A paradigm for developing earthquake probability forecasts based on geoelectric data

We examine the precursory behavior of geoelectric signals before large earthquakes by means a previously published algorithm including an alarm-based model and binary classification [H.-J. Chen, C.-C. Nat. Hazards 84, 877 (2016)]. The original method has been improved removing time parameter used for coarse-graining earthquake occurrences, as well extending single-station into joint-stations method. Analyzing filtered data with different frequency bands, we determine optimal bands earthquake-related featuring highest signal-to-noise ratio. Based on significance tests, also provide evidence relationship between seismicity. suggest using machine learning to extract this underlying relationship, which could be quantify probabilistic forecasts impending get closer operational prediction.