipated. This results in the energization of all of the lectrons in each large-scale (5 x 10(exp 7)cm) turbulent eddy to 25 keV. The Fermi process also requires efficient scattering so that the pitch-angle distribution of the accelerating electrons remains isotropic. We propose that the electrons undergo resonant scattering by high-frequency plasma R-waves that, as suggested by others (Hamilton & Petrosian 1992), are generated by the reconnection. Ions are not scattered by R-waves. Provided that there is negligible generation of ion-scattering plasma turbulence (e.g., L-waves) by the reconnection or the MHD turbulence, the ions will not Fermi-accelerate and the cascading energy is transferred only to the electrons. We conclude that, given this situation, electron Fermi acceleration can plausibly account for the electron bulk energization in impulsive solar flares.

Description: We present a formaldehyde map of the translucent high-latitude molecular cloud MBM 16. The molecular gas traced by the H2CO is located in spatially distinct large structures that exhibit velocity coherence on a scale of 0.5 pc. These structures are not pressure-confined and are probably not self-gravitating. They may be transient structures. If so, we suggest that they are produced by shear flows whose scale length is of order the size of the cloud.

Description: A time-dependent analysis of emerging flux is carried out, and the time evolution of both the current sheet energetics and the plasma state is calculated. This evolution is determined in two different regimes. In the first case the width of the current sheet is assumed to be independent of the sheet thermodynamics and is fixed by the initial conditions. In the second, the width of the current sheet is a function of the resistivity and is allowed to decrease to its minimum given by the electron gyroradius. In both cases the resistivity is computed according to the marginal stability hypothesis. In each case the thermodynamic evolution is found to be quite rapid, with the temperature increasing from 10,000 to 1,000,000 K in a second or less. In contrast to previous studies, it is found that the resistivity is not significantly enhanced by the current-driven plasma wave turbulence. It is concluded that a laminar current sheet cannot be responsible for the activity associated with emerging flux.

Description: It is shown here that the generation of return current in large-scale astrophysical plasmas such as solar flares is quite different from the laboratory environment. Whereas in the laboratory the return current is established inductively and subsequently decays on a resistive timescale, the return current in a solar flare is established electrostatically and therefore does not decay. An explanation of the difference is given in terms of the characteristics of the laboratory circuit.

Description: We propose that the large production rate (approximately 10(exp 36)/s) of energetic electrons (greater than or approximately equal to 25 keV) required to account for the impulsive-phase hard X-ray burst in large flares is achieved through MHD turbulent cascade of the bulk kinetic energy of the outflows from many separate reconnection events. Focusing on large two- ribbon eruptive flares as representative of most large flares, we envision the reconnection events to be the driven reconnection of oppositely directed elementary flux tubes pressing into the flare-length current-sheet interface that forms in the wake of the eruption of the sheared core of the preflare bipolar field configuration. We point out that, because the outflows from these driven reconnection events have speeds of order the Alfven speed and because the magnetic field reduces the shear viscosity of the plasma, it is reasonable that the outflows are unstable and turbulent, so that the kinetic energy of an outflow is rapidly dissipated through turbulent cascade. If the largest eddies in the turbulence have diameters of order the expected widths of the outflows (10(exp 7)-10(exp 8)cm), then the cascade dissipation of each of these eddies could produce approximately 10(exp 26) erg burst of energized electrons (approximately 3 x (10(exp 33) 25 keV electrons) in approximately 0.3 s, which agrees well with hard X-ray and radio sub-bursts commonly observed during the impulsive phase. Of order 10(exp 2) simultaneous reconnection events with turbulent outflow would produce the observed rate of impulsive-phase plasma energization in the most powerful flares (approximately 10(exp 36) 25 keV electrons/ s); this number of reconnection sites can easily fit within the estimated 3 x 10(exp 9) cm span of the overall current-sheet dissipation region formed in these large flares. We therefore conclude that MHD turbulent cascade is a promising mechanism for the plasma energization observed in the impulsive phase of solar flares.

Description: LaRosa and Moore (1993) recently proposed that the bulk dissipation of magnetic field that is required for the electron energization in the explosive phase of solar flares occurs in a 'fat current sheet', a wall of cascading magnetohydrodynamic (MHD) turbulence sustained by highly disordered driven reconnection of opposing magnetic fields impacting at a turbulent boundary layer. Here, we use the well-observed great two-ribbon eruptive flare of 1984 April 24/25 to assess the feasibility of both (1) the standard model for the overall three-dimensional form and action of the magnetic field and (2) the turbulent reconnection wall within it. We find (1) that the morphology of this flare closely matched that of the standard model; (2) the preflare sheared core field had enough nonpotential magnetic energy to power the flare; (3) the model turbulent wall required to achieve the flare's peak dissipative power easily fit within the overall span of the flaring magnetic field; (4) this wall was thick enough to have turbulent eddies large enough (diameters approximately 10(exp 8 cm) to produce the approximately ergs energy release fragments typically observed in the explosive phase of flares; (5) the aspect ratio (thickness/vertical extent) of the turbulent reconnection wall was in the 0.1-1 range expected by (Parker 1973). We therefore conclude that the viability of our version of the standard model (i.e., having the magnetic field dissipation occur in our turbulent reconnection wall) is well confirmed by this typical great two-ribbon eruptive flare.

Description: This paper shows quantitatively that when a limiting X-ray yield is considered, a universal area of 10 to the 17th sq cm used in the thick-target electron bombardment model of McClymont and Canfield (1986) cannot explain the observed hard X-ray flux in large solar flare events without recourse to extreme values of the physical parameters of the flaring corona. The return current ohmic heating produced by a beam of flux 10 to the 13th ergs/sq cm/s results in a coronal temperature in excess of 100 million K. In this case, the thermal hard X-ray emission dominates the nonthermal emission and the EUV-to-hard X-ray ratio would not decrease with increasing hard X-ray flux, as observed. Hence, any model that requires a beam flux of 10 to the 13th ergs/sq cm/s is untenable. It is proposed that these apparently contradictory results can be reconciled if the X-ray emitting area is substantially larger than the area of the chromospheric precipitation site.