Search

Aurora Computed Tomography (ACT) is a method for retrieving the three-dimensional (3-D) distribution of the volume emission rate from monochromatic auroral images obtained simultaneously by a multi-point camera network. We extend this method to a Generalized-Aurora Computed Tomography (G-ACT) that reconstructs the energy and spatial distributions of precipitating electrons from multi-instrument data, such as ionospheric electron density from incoherent scatter radar, cosmic noise absorption (CNA) from imaging riometers, as well as the auroral images. The purpose of this paper is to describe the reconstruction algorithm involved in this method and to test its feasibility by numerical simulation. Based on a Bayesian model with prior information as the smoothness of the electron energy spectra, the inverse problem is formulated as a maximization of posterior probability. The relative weighting of each instrument data is determined by the cross-validation method. We apply this method to the simulated data from real instruments, the Auroral Large Imaging System (ALIS), the European Incoherent Scatter (EISCAT) radar at Tromsø, and the Imaging Riometer for Ionospheric Study (IRIS) at Kilpisjärvi. The results indicate that the differential flux of the precipitating electrons is well reconstructed from the ALIS images for the low-noise cases. Furthermore, we demonstrate in a case study that the ionospheric electron density from the EISCAT radar is useful for improving the reconstructed electron flux. On the other hand, the incorporation of CNA data into this method is difficult at this stage, because the extension of energy range to higher energy causes a difficulty in the reconstruction of the low-energy electron flux. Nevertheless, we expect that this method may be useful in analyzing multi-instrument data and, in particular, 3-D data, which will be obtained in the upcoming EISCAT_3D.

Using the auroral boundary index derived from DMSP electron precipitation data and the Dst index, changes in the size of the auroral belt during magnetic storms are studied. It is found that the equatorward boundary of the belt at midnight expands equatorward, reaching its lowest latitude about one hour before Dst peaks. This time lag depends very little on storm intensity. It is also shown that during magnetic storms, the energy of the ring current quantified with Dst increases in proportion to Le–3, where Le is the L-value corresponding to the equatorward boundary of the auroral belt designated by the auroral boundary index. This means that the ring current energy is proportional to the ion energy obtained from the earthward shift of the plasma sheet under the conservation of the first adiabatic invariant. The ring current energy is also proportional to Emag, the total magnetic field energy contained in the spherical shell bounded by Le and Leq, where Leq corresponds to the quiet-time location of the auroral precipitation boundary. The ratio of the ring current energy ER to the dipole energy Emag is typically 10%. The ring current leads to magnetosphere inflation as a result of an increase in the equivalent dipole moment.Key words. Ionosphere (Auroral ionosphere) · Magnetospheric physics (Auroral phenomena; storms and substorms)