The evolution of fluid injection-induced seismicity, generally characterized through the number of events or their seismic moment, depends on, among other factors, the injected fluid volume. Migration of seismicity is observed during those sequences and might be caused by a range of mechanisms: fluid pressure diffusion, fluid-induced aseismic slip propagating along a stimulated fault, interactions between earthquakes. Recent theoretical and observational developments underline the important effect on seismicity migration of structural parameters, like fault criticality, or injection parameters, like flow rate or pressurization rate. Here, we analyze two well-studied injection-induced seismic sequences at the Soultz-sous-Fôret and Basel geothermal sites, and find that the evolution of the seismicity front distance primarily depends on the injected fluid volume. Based on a fracture mechanics model, we develop new equations relating seismicity migration to injected fluid volume and frictional and structural properties of the fault. We find that the propagation of a fluid-induced aseismic slip front along the stimulated fault, triggering seismicity, explains well the observations made on the two sequences. This model allows us to constrain parameters describing the seismicity front evolution and explains the diversity of migration patterns observed in injection-induced and natural earthquake swarms.

Anthropogenic fluid injections at depth induce seismicity which is generally organized as swarms, clustered in time and space, with moderate magnitudes. While some similarities between swarms have already been observed, whether they are driven by the same mechanism is still an open question. Pore fluid pressure or aseismic processes are often proposed to explain observations, while recent models suggest that seismicity is triggered by fluid-induced aseismic slip. Using 22 natural and anthropogenic swarms, we observe that duration, migration velocity and total moment scale similarly for all swarms. This confirms a common driving process for natural and induced swarms and highlights the ubiquity of aseismic slip. We propose a method to estimate the seismic-to-total moment ratio, which is then compared to a theoretical estimation that depends on the migration velocity, the effective stress drop and the slip velocity. Our findings lead to a generic explanation of swarms driving process.