Unraveling Scaling Properties of Slow Slip Events through Kinematic Simulations

A long-standing debate in geophysics concerns the moment–duration scaling of fast and slow earthquakes. While many studies suggest that slow slip events (SSEs) follow a linear moment–duration relationship, in contrast to the cubic scaling typical of regular earthquakes, some observational and modeling studies have reported cubic-like scaling in some SSEs, raising questions about the physical origin of these differences. In this study, we investigate the scaling properties of SSEs using kinematic simulations constrained by empirical source scaling relationships specific to slow slip, combined with stochastic kinematic approaches widely used for regular earthquake rupture modeling. The simulations are conducted on realistic fault geometry in the Cascadia subduction zone and incorporate kinematic constraints characteristic of slow slip. Synthetic SSE scenarios are generated over a wide range of moments and rupture dimensions, allowing systematic exploration of moment–duration behavior without prescribing rupture duration a priori. Our results are consistent with observations and show that the total duration of the simulated SSEs lies near the upper envelope of a cubic scaling trend (M ∝ T³) at smaller moments and gradually transitions toward a linear scaling (M ∝ T) with increasing moment magnitude. When event duration is identified using threshold-based criteria, the resulting moment–duration scaling appears predominantly linear. This finding suggests that biases introduced by observational criteria influence the inferred scaling relationships, providing an explanation for why some studies report linear scaling whereas others report cubic scaling. Furthermore, these results suggest that slow slip events involve rupture processes fundamentally distinct from those of regular earthquakes, as their apparent duration and scaling behavior emerge from kinematic constraints characteristic of slow slip.