A multi-lag transfer entropy framework for seismic network analysis: persistent marmara isolation following the 2023 Kahramanmaraş sequence

Abstract We develop a mathematical framework for analyzing inter-regional earthquake triggering using multi-lag transfer entropy. While transfer entropy itself is well-established in information theory, its systematic application to discriminate earthquake triggering mechanisms across tectonic regions has not been previously formalized. We introduce the Multi-Lag Significance Profile (MLSP) as a binary vector encoding statistical significance across multiple time lags and prove that specific MLSP patterns correspond to distinct physical triggering mechanisms. Applying this framework to 53,951 earthquakes ( $$M \ge 2.0$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>M</mml:mi> <mml:mo>≥</mml:mo> <mml:mn>2.0</mml:mn> </mml:mrow> </mml:math> ) from Turkey (2021–2025), including the catastrophic Kahramanmaraş $$M_w$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msub> <mml:mi>M</mml:mi> <mml:mi>w</mml:mi> </mml:msub> </mml:math> 7.7 doublet, we establish three principal results: (1) Marmara remains highly isolated across both windows (Regional Isolation Index ( $$\textrm{RII}= 1.000$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mtext>RII</mml:mtext> <mml:mo>=</mml:mo> <mml:mn>1.000</mml:mn> </mml:mrow> </mml:math> pre-earthquake; 0.917 post-earthquake)), with no robust coupling pathways after applying surrogate- and embedding-based robustness checks—this is the most robust finding of the study; (2) The Eastern–Southeastern Anatolia connection exhibits $$\textbf{S}= (1,1,1)$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mi>S</mml:mi> <mml:mo>=</mml:mo> <mml:mo>(</mml:mo> <mml:mn>1</mml:mn> <mml:mo>,</mml:mo> <mml:mn>1</mml:mn> <mml:mo>,</mml:mo> <mml:mn>1</mml:mn> <mml:mo>)</mml:mo> </mml:mrow> </mml:math> in both directions in the immediate post-earthquake period; however, temporal windowing and declustering analyses reveal this coupling is primarily aftershock-driven rather than persistent tectonic interaction; (3) The earthquake triggered significant network reorganization, with new coupling pathways emerging in the epicentral region. All findings are validated through false discovery rate (FDR) correction, dependence-preserving surrogate tests, embedding sensitivity analyses, and comprehensive robustness analyses. These mathematically rigorous findings have direct implications for Istanbul seismic hazard assessment, particularly given recent observations of progressive eastward rupture along the Main Marmara fault [1, 2].