Alan L. Kafka
Weston Observatory
Department of Geology and Geophysics
Boston College

Yes, the Earth Does Quake in New England!

In the evening of October 2, 2006 residents of coastal Maine felt the ground shake from a magnitude 4.2 earthquake located near Bar Harbor. In nearby Acadia National Park, falling rocks forced the closure of the Park Loop Road.

On December 20, 1940, residents of Springfield, MA felt vibrations from an earthquake centered near Ossipee, NH. The Springfield Republican newspaper reported that "dishes toppled from the shelves in many kitchens and wall pictures swung pendulum fashion on their fastenings."

On November 20, 1755, the Boston Weekly News-Letter reported "a most terrible Shock of an Earthquake: The conditions were so extreme as to wreck the Houses in this Town to such a degree that the Tops of many Chimnies, and some of them quite down to the Roofs, were thrown down. . . ." This earthquake was centered off the coast of Cape Ann, MA.

Yes, it's true. Earthquakes do occur on a regular basis in New England. Notable examples of earthquakes that caused damage in New England and adjacent areas are: the earthquake off the coast of Cape Ann, MA in 1755; two earthquakes near Ossipee, NH in 1940; and an earthquake near New York City in 1884. In spite of this regular occurrence of earthquakes in the northeastern United States, most people living in New England probably think of places like California or Japan when they hear the word "earthquake." While Californians have learned to expect earthquakes, residents of New England typically consider the ground beneath their feet to be "solid as a rock." Nonetheless, the record of earthquake activity in the United States shows that, while the highest level of activity is, of course, in the western part of the country, earthquakes are quite common in many areas of the eastern United States, including New England.

Why then do earthquakes occur in New England? What causes earthquakes in major seismic zones like California and Japan? Why are earthquakes more common in California than in the northeastern United States, and how does earthquake activity in California differ from that in the Northeast?

Plate Tectonics and California Earthquakes

Unlike the situation in the northeastern United States, most California earthquakes occur near a major boundary separating two of the Earth's tectonic "plates", the North American plate and the Pacific plate. In the late 1960's, a revolution occurred in the geosciences when scientists discovered that earthquakes, volcanoes and geological activity are concentrated in well-defined zones that outline the edges of plates. There are about a dozen such plates. This theory of the Earth, known as plate tectonics, is now accepted as the overall explanation of how the Earth works. Plate interiors are much more stable than plate boundaries, and the theory of plate tectonics accounts for most of the large earthquakes in California, Japan, and Alaska, to name a few of the infamous seismic zones.

Of course, the real world is more complicated than this model suggests. While the two largest historically-documented earthquakes to occur in California (1906, San Francisco and 1857, Fort Tejon) did rupture a plate boundary along the well-known San Andreas fault, when looked at in detail, earthquakes in California are rather diffusely distributed in the vicinity of the plate boundary. Furthermore, earthquakes are scattered throughout the interior of the North American plate (as well as throughout the interiors of all plates).

With this perspective on plate boundary earthquakes, let us look at earthquakes in the northeastern United States.

Earthquakes in the Northeastern United States

The historical record of earthquakes in the northeastern United States and adjacent areas goes back to the 1500's, and a number of seismographs were operating in this region beginning in the early 1900's. Routine reporting of instrumental data on earthquakes in this region began in the late 1930's. The number of seismic stations in the Northeast increased significantly between 1970 and 1974. By 1974, area seismologists were operating a regional seismic network in the northeastern United States. The figure shown below illustrates the pattern of earthquake activity in New England and  surrounding areas.

Earthquakes in New England and surrounding areas. The data for 1924-1974 is from the U.S. Geological Survey, and the data for 1975-2006 is from the Weston Observatory archives of earthquakes recorded by the Northeastern United States Seismic Network.

There is a general correlation between the distribution of earthquake epicenters determined from the 1975-2006 network data and that of the epicenters of earthquakes that occurred between 1924 and 1974. In both the 1924-1974 and the 1975-2006 records, clusters of seismic activity occur in northern New York State, and extend from there along an approximately northwestern trend into Canada. During both periods of time there was earthquake activity in New Jersey, eastern Pennsylvania, and the greater New York City area, as well as along coastal New England and in most of Maine and New Hampshire. Also, in both maps, we see intense activity in the vicinity of the St. Lawrence River. In addition to these clusters of activity, diffuse seismicity has occurred throughout most of the region during both periods of time. In spite of these similarities between the 1924-1974 seismicity and that observed between 1975 and 2006, during the past few decades some relatively large earthquakes occurred in areas that were not particularly active in the earlier part of the 20th century. Also, the activity seen in Vermont during the earlier period is not observed in the more recent record.

