Development of a community-engaged, low-cost earthquake early warning system using MEMS-based sensors : enhancing and adapting the PLUM algorithm with decentralised processing and P-wave integration : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Emergency Management at Massey University, Wellington, New Zealand
Loading...
Files
Date
2025-03-04
DOI
Open Access Location
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Massey University
Rights
© The Author
Abstract
Earthquakes pose a significant threat to people and infrastructure, particularly in regions near active faults or offshore subduction zones, which are more frequently exposed to moderate to strong shaking. Earthquake Early Warning Systems (EEWS) provide crucial alerts immediately following an earthquake, offering a warning window ranging from a few seconds to tens of seconds. These systems have effectively reduced damage and allowed individuals to take protective actions. However, the high cost of establishing high-end EEWSs makes them unaffordable for many countries.
To address this, there is growing interest in using low-cost technologies such as Micro-Electromechanical Systems (MEMS)-based ground motion sensors to implement EEWSs. However, despite their potential, several knowledge gaps must be addressed to enhance their efficiency and effectiveness. Firstly, further investigation into decentralised processing for earthquake detection and alert generation is required. Traditional high-end EEWSs often rely on centralised processing units, which have proven vulnerable to critical delays and communication failures during major seismic events. Secondly, adapting and improving ground-motion-based or wave-field-based EEW algorithms is crucial for enhancing their performance, ensuring that EEWSs can provide timely and effective warnings in all regions during an earthquake, as opposed to the limitations posed by traditional source-based methods.
This doctoral research addresses these gaps by developing and evaluating a community-engaged, low-cost MEMS-based EEWS. The system utilises a ground-motion-based EEW algorithm adapted for decentralised processing, enabling rapid earthquake detection. It also integrates a P-wave detection algorithm to enhance the performance of the ground-motion-based approach. Guided by the Design Science Research methodology, this study seeks to answer three key research questions: (1) How can the Propagation of Local Undamped Motion (PLUM) ground-motion-based EEW algorithm be adapted and implemented for New Zealand’s seismic conditions using decentralised processing? (2) How can high-accuracy P-wave detection be achieved in a community-engaged EEW network with high ambient noise? (3) How can the P-wave detection algorithm be integrated into the adapted PLUM algorithm to extend the warning window?
The study begins with a comprehensive literature review to identify research gaps in low-cost MEMS-based EEWSs, leading to the formulation of the research questions addressed in this thesis. To answer these questions, an experimental community-engaged EEW network was implemented in Greater Wellington, NZ, using low-cost MEMS-based sensors. This implementation was followed by adapting the PLUM algorithm to NZ-PLUM, making it compatible with New Zealand’s seismic intensity by employing region-specific Ground Motion Intensity Conversion Equations. The NZ-PLUM algorithm was then integrated into a sensor network operating under a decentralised processing architecture using a two-tier communication model, ensuring rapid and reliable data transmission and processing.
Building upon the implementation of the NZ-PLUM algorithm, integrating a P-wave detection algorithm into the NZ-PLUM approach was explored to extend the warning window. A performance analysis is conducted to identify the most effective P-wave detection algorithm for integration into the community-engaged EEWS. Subsequently, an empirical relationship between P-wave and S-wave amplitudes is established, leading to the development of a P-wave-based PLUM algorithm (NZ-PLUM-P), which provides an extended warning window before the onset of seismic shaking.
The outcomes of this doctoral research make significant advancements in community-engaged, low-cost EEWSs. A key contribution is developing a real-life experimental EEW network using two distinct algorithms, NZ-PLUM and NZ-PLUM-P, tailored to NZ’s seismic context within a decentralised processing architecture. This study offers a versatile framework applicable to implementing community-engaged EEW networks at a low cost, making a substantial contribution to theory and practice. The methods developed for P-wave detection, constructing P-S wave amplitude relationships, executing EEW algorithms using decentralised processing and evaluating EEW network performance provide valuable tools for future research and implementation. Further, this cost-effective, community-driven model not only offers a viable solution for seismically active nations with limited resources but also has the potential to enhance the performance of existing high-end EEWS by increasing sensor density and extending warning capabilities. Providing earthquake early warnings can potentially be crucial in saving lives, protecting critical infrastructure, and enhancing public preparedness.
Description
Keywords
Earthquake Early Warning (EEW), earthquake detection, community-engaged EEW, low-cost network, decentralised processing, seismology & signal processing, sensor networks & data processing, disaster preparedness, seismic risk reduction, Internet of Things (IoT)