JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 93, NO. A7, PAGES 7441-7465, JULY 1, 1988 DETERMINATION OF AURORAL ELECTROSTATIC POTENTIALS USING HIGH- AND LOW-ALTITUDE PARTICLE DISTRIBUTIONS P. H. Reiff, • H. L. Collin, z J. D. Craven,' J. L. Bureh, • J. D. Winningham, • E.G. Shelley, z L. A. Frank,' and M. A. Friedman 5 Abstract. The Dynamics Explorer (DE) pair of average energies of O + and He + are comparable to, spacecraft provide a unique opportunity to search but typically somewhat larger than, that of H +, for the presence of electric fields aligned par- indicating that a two-stream instability may be allel to magnetic field lines by sampling, nearly the heating mechanism. Electron heating is also simultaneously, the velocity-space distribution detected within the auroral acceleration region, functions of ions and electrons at two points on with gains in characteristic energies of 10-15• auroral field lines: DE 1 at high altitudes of e•. From the high-altitude electron measure- (9000-15,000 km in this study) and DE 2 at low ments, we can determine a minimum potential altitudes (400-800 kin). Three independent tech- distribution as a function of altitude in order niques are used to infer the auroral electro- to overcome the mirror force. We find that in static potential difference from the particle one case at least 94 V of the 2300-V potential distributions: (1) the energy of the precipita- drop must occur above 7000 km, and at least 960 V ting electrons at DE 2 (compared to that at DE above 2000 kin. 1), (2) the energy of the upgoing ions at DE 1, and (3) the widening of the loss cone for elec- Introduction trons at DE 1. The three estimates are in gene- ral agreement, confirming the long-standing, but The velocity distribution functions for pre- not fully accepted, hypothesis that parall el cipitating auroral electrons measured at low electrostatic fields of 1-10 kV potential drop at altitudes (400-1600 km) have long been known to 1-2 R E altitude are an important source for be consistent with Maxwellian thermal distribu- auroral particle acceleration. The upflowing ion ticns accelerated through a parallel potential distribution typically can be characterized by a difference of 1-10 kV [Gurnett, 1972; Reasoner sharp peak and a falloff at high energies of the and Chappell, 1973] (see the seminal theoretical form exp-{(E-Epeak)/Eo}, with Epeak being the paper of Knight [1973]). At energies belowthe peak energy and E o the characteristic energy. accelerating energy e•, the distribution function This is the functional dependence one expects is dominated by secondary and energy-degraded if a MaxwellJan of thermal energy E o is accel- primary electrons which are bounded by the absor- erated upward by a parallel electric field with bing and scattering atmosphere below and the e•zEpeak. The fact that the peak energies and accelerating potential above [Evans, 1974; not the flow velocities of the various ion Stamnes, 1981; PullJam et al., 1981]. species are in agreement also lends strong High-altitude (>7000 kin) particle measurements credence to the parallel electric field hypothe- have also been shown to be consistent with the sis. The acceleration mechanism cannot be a parallel electric field hypothesis. The presence simple pa-allel electric field, however, for two of an auroral electrostatic potential difference reasons: first, the characteristic energy E o is below a spacecraft can be inferred by two inde- considerably larger than the ionospheric thermal pendent additional techniques: the energy of an energy (E o is typically hundreds of electron upflowing (pitch angle • 180 ø) ion beam, and the volts and 20-30% of Epeak), and second, the energy-dependent widening of the loss cone of energy Epeak is typically 30-50% smaller than electrons mirroring belowthe spacecraft [Ghiel- that inferred by the two other independent tech- metti et al., 1978; Cladis and Sharp, 1979; o niques. The distribution does appear to be Gorneyet al., 1981; Fennell et al., 1981; Mizera consistent with an ionospheric source, heated et al., 1981; Greenspan et al., 1981]. Thesetwo within (or above)the acceleration region, since techniques have beenshown to be consistentwith the ion average energy is comparable to e•. The each other [Mizera et mi., 1981; Coilin et al., 1 986a ]. Although many observations are consistent with this picture (see Reiff [1983] for a list of •Department of Space Physics and Astronomy, recent papers and review articles), models have Rice University, Houston, Texas. been proposed in which the primary acceleration 2Lockheed Palo Alto Research Laboratory, Palo mechanism is lower hybrid wave turbulence [Bing- Alto, California. ham et al., 1984] or ion acoustic turbulence 'Department of Physics and Astronomy, Univer- [Stasiewiez, 1984]. Stasiewiez's argument slty of Iowa, Iowa City. against the standard model is that the secondary •Southwest Research Institute, San Antonio, electron distribution should have an E-2 distri- Texas. bution, whereas E -• is more commonly observed. 5Institute of Geophysics and Planetary However, including the flux of scattered and Physics, University of California, Los Angeles, energy-degraded primary electrons with the secon- California. daries, as was done by PullJam et al. [1981], does yield good agreement with particle flux Copyright 1988 by the American Geophysical Union. measurements. Wave-particle interactions have also been suggested as the cause of the electron Paper number 6A8645. fluxes observed below the energy of the primary 0148-0227/88/006A-8645505.00 peak [e.g., Stasiewiez, 1985]. 7441