Earthquakes in the Interior of the North American Plate

One of the largest historically-documented earthquakes in the United States occurred in New Madrid, Missouri, deep in the interior of the North American plate. This earthquake--actually a sequence of three major shocks in late 1811 and early 1812--is a dramatic example of the fact that quite large earthquakes occur in plate interiors. These earthquakes are called intraplate earthquakes. Intraplate earthquakes are common occurrences, globally as well as specifically in the northeastern United States.

The occurrence of earthquakes in the northeastern United States apparently violates the plate tectonic model. The past several decades of research on this topic, however, have demonstrated that it may be possible to explain the occurrence of earthquakes in the Northeast within the framework of plate tectonics. The challenge in figuring out why the Earth quakes in New England is that the earthquake process in plate interiors is more complex than at plate boundaries. Unlike the situation in California, there is no obvious relationship between earthquakes and geologically mapped faults in most intraplate areas.

With plate tectonics under their belt and a seemingly endless supply of research funds flowing in, a number of geologists and geophysicists in the mid-1970's turned their attention to a (presumably) straightforward problem: What causes intraplate earthquakes? These were heady times for seismologists, and some were quite confident that the problem would be easily resolved. Hypotheses abounded. With a minimum of data available, it sometimes seemed like there was more fanciful hypothesis-generating than hypothesis testing. Two well-known hypotheses to explain earthquakes in the Northeast were the supposed existence of a "Boston-Ottawa seismic zone" and the presumed activity of the Ramapo fault (an ancient geological fault in northern New Jersey and southeastern New York State). In spite of many such attempts to explain the cause of earthquakes in the Northeast, however, none of these specific hypotheses have fared very well in light of new data gathered in the past few decades.

At the present time, a commonly accepted explanation for the cause of earthquakes in the Northeast is that "ancient zones of weakness" are being reactivated in the present-day stress field. In this model, preexisting faults and/or other geological features formed during ancient geological episodes persist in the intraplate crust, and, by way of analogy with plate boundary seismicity, earthquakes occur when the present-day stress is released along these zones of weakness. Using this model as a guide, much of the research on northeastern United States earthquakes has involved attempts to identify preexisting faults and other geological features that might be reactivated by the present-day stress field. While this concept of reactivation of old zones of weakness is commonly assumed to be valid, in reality the identification of individual active geologic features has proven to be quite difficult. Unlike the situation for many plate boundary earthquakes, it is not at all clear whether faults mapped at the earth's surface in the Northeast are the same faults along which the earthquakes are occurring.

Earthquakes and Ancient Plate Boundaries in New England

How well then does this pre-existing zones of weakness model work in the northeastern United States? Like all other parts of the Earth, the bedrock of New England and adjacent areas has been affected by many episodes of geological activity, each lasting many millions of years. Geological features (including ancient faults) that are "scars" of these geological episodes are found throughout New England and adjacent areas; nonetheless, it has been hard to find any unequivocal relationship between these geological features and seismicity in this area.

During the past half billion years, the Earth's crust underlying New England has been the site of two major geological episodes, each of which has left its imprint on the New England bedrock. Between about 450 million years ago and about 250 million years ago, this area was the site of a "continental collision," in which the ancient African continent collided with the ancient North American continent to form the supercontinent known as Pangaea. Beginning about 200 million years ago, the present-day Atlantic ocean began to form as plate tectonic forces rifted the continent of Pangaea. The last major episode of geological activity to affect the bedrock in New England occurred about 100 million years ago, during the Mesozoic era, when "continental rifting" led to the opening of the present-day Atlantic ocean.

Earthquakes and Ancient Continental Rifts: A Tale of Two Basins

Since this Mesozoic rifting episode was the most recent major tectonic event in the northeastern United States, one might expect that the geological features formed during the Mesozoic era would be likely places for present day earthquakes to occur. Also, research on intraplate earthquakes on a global scale suggests that anient continental rift zones are likely locations for major intraplate earthquakes. Thus, one example of proposed hypotheses to explain the connection between geological features and the occurrence of earthquakes in this region is that the current seismicity is correlated with "Mesozoic rift basins."

The remains of the Mesozoic rifting episode in the Northeast are found in a number of ancient continental rift zones, most notably the Hartford rift basin in central Connecticut and central Massachusetts, and the Newark rift basin in the greater New York City area. In addition, Mesozoic rifts are believed to be buried beneath sediments along continental margins of the eastern United States. In my own research, I have  found that there does appear to be a concentration of earthquakes in the area surrounding the Newark basin, which would be expected if Mesozoic basins are the active features in this intraplate environment. It is less obvious, however, whether such a concentration of earthquakes exists in the area surrounding the Hartford basin. Furthermore, it is very difficult to determine whether the observed concentation of earthquakes around the Newark basin is a long-term (i.e., geological) phenomenon or just an artifact of the particular period of time during which we have been observing the historical and insturmentally recorded seismicity.

Such an ambiguous relationship between geological features and earthquakes is typical of seismological studies in intraplate areas, which makes it very difficult to determine which (if any?) parts of New England are more prone to earthquakes than others.

