QuakeFinder's Algorithm Results for Forecasting Earthquakes
Abstract
A >10 year effort of data acquisition has yielded a dataset with outstanding spatial and temporal sampling of Earth's magnetic field. The QuakeFinder (QF) network consists of 150 stations with the majority in California along the San Andreas Fault and some along faults in Greece, Taiwan, Peru, Chile and Indonesia (Warden et al. 2018). Each station is equipped with 3 feedback induction magnetometers, 2 ion sensors, a horizontal 4 Hz geophone, a temperature sensor, and a humidity sensor. The data are continuously recorded at 50 samples per second with GPS antennas supplying reference timestamps.
QF is attempting to detect anomalous electromagnetic activity occurring ahead of earthquakes and is currently focused on their magnetometer data. Since QF has acquired >70 TB of data, they have developed an algorithmic framework to support an automated approach for finding earthquake precursory statistical significance. The algorithm framework can be scaled to an arbitrary number of stations and time intervals of investigation (one implementation is currently in review). At previous AGU conferences, the QF algorithmic framework was outlined and the importance of data QC was presented. Now, the results and their statistical significance will be discussed from one implementation of the algorithm framework applied to the QF California dataset. Further, QF will present various experiments towards improving results. These experiments may include extending the time span of analysis, removing large known noise sources (such as the Pacific DC Intertie), bandpassing the data, etc.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMNH52B..05S
- Keywords:
-
- 2427 Ionosphere/atmosphere interactions;
- IONOSPHERE;
- 4302 Geological;
- NATURAL HAZARDS;
- 7223 Earthquake interaction;
- forecasting;
- and prediction;
- SEISMOLOGY;
- 7999 General or miscellaneous;
- SPACE WEATHER