Capturing Seismic Signals From Karst Aquifer Injection Experiments and a Natural Recharge Event

Variations in subsurface flow processes through a karst aquifer that feeds Bear Spring in southeastern Minnesota were captured on a temporary seismic network during injection experiments and a natural recharge event. Each experiment involved injecting similar to 13,000 L of water into an overflow spring, and the natural event was triggered by a large rainstorm of similar to 70 min in duration. During the injection experiments, the largest amplitude signals in the ground velocity seismograms occurred as the water first fell onto the rock at the overflow spring and as the large slug of water reached a sump or water-filled passage. During the natural rainstorm event, the overflow spring began flowing and total spring discharge (perennial emanation points and the overflow spring) increased from similar to 100 to 300 L/s. Seismic signals during and following the rain event include broadband noise from raindrops impacting the ground, as well as large amplitude signals while water levels rose; the latter occurred over a 5-s period, producing multiple pulses of ground motion up to similar to 0.5 mm/s. Based on seismic array analysis, high frequency signals during the natural recharge event and one of the injection experiments are largely sourced from south of the array, where a sump exists and the conduit orientation changes, but additional modeling is required to further understand which of a set of possible mechanisms is mostly likely the cause of these seismic signals.

Automated summary

This study analyzed variations in subsurface flow processes through a karst aquifer in southeastern Minnesota by conducting injection experiments and observing a natural recharge event using a temporary seismic network. The experiments involved injecting approximately 13,000 liters of water into an overflow spring, while the natural event was triggered by a 70-minute rainstorm. Seismic signals recorded during these events provided insights into the largest amplitude signals and changes in groundwater levels, but further modeling is necessary to determine the exact cause of these signals.