The 2023 Mw 6.8 Morocco earthquake induced atmospheric and ionospheric anomalies

Earth observations through Global Navigation Satellite System (GNSS) and Remote Sensing (RS) technologies play a significant role in natural hazard surveillance, particularly in the context of earthquake prediction and detection. This study introduces a distinctive Deep Learning (DL) based approach to identify ionospheric and atmospheric precursors, utilizing data from multiple satellite sources and provides a comprehensive analysis of spatiotemporally varying precursors, contributing to the understanding and monitoring of seismic activity in earthquake-prone regions. In our investigation of the Morocco earthquake on September 08, 2023 (Mw 6.8), we analyzed various precursors including Total Electron Content (TEC), Air Pressure (AP), Relative Humidity (RH), Outgoing Longwave Radiation (OLR), and Air Temperature (AT). Our study aims to identify a synchronized anomalous window of potential earthquake precursors using Standard Deviation (STDEV), Continuous Wavelet Transform (CWT), and Long Short-Term Memory Inputs (LSTM) network. Both statistical and deep learning methods revealed abnormal fluctuations as precursors occurring within 8-9 days before the earthquake near the epicenter. Additionally, we detected geomagnetic anomalies in the ionosphere 6 days prior to and 4 days after the earthquake, coinciding with active geomagnetic storm days. This research underlined the importance of combining multiple earthquake precursors using statistical and deep learning approaches to support the understanding of the Lithosphere-Atmosphere-Ionosphere-Coupling (LAIC) phenomena.