Search for a Possible Triggering Effect Between Tectonic Earthquakes and Swarms at Volcanoes

We searched for possible relationships between earthquake swarms at volcanoes and tectonic earthquakes nearby, using the Global Volcanic Earthquake Swarm Database (GVESD). From the GVESD we extracted a sample of 358 swarms with reported start dates. The durations of these swarms are in the range 15 minutes to 4.6 years and they occurred at 93 volcanoes around the world. Then we used the NEIC earthquake database to search systematically for earthquakes with magnitude >6 and distance <1000 km from the volcano, occurring in the ten years before the swarm start date. The relation between tectonic earthquakes and swarm start dates was analyzed by plotting the number of earthquakes, their magnitudes, and their distances to the volcano versus delta T, the time difference between earthquake occurrence and swarm start date. The time distribution of precursor earthquakes was divided into bins of 6 months and 1 month and all distributions were tested for significance of peaks in the number of earthquakes, magnitudes, or distances. For 50 randomly selected swarms with durations less than 120 days we found only one peak in the number of precursor earthquakes that appears to be statistically significant. This peak occurs in the 4th month bin before the swarm start dates. We also tested for smaller time scales and for earthquakes after the swarm start dates. We constructed plots of the number of earthquakes, magnitude, and distance to the volcano dividing the time axis in bins of 1 day and extending the search range up to one month after the swarm start dates. A peak in the number of precursor earthquakes occurs at the 15th day bin before the swarms start, but the peak is not significant as estimated using the t-test. The results are somewhat surprising considering that the group of precursor earthquakes includes events that are large (M=7.9), nearby to the volcanoes, and relatively close in time to the swarm start dates. Recent well-documented cases for triggering of seismicity at volcanoes in Alaska indicate that the triggered swarms are small and short (e.g. 17 events in 13 hours at Katmai). This suggests that most swarms in the GVESD may be too large or long lasting. Those likely to be triggered may be too small to be reported. Future studies would benefit from knowledge of the initial conditions at the volcanoes. This would help to determine whether there is some magnitude-distance threshold that the tectonic earthquake needs to overcome to trigger a volcanic swarm.