The Red Sprite type of upward discharges exhibit predominantly red color, enduring for few milliseconds and extending in altitude from ~50 to 90 km [e.g., Sentman et al., 1995]. Red Sprites are typically associated with 1 out of 20-40 especially energetic positive cloud-to-ground (CG) discharges [Sentman and Wescott, 1993; Sentman et al., 1995]. Positive CG discharges can involve transfer (to the ground) of up to 300 C in several ms [e.g., Brook et al., 1982], resulting in large (up to ~1000 V/m at 50 km altitude) QE fields due to the uncompensated negative charge left in and above the cloud. The ambient electrons and ions at all altitudes above the cloud are heated by the large QE fields, leading to optical emissions. The observed several to tens of ms duration of Red Sprites is consistent with the characteristic relaxation time of QE fields due to finite conductivity of the medium [Dejnakarintra and Park, 1974; Baginski et al., 1988].
In our model [Pasko et al., 1995] we consider a QE field established in the mesosphere and lower ionosphere due to the accumulation of large thundercloud charge and its evolution in time when the charge is brought to the ground (exponentially with a time constant of 1 ms) by a CG lightning stroke. We use a cylindrical coordinate system (r, phi, z) with the z axis representing altitude. The ground and the cylindrical boundaries at z=90 km and r=60 km are assumed to be perfectly conducting. The effect of the artificial conducting boundary at r=60 km on the QE fields inside the system is small (e.g., ~10% at r=50 km, z=10 km). The ambient ion conductivity is assumed to vary exponentially with altitude, with a scale height of 6 km [Dejnakarintra and Park, 1974]. The electron component of the ambient sigma is based on the `tenuous' ambient density model of Taranenko et al. [1993a], ionization rates for atmospheric and ionospheric breakdown of Papadopoulos et al. [1993], and the electron mobility as a function of electric field obtained from experimental data [Davies, 1983]. Electrons are assumed to have a small but non-negligible number density at mesospheric altitudes [Reagan et al., 1981]. We neglect chemical effects due to their relatively long time scales [e.g., Mitra, 1981] and the geomagnetic field since the plasma is highly collision dominated even after minor heating. Numerically we solve Poisson equation for the electric field and conservation of charge equation for the charge density with a given (in space and time) thundercloud charge distribution (please see [Pasko et al., 1995] for detailed description of the model).
Results will be shown for a case of dipole charges +Q, -Q (Q=200C) which are assumed to form in a thundercloud over a time tau_f (0.5 sec). The positive half of the dipole charge is then discharged to ground with a time constant tau_s=1 ms. As the charges accumulate, high altitude regions where tau_f is greater than the local relaxation time are shielded from the QE fields of the thundercloud charges by induced space charge at lower altitudes. When charge is quickly removed, a QE field appears at all altitudes above the shielding space charge. Since charge removal can be viewed as `placement' of an identical charge of opposite sign, the initial field at ionospheric altitudes is approximately the free space field due to the `newly placed' charge. Thus, the important physical consequences of the QE system depend on the magnitude and altitude of the removed charge and are essentially independent of the initial charge configuration.
Lets get started.
Title of the Spring 1995 AGU talk: title.mpg (21K)