Federico Speranza

Existing, commonly used, models of high-latitude electric fields and conductivities have been derived from ground-based radars and satellites since the 1970s. These have largely used measurements of plasma densities and velocities, or magnetic field measurements to derive the electric field, or the conductivity, individually and separately. These models are then used as drivers in coupled ionosphere-thermosphere models, which then calculate the response of the whole upper atmosphere system. However, the feedback contribution of the neutral atmosphere is not incorporated into these electric field and conductivity models. So each new model step uses an unmodified driver. For this preliminary study we will apply Machine Learning techniques to look for signatures of feedback between electric field and conductivities in EISCAT radar and satellite databases. We will also look at the contribution of the neutral atmosphere to the coupled system, in terms of neutral dynamics and temperatures, and, in particular, to the chemical changes due to upwelling of the neutral atmosphere measured by Fabry-Perot Interferometers co-located with the EISCAT radars.