ELVES Measurements in the “UV Atmosphere” (Mini-EUSO) Experiment Onboard the ISS and Their Reconstruction

More than three dozen submillisecond events of ELVES type ("elves"), which are the result of the interaction of the front of an electromagnetic pulse from a lightning discharge and the lower layer of the ionosphere, have been identified in the data of a UV Atmosphere orbital multichannel detector (Mini-EUSO). Each event has a characteristic annular glow pattern and occupies a significant part of the detector's field of view, and the signal in a separate channel has an asymmetric profile with a pronounced peak. The distribution of peak times contains information about both the localization of the discharge and the altitude of the glow. In this paper, we propose a Bayesian (probabilistic) model for reconstructing ELVES events, implemented using probabilistic programming methods in PyMC-5. The capabilities of the model for determining the position of the discharge are shown using the example of several events. Methods for modifying the model to restore the discharge orientation and refine the glow height are outlined.