Automatic Detection of Naturally Enhanced Ion-Acoustic Lines in EISCAT data

Abstract

Naturally Enhanced Ion-Acoustic Lines (Neials) are small-scaled and short-lived phenomena in the ionosphere, which is widely researched internationally. At present, the search for Neials is a tedious task that has to be done through visual inspection. This is often a time-consuming process and limits the research on Neials. By automating the means of locating the Neials, the access to them will be improved and it will be more feasible to study the phenomenon.

In this thesis, routines for automatically detection Neials in EISCAT data will be presented. These routines will then be utilized to construct a program which will search for Neials. This program will be using the spectra from incoherent scatter radar experiments as the main resource. The spectra are generated using Guisdap on raw data obtained from the EISCAT database.

The program may be used such that it either picks out as many Neials as possible, as few false positives as possible, or a mixture of these two modes. In the case of picking out as many Neials as possible, the program is able to detect almost 90% of all Neials, where roughly 60% of the reported dumps are false positives. In the case of minimizing false positives, the program picks out roughly 60% of all Neials, but selects almost zero false positives.