Enhanced Detection of P-wave onset: A Novel Algorithm for Accurate P-Wave Picking

Sandeep .

Submitted September 7, 2025, SCEC Contribution #14239, 2025 SCEC Annual Meeting Poster #TBD

Globally, Earthquake Early Warning (EEW) system function as a national disaster mitigation
technique for seismically active countries. Accuracy in P-wave picking being one of the
major hurdles in EEW system for providing timely alerts in different regions worldwide. This
research introduces a novel P-wave picking algorithm designed for P-Alert data, aiming to
enhance the accuracy and reliability of EEW system. It utilizes the concept of an envelope,
which is constructed from the absolute value of the analytic signal ℵ(s(t)). This analytic
signal composed of the original waveform s(t) and its Hilbert transform h(t). Subsequently, a
threshold value is calculated based on the standard deviation of the envelope to identify
sudden changes in envelope amplitude. Unlike traditional (STA/LTA, STA-LTA, AIC and
wavelet transform) and advance methods (Artificial neural networks, Wave polarization
analysis) this algorithm doesn’t require prior knowledge or tuning of parameters related to
window length, signal intensity, frequency content, and detection thresholds. Additionally,
this developed algorithm is not sensitive to noise spikes in the data and performs well even in
low SNR conditions. Furthermore, this algorithm has been tested on 900+ earthquake records
of various magnitudes. The maximum relative error of 4% from all records confirms the
applicability of this algorithm. The newly developed algorithm is crucial for minimising false
alarms and enhancing the reliability of EEW systems, which are essential in earthquake-
prone regions worldwide.

Key Words
Earthquake Early warning System

Citation
., S. (2025, 09). Enhanced Detection of P-wave onset: A Novel Algorithm for Accurate P-Wave Picking. Poster Presentation at 2025 SCEC Annual Meeting.

Related Projects & Working Groups
Earthquake Forecasting and Predictability (EFP)