According to the USGS the water-level fluctuates in wells in response to seismic waves. The occurrence and size of the water level response of water wells to earthquakes are influenced by a variety of factors such as the magnitude and depth of the earthquake, the distance of the well from the epicenter, the geology surrounding the well, the depth of the well, whether the aquifer is confined or unconfined and the well construction. The most common observed groundwater response to an earthquake is an instantaneous water level offset or step up. An instantaneous increase or decrease in water level. This response is commonly observed because there are hundreds of wells used to monitory water levels nationally. The change and recovery in water level can be so rapid that it is barely detectable, or it may take minutes, hours, days, or months for a well to return to previous water levels. There have also been instances where the well never returned to pre-earthquake levels.
Lots of things can impact the response of a well to a seismic event and scientists cannot predict which wells will be impacted and whether the impact is permanent or transitory. The USGS reports that within 3 months of the 1998 magnitude 5.2 earthquake in northwestern Pennsylvania that over a hundred private water wells in the area went permanently dry. The 2002 Alaskan Denali Fault earthquake which was a magnitude 7.9 caused a 2-foot water-level rise in a well in Wisconsin, more than a thousand miles from the epicenter. That rise also appears to be permanent.
The USGS says that water level offsets in the area of an earthquake because the earthquake “subjects the earth’s crust and its aquifer systems, to stress and permanent strain (deformation). This deformation process results in altered fluid pressure within the aquifer systems, and consequently, a step like change in water level would be expected.” The USGS cites various mechanisms for well water responses based on type of geology. For increased water level in shallow wells, the USGS suggests that compaction of overlying alluvium similar to liquefaction may be the mechanism producing the offset. Fluid-pressure declines are suggested to be caused by the escape of small amounts of dissolved gas from pore spaces in the aquifer in response to seismic waves. In a fractured rock system like the one here and in northwestern Pennsylvania where the wells went dry, permeability of the ground may be changed by the unclogging, widening, or narrowing of fractures, or the creation of new fractures. Similarly, an increase in ground-water discharge though springs, seeps, or to streams could be caused by an increase in the subsurface fluid pressure or permeability of the geologic formation.
There have also been several instances of reported changes in water levels in well before an earthquake. Many of these documented cases come from seismically active Japan where they continue to search for predictors of earthquakes. Many scientists do not believe that groundwater and well response can be used to predict earthquakes; they believe that there are too many other explanations for well water changes. Truthfully in the last couple of years (in my volunteer work with the VAMWON) I’ve seen several instance of what I think of as transitory benign well response where a well level falls dramatically and then recovers seemingly unconnected to precipitation and use. I have also seen transitory turbidity that just seems to pass through and then the well returns to normal. There are many potential explanations for these observations, so I simply note them and test the water quality when it returns to normal to make sure it is safe to drink.
However these episodes make you think about how interconnected the groundwater system is to the earth. Recently Dr. Robert Jackson et al. published an analysis of all the peer reviewed research done on hydraulic fracking. Though there has been documented impact to groundwater from improperly constructed wells, there has been no confirmed impact to groundwater from fracking itself. Rather than testing for direct chemical contamination in areas surrounding a hydraulic fracking maybe scientist should be looking for changes in water quality and quantity in nearby groundwater wells and systems that are more typical in seismic events.
For more information on stream flow and water well responses to earthquakes and discussion of potential mechanisms see research from Department of Earth and Planetary Science, University of California, at Berkeley by Chi-Yuen Wang, Michael Manga and others.