Automated Identification of Auroral Luminosity Boundaries using pyIntensityFeatures

Imagers that observe emissions from the atmosphere are commonly used to study various ionospheric phenomena.  These phenomena include the auroral oval, equatorial plasma bubbles, and travelling ionospheric disturbances.  A difficulty in using imager observations is accurately and automatically retrieving the locations of interest from these images.  We present an automated method designed to identify the auroral luminosity boundaries from space-based imager data.   These boundaries are important for high-latitude studies that use statistical or machine learning approaches, as geographic and magnetic coordinate systems that do not account for changes in the polar cap or equatorward auroral oval boundaries will mix together data from regions experiencing different types of coupling with the magnetosphere.

The boundary identification method was originally developed for the Imager for Magnetopause-to-Aurora Global Exploration (IMAGE) observations, and has been further adapted for use in a wider variety of situations.  We will discuss the updated detection method and demonstrate the process on two different satellite data sets.  The updated detection method will be made publicly accessible through a new Python package, pyIntensityFeatures.