Northeastern Section - 42nd Annual Meeting (12–14 March 2007)

Paper No. 6
Presentation Time: 1:00 PM-4:45 PM

A LINK BETWEEN SEASONAL GROUNDWATER LEVELS AND EARTHQUAKES IN THE NORTHEASTERN UNITED STATES?


BUCK, Nicole L., DADE, William Brian and POSMENTIER, Eric S., Earth Sciences, Dartmouth College, 6105 Fairchild, Hanover, NH 03755, nicole.l.buck@dartmouth.edu

An analysis of shallow seismic activity in the northeastern United States during the past three decades reveals that earthquakes exhibit near semi-annual periodicity to a statistically significant degree. Specifically, we examined records of earthquakes for which M > 2 and with hypocenters at crustal depths of 5 to 20 km, and found that their monthly rates of occurrence during March and October are 28 percent and 43 percent greater, respectively, than the historical average monthly rate of occurrence. We hypothesize that this apparent periodicity is related to seasonal fluctuations in the degree of regional, shallow crustal saturation with groundwater or, equivalently for brevity, hydrologic loading. As a proxy for hydrologic loading, we evaluated the mean depth to groundwater in 17 wells, distributed more or less uniformly throughout the region. Using this proxy, we tested the statistical significance of differences in loading on days with an earthquake vs. days without, and found the degree of loading to be significantly greater (p < 0.015) on days with an earthquake than on days without. However, the difference in loading, while statistically significant owing to large sample size, was small compared to the standard deviation of well depths to groundwater. This observation indicates that hydrologic loading alone, as defined by the proxy used here, is not a useful predictor of earthquakes. Upon further analysis, we found that the rate of change of regional hydrologic loading was not significantly different (p = 0.10) between days with and without earthquakes. These exploratory results suggest that hydrologic loading may contribute to subtle crustal deformation and thus influence triggering of some earthquakes, but that the statistically significant recurrence of earthquakes in March and October remains an intriguing and largely unexplained phenomenon.