Traveling Ionospheric Disturbance (TID) Detection using Deep Learning Models

Traveling Ionospheric Disturbances (TIDs) are wave-like perturbations in ionospheric electron densities and play a significant role in exchange of momentum and energy between various regions of the upper atmosphere. In this study, we focus on concentric TIDs induced by different climatological events like hurricanes, tornadoes, or convective storms. The aim is to develop deep learning models that can effectively detect the concentric TIDs and extract their characteristic features, including wavelength, wave speed, and frequency. The detrended Total Electron Content (dTEC) data collected from dense GNSS network over the contiguous US have been used for training the deep-learning TID detection models. Our results show that these models can detect the TID regions in the dTEC maps with bounding boxes effectively. These models can be deployed as an automatic tool for real-time TID detection and further developed to extract their wave parameters. This study will strongly enhance our capability in wave pattern recognition from GNSS observations.