Can New England Earthquakes be Predicted?

In spite of all these complications and enigmas -- we are not even sure where the earthquakes are most likely to occur, let alone when they are likely to occur -- New England seismologists are often asked if it is possible to predict earthquakes in this area. Not only is it impossible to predict earthquakes in New England, but even at plate boundaries routine prediction of earthquakes remains an elusive goal. Seismologists make the distinction between specific predictions of earthquakes and long-term forecasts of earthquakes. In the late 1970's, seismologists were quite optimistic about eventually being able to forecast, and perhaps even predict, earthquakes. Today, in light of new data gathered in recent decades, predicting earthquakes seems to be far more complex than it was once thought to be.

At least for the time being, the future time and place of any particular earthquake is hard to pin down. We can of course say with certainty that future earthquakes will occur in New England. But as for when and where . . . the scientific study of these New England earthquakes has a long way to go.

Evaluating the Earthquake Hazard

"We muddle through life making choices based on incomplete information..."
- Larry Gonick and Woollcott Smith, The Cartoon Guide to Statistics
Life is uncertain. No matter what we do, there will always be risks that we accept as part of our daily lives. No matter how advanced our scientific knowledge of the Earth becomes, it is unlikely that there will ever come a time when all natural disasters are predictable. Nonetheless, although we can't predict when specific earthquakes will strike, our knowledge of the earthquake process in New England does make it possible to estimate the odds of a damaging earthquake vibrations striking your town.

Recognizing that there are many unknowns and uncertainties regarding the cause of earthquakes and where they are likely to strike, the U.S. Geological Survey (USGS) has produced a series of National Seismic Hazard Maps based on the current state of knowledge regarding earthquake activity across the United States. These maps were developed to provide the most accurate information possible regarding ground shaking hazard, and are updated as the knowledge base increases and new scientific findings on earthquake processes become available. The maps show the amount of earthquake generated ground shaking that, over a specified period of time, is predicted to have a specified chance of being exceeded. Ground shaking caused by earthquakes is often expressed as a percentage of the force of gravity. For example, the earthquake hazard map shown below gives contours of the percentage of the force of gravity that has a 2% chance of being exceeded in 50 years. Based on such a map, if you were living in the same house for 50 years, and that house was in a zone labeled "0.10 g", then there would be 1 chance in 50 (i.e., 2% chance) that (at some point during those 50 years) an earthquake would shake your house at a level of at least 10% of the force of gravity.

A recent version of the National Seismic Hazard Maps developed by the USGS is shown below. Because of the difficulty of identifying specific seismically active geological faults in the Northeast, the level of seismic hazard in the northeastern part of the USGS map is based primarily on the past record of seismic activity. Earlier versions of the National Seismic Hazard Maps placed more emphasis on the locations of geologically mapped faults (and other types of geological features) in the Northeast.

National Seismic Hazard Map developed by the U.S. Geological Survey in 2008. This map shows contours of the percentage of the force of gravity that is predicted to have a 2% chance of being exceeded in 50 years.
To appreciate why these values of ground shaking are expressed as a percentage of the force of gravity, note that it requires more than 100% of the force of gravity to throw objects up in the air. In terms of felt effects and damage, ground motion at the level of several percent of gravity corresponds to the approximate threshold of damage to buildings and houses (an earthquake intensity of approximately V). For comparison, reports of "dishes, windows and doors disturbed" corresponds to an intensity of about IV, or about 2% of gravity. Reports of "some chimneys broken" corresponds to an intensity of about VII, or about 20% to 30% of gravity.

Although seismologists have not yet found an unequivocal answer to the question, "Why Does the Earth Quake in New England?", earthquake hazard maps generally show that in most parts of New England, there is about a 2% chance, that in any given 50 year period of time, earthquake vibrations that are potentially damaging will strike. This type of information can be used as input for the design and construction of buildings and for public policy decisions requiring an assessment of earthquake risk. The better we can understand the cause of earthquakes in the northeastern United States and the nature of ground motion generated by those earthquakes, the better we will be able to provide the necessary information to estimate (and plan for) the earthquake risk in New England.

Bibliography of Research Articles and Books for Further Reading

Bolt, B.A., Earthquakes and Geological Discovery, Scientific American Library, 1993.

Ebel, J.E. and A.L. Kafka, Earthquake Activity in the Northeastern United States, in Neotectonics of North America, D.B. Slemmons, E.R. Engdahl, M.D. Zoback, and D.D. Blackwell, eds., Geological Society of America, 1991.

Johnston, A.C. and L.R. Kanter, Earthquakes in Stable Continental Crust, Scientific American, March 1990.

Kafka, A.L., Does Seismicity Delineate Zones Where Future Large Earthquakes will Occur in Intraplate Environments?, In Continental Intraplate Earthquakