Patterns of the meteor head echoes observed by the Jicamarca High-Power Large-Aperature radar

Millions of interplanetary and interstellar particles constantly hit the Earth's upper atmosphere. Those particles can be detected as meteor echoes by radar at 70km to 140km altitude. In this study, we present a detailed analysis of the meteor head echoes observed by the Jicamarca High-Power Large-Aperture radar and the corresponding processing algorithms for obtaining the results. The meteor head echoes are detected by the Algorithm for Head Echo Automatic Detection (AHEAD), and a ResNet-18 Convolutional Neural Network (CNN) is used to remove false alarms. The workflow highlights the capabilities of automatically resolving interferometric aliasing by stochastic reconstruction. Doing so substantially increases the accuracy of inferring the direction of vector velocity of the meteors. Then, the detected meteor head echoes are automatically classified into five categories by a Long Short-Term memory (LSTM) network. These categories included meteor head echoes exhibiting regular or irregular interference patterns, sudden drops in received power, and displaying radar patterns, i.e., traveling between radar lobes and others. This study presents meteor head echo examples of different categories and discusses the potential causes of those features.