Magnitude Estimation for Earthquake Early Warning with Multiple Parameter Inputs and a Support Vector Machine

Abstract

Accurately estimating the magnitude within the initial seconds after the P-wave arrival is of great significance in earthquake early warning (EEW). Over the past few decades, single-parameter approaches such as the τc and Pd methods have been applied to EEW magnitude estimation studies considering the first 3 s after the P-wave onset. However, these methods present considerable scatter and are affected by the signal-to-noise ratio (SNR) and epicentral distance. In this study, using Japanese K-NET strong-motion data, we propose a machine-learning method comprising multiple parameter inputs, namely, the support vector machine magnitude estimation (SVM-M) model, to determine earthquake magnitudes and resolve the aforementioned problems. Our results using a single seismological station record show that the standard deviation of the magnitude prediction errors of the SVM-M model is 0.297, which is less than those of the τc (1.637) and Pd (0.425) methods. The magnitudes estimated by the SVM-M model within 3 s after the P-wave arrival are not obviously affected by the SNR or epicentral distance, and not overestimated for MJMA≤5. In addition, in an offline EEW application, the magnitude estimation error of the SVM-M model gradually decreases with increasing time after the first station is triggered, and the underestimation of event magnitudes for 6.5≤MJMA gradually improves. These results demonstrate that the proposed SVM-M model can robustly estimate earthquake magnitudes and has potential for EEW.

Advanced Seismic Magnitude Classification Through Convolutional and Reinforcement Learning Techniques

“…During the evaluation phase, a comprehensive comparative analysis was conducted to assess the performance of our proposed model in comparison to five distinct deep learning frameworks: SeisNet [25], EEWMagNet [32], MagEstNet [39], QuakeClassNet [40], QuakeNet [41]. This rigorous examination aimed to offer a holistic perspective on our model's efficacy in relation to contemporary methodologies.…”

Application of Model-Based Time Series Prediction of Infrared Long-Wave Radiation Data for Exploring the Precursory Patterns Associated with the 2021 Madoi Earthquake

“…SVM is a supervised learning method in machine learning that transforms multiple features nonlinearly via the kernel function, aiming to extract more information from multiple feature inputs and obtain more accurate prediction results [31]. The specific mathematical description is as follows:…”

Respecting Nature: The Centrality of Village Life in Jiangxi Province, South East China

“…Implementing overly strict criteria, such as requiring too many or a large number of stations to trigger, can negatively impact the real-time efficiency of EEW systems, while loose criteria can result in false alarms 5 .Theoretically, earthquake parameter determination may require data from at least four triggered stations to ensure accuracy [5][6][7] . The magnitude determination often requires 3 s P arrivals for a single station [8][9] . Therefore, the time delay for issuing a warning is the duration from the origin time to 3 seconds after the last station triggers when multiple stations are used to estimate the magnitude.…”

“…Theoretically, earthquake parameter determination may require data from at least four triggered stations to ensure accuracy [5][6][7] . The magnitude determination often requires 3 s P arrivals for a single station [8][9] . Therefore, the time delay for issuing a warning is the duration from the origin time to 3 seconds after the last station triggers when multiple stations are used to estimate the magnitude.…”

“…evaluation [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23] . With recent advancements in machine learning, the automatic construction of earthquake catalogs is now feasible without the need for manual intervention [21][22][23] , which significantly enhances our ability to monitor and comprehend seismic activity.…”