The observations were made at the same time that ionospheric electric fields and plasma number densities were measured in situ by instruments on sounding rockets. Neutral wind profiles were also measured during the campaign from triangulation of chemiluminescent trails from rocket releases. Aperture synthesis radar imaging techniques permit the sorting of the coherent backscatter into small azimuth and range bins and the determination of the scattering altitude. Individual Doppler spectra could thereby be unambiguously associated with in situ electric field measurements in the same small volume. We find that the Doppler shifts of the auroral echoes correspond to the ion acoustic speed times the cosine of the flow angle, where the former is predicted according to an empirical wave heating law. Type I echoes are only observed for very small flow angles regardless of the convection speed.

This article presents measurements of electron heating in the high‐latitude E region during rare, extreme convection electric fields (up to 160 mV/m). As inferred from the merged Sondrestrom incoherent scatter spectra measurements, the electron temperature monotonically increases with the electric field reaching 4000+ K, without any indication of saturation, and the radar backscatter power monotonically decreases.

An algorithm has been developed to image the local structure in the convection electric field using multibeam incoherent scatter radar (ISR) data. The imaged region covers about 4° in magnetic latitude and 8° in magnetic longitude for the specific geometry considered (that of the Poker Flat ISR). The algorithm implements the Lagrange method of undetermined multipliers to regularize the underdetermined problem posed by the radar measurements. The error on the reconstructed image is estimated by mapping the mathematical form to a Bayesian estimate and observing that the Lagrangian method determines an effective a priori covariance matrix from a user‐defined regularization metric. There exists a unique solution when the average measurement error is smaller than the average measurement amplitude. The algorithm is tested using synthetic and real data and appears surprisingly robust at estimating the divergence of the field. Future applications include imaging the current systems surrounding auroral arcs in order to distinguish physical mechanisms.

[1] Empirical models of the Poynting flux and particle kinetic energy flux, associated with auroral processes, have been constructed using data from the FAST satellite and are available online. The models output flux maps as a function of several geophysical parameters: (1) clock angle of the interplanetary magnetic field (IMF), (2) magnitude of the IMF in the GSM y-z plane, (3) solar wind speed, (4) solar wind number density, (5) magnetic dipole tilt angle, and (6) the AL index (optional for Poynting flux). The choice of parameters is motivated by the Weimer potential model. Because the Poynting flux distribution has a heavy tail, care must be taken in applying the model to events that may be uncommon, and the model output includes a measure of quality based on the density of FAST orbits in the parameter space. The models are constructed by fitting the data to a sum of empirical orthogonal functions (EOFs), with coefficients modeled by quadratic equations in the geophysical parameters. The EOFs are constructed from the same data set using singular value decomposition, along with a smoothing/interpolation algorithm that minimizes curvature and incorporates uncertainty. Potential applications include specification of the auroral energy input for general circulation models. Basic findings from the model include verification of the importance of the auroral electrojets as features of auroral Poynting flux deposition and identification of the cusp as a third important feature. The cusp makes an important contribution to the overall energy budget when the IMF is north of˙90 ı .