Kernel density estimation approach to characterizing field-aligned irregularities from equatorial atmosphere radar observations

Abstract Field-aligned irregularities (FAI) are a persistent feature of the equatorial ionosphere and can significantly impact satellite-based communication and navigation systems. Despite extensive documentation of their large-scale occurrence patterns, there is still a lack of understanding regarding their short-term temporal variability and detection uncertainty. To address this, a statistical framework based on Kernel Density Estimation (KDE) was developed to investigate the temporal characteristics of FAI events with higher resolution. Data from the Equatorial Atmosphere Radar (EAR) in West Sumatra, Indonesia, were analyzed, focusing on two years: 2016 (a leap year) and 2017 (a non-leap year). KDE was applied to generate smoothed daily probability estimates of FAI occurrence, along with associated confidence intervals, allowing the temporal evolution of FAI activity to be visualized more clearly. To examine short-term variability, FAI events are grouped into 1-day, 2-day, and 3-day sequential patterns. Results show consistent seasonal signatures across both years, suggesting stable ionospheric behaviours despite differences in calendar structure. The KDE approach captures fluctuations more clearly than standard methods and highlights subtle patterns in event occurrences. This method offers a reproducible way to study FAI dynamics and can be extended to multi-year or multi-site analyses. It supports a more complex conception of equatorial ionospheric variability and its relevance to space weather monitoring and forecasting, where precise characterization of ionospheric disturbances is essential.