Temporal evolution of entropy and chaos in low amplitude seismic wave prior to an earthquake

Abstract

This study investigates the temporal changes of entropy and chaos in low-amplitude continuous seismic wave data prior to two moderate-level earthquakes. Specifically, we examine seismic signals before and during the Istanbul-Turkey earthquake of September 26, 2019 (M = 5.7), and the Duzce-Turkey earthquake of November 17, 2021 (M = 5.2), which occurred near the Marmara Sea region on the north-Anatolian fault line. We aim to identify changes in complexity and chaotic characteristics in the pre-earthquake seismic waves and explore the possibility of earthquake forecasting minutes before an earthquake. To accomplish this, we utilize windowed scalogram entropy and sample entropy methods and compared the results with Lyapunov exponents and windowed scale index. Our findings indicate that measuring the temporal change of entropy using windowed scalogram entropy is sensitive to the change in complexity due to the frequency shifts during the weak ground motion approaching an earthquake. On the other hand, Lyapunov exponents and sample entropy appear more effective in their response to the change in complexity and chaotic characteristics due to the change in the signal amplitude. Additionally, the windowed scale index can detect temporal fluctuations in the aperiodicity of the signal. Overall, our results suggest that all four methods can be valuable in characterizing complexity and chaos in short-time pre-earthquake seismic signals, differentiating earthquakes, and contributing to the development of earthquake forecasting techniques. © 2023 Elsevier Ltd