Investigation of precursor signals of earthquakes using triangular area change method and CORS-GNSS networks

This thesis aims to systematically investigate the detectability of potential precursor signals preceding destructive earthquakes in Türkiye between 2011 and 2023 by employing the "Triangular Area Change Analysis" method. The study focuses on the earthquakes that occurred on 23 October 2011 in Van (Mw = 7.2), 21 July 2017 in the Gökova Gulf (Mw = 6.6), 30 October 2020 in İzmir (Mw = 6.9), and 6 February 2023 in Kahramanmaraş (Mw = 7.7–Mw = 7.6). Using the North (N), East (E), and Up (U) coordinate data from GNSS stations operating within the TUSAGA-Aktif network located around the epicenters of these earthquakes, the temporal variations of triangular areas formed by stations surrounding each epicenter were analyzed via a Python-based code developed on the Google Colab platform. Triangular area changes were calculated and interpreted separately for all planes of the 3D triangles (X–Y, Y–Z, and X-Z).The graphical analyses revealed that, for the Kahramanmaraş earthquake, distinct anomalies in area change emerged up to five days before the event. In contrast, for the İzmir earthquake, due to the insufficient number and heterogeneous distribution of GNSS stations, no significant pattern could be identified in the area change graphs, indicating that the applicability of the method is directly dependent on network geometry and station density. For the Van and Gökova Gulf earthquakes, notable area changes were also observed up to five days prior to the events.The results suggest that increasing the number of high-precision GNSS stations, homogeneously distributed along fault zones across Türkiye, would enhance the reliability of the applied method.