Although researchers accumulated knowledge about seismogenesis and decadeslong earthquake data, predicting imminent individual earthquakes at a specific time and location remains a long-standing enigma. This study hypothesizes that the observed data conceal the hidden rules which may be unraveled by a novel glass-box (as opposed to black-box) physics rule learner (GPRL) framework. Without any predefined earthquake-related mechanisms or statistical laws, GPRL's two essentials, convolved information index and transparent link function, seek generic expressions of rules directly from data. GPRL's training with 10-years data appears to identify plausible rules, suggesting a combination of the pseudo power and the pseudo vorticity of released energy in the lithosphere. Independent feasibility test supports the promising role of the unraveled rules in predicting earthquakes' magnitudes and their specific locations. The identified rules and GPRL are in their infancy requiring substantial improvement. Still, this study hints at the existence of the data-guided hidden pathway to imminent individual earthquake prediction.
1Analytically or computationally simulated earthquakes are widely used to offer valuable insights into the long-standing enigma of seismogenesis. Researchers seek clues from basic physics -the thermal instability for computational reproductions of deeper slow earthquakes (1), natural fluid injections into fault zones for earthquake swarms (2), or sliding frictional blocks for the chaotic slip pulse behaviors (3). By combining a number of mechanics-/physics-based rules, researchers can reproduce "virtual" earthquakes on computer (4,5). For illustration purposes, this paper calls these methods as "bottom-up" approach since their common starting point is the adopted mechanics-or physics-based rules and the associated parameters.Despite their important roles and values, the bottom-up approaches may explain real earthquake behaviors from a restricted angle, bounded by the intrinsic limits of the adopted rules and experiments used for determining the rules' key parameters. For instance, many studies used the frictional strength of the fault, e.g., rate-and-state friction (6-8), f (ψ, V ) = asinh −1 V2V 0 e ψ/a along with a rule about state evolution (9, 10), G(ψ, V ) = b V 0 dc e (f 0 −ψ)/b − V V 0 , where a is the direct effect parameter, V is the slip velocity, V 0 is the reference velocity, ψ is the state variable, b is the state evolution parameter, d c is the state evolution distance, and f 0 is the reference friction coefficient for steady sliding at V 0 . Parameters (a, b) are useful to simulate the depth of earthquake arrest or nucleation as well as physically sound fault behaviors (e.g., a − b > 0 for the stable sliding, the so-called velocity-strengthening whereas a − b < 0 for unstable sliding, the velocity-weakening). To determine the parameters, researchers often assume a "link" between laboratory tests and real-world earthquakes, e.g., wet granite laboratory tests for deriving (a, b) (6,11,12). Th...