← Back to Sessions

False and Missed Alerts: A Performance Analysis of a Community-Engaged Earthquake Early Warning System

Description: 

False and missed alerts are two key concerns regarding Earthquake Early Warning (EEW) Systems. Every year, earthquakes strike New Zealand, making it one of the most earthquake-prone countries on the planet. Thousands of quakes occur each year, and around 150 earthquakes are felt. Citizen seismology, which refers to research partnerships between seismologists and non-scientist volunteers, is a developing field. In this study, citizen seismology encourages the public to participate in gathering ground motion data for EEW. Community volunteers host low-cost Raspberry Shake ground motion sensors in their homes. The volunteers have the flexibility to install the sensors at their own discretion; the sensor's location in their homes and installation methods are not pre-defined. On 22 September 2022, a moderate M5.2 earthquake occurred in the Cook Strait region of New Zealand, which caused moderate levels of shaking in the lower part of the North Island, mainly in the greater Wellington region. Fortunately, the experimental community-engaged Earthquake Early Warning System (EEWS) was operating during this earthquake. Preliminary analysis of the ground motion data collected during the earthquake event demonstrates that EEW is achievable with a citizen seismology-based approach. However, the reliability and accuracy of the EEWS must be evaluated in depth to understand whether this network is capable of generating accurate warnings. Therefore, this study analyses the false and missed alerts of the implemented EEWS during a 48-hour time window of the mentioned earthquake using four different P-wave detection algorithms. Results show that a wavelet transformation-based P-wave picker is the most suitable algorithm for the community-engaged EEWS in detecting an earthquake with minimal false and missed alerts. By assessing the performance through false and missed alerts, this study discusses whether a citizen seismology-based approach is appropriate for an EEWS.

Session: Earthquake Early Warning Optimization and Efficacy

Type: Oral

Date: 4/20/2023

Presentation Time: 02:00 PM (local time)

Presenting Author: Chanthujan Chandrakumar

Student Presenter: Yes

Invited Presentation: 

Authors