Improving the Foreshock Traffic Light Systems for Real-Time Discrimination Between Foreshocks and Aftershocks

After a moderate-to-large earthquake, Civil Protection, scientists, the population, and decision makers immediately ask: "Was this the mainshock, or is a bigger event yet to come?" According to the current state of the art, the probability that an even larger event will occur within five days and 10 km after a moderate earthquake is typically 5%. Recent work proposes that a more specific answer to this question is possible (Gulia and Wiemer, 2019). The method allows real-time discrimination between foreshocks and aftershocks in well-monitored regions. However, so far, some expert judgments are required to adapt the technique to the case-specific data, such as defining the overall magnitude of completeness and the duration of the short-term aftershock incompleteness. Here, we introduce and retrospectively test an improved version of the foreshock traffic light system (FTLS) approach that successfully overcomes the aforementioned limits using the b-positive estimator (van der Elst, 2021). With these improvements, an FTLS assessment can be fully automated as early as one hour after an M ≥ 6 event without requiring expert judgments. We also test the updated approach on six additional sequences; the results align with the FTLS hypothesis.