Magnitude Clustering During Stick‐Slip Dynamics on Laboratory Faults

We present an analysis of magnitude clustering of microfractures inferred from acoustic emissions (AEs) during stick‐slip (SS) dynamics of faulted Westerly granite samples in frictional sliding experiments, with and without fluids, under triaxial loading with constant displacement rate. We investigate magnitude clustering in time across periods during, preceding and after macroscopic slip events on laboratory faults. Our findings reveal that magnitude clustering exists such that subsequent AEs tend to have more similar magnitudes than expected. Yet, this clustering only exists during macroscopic slip events and is strongest during major slip events in fluid‐saturated and dry samples. We demonstrate that robust magnitude clustering arises from variations in frequency‐magnitude distributions of AE events during macroscopic slip events. These temporal variations indicate a prevalence of larger AE events right after (0.3–3 s) the SS onset. Hence, magnitude clustering is a consequence of non‐stationarities. Plain Language Summary Can we determine the size of a future earthquake based on the size of past earthquakes? This fundamental question has been controversially debated over the years, without an agreed‐upon answer. Here, we tackle this question under controlled conditions in a lab setting by studying the frictional stick‐slip dynamic of rough granite faults, which gives rise to mm‐scale seismic events. We find that the sizes of these seismic events are not independent during periods containing a macroscopic slip but instead are clustered such that larger seismic events tend to directly follow other large seismic events. We show that this can be explained by temporal changes in the frequency of occurrence of seismic events associated with the sliding motion of the fault. Our findings link the properties of mm‐scale seismic events with macroscopic slip on lab faults. Key Points Magnitude clustering of acoustic emissions (AEs) is strongest during major slip events in fluid‐saturated samples of Westerly granite Variations in the frequency‐magnitude distribution of AEs during slip events explain the strong magnitude clustering These variations indicate a prevalence of larger AEs right after the onset of slip events