Segmentation of PMSE Data Using Random Forests

EISCAT VHF radar data are used for observing, monitoring, and understanding Earth’s upper atmosphere. This paper presents an approach to segment Polar Mesospheric Summer Echoes (PMSE) from datasets obtained data. The consist of 30 observations days, corresponding 56,250 samples. We manually labeled the into three different categories: PMSE, Ionospheric background, Background noise. For segmentation, we employed random forests on a set simple features. These features include: altitude derivative, time mean, median, standard deviation, minimum, maximum values neighborhood sizes ranging 3 by 11 pixels. Next, in order reduce model bias variance, method that decreases weight applied pixel labels with large uncertainty. Our results indicate that, first, it is possible PMSE using forests. Second, weighted-down technique improves performance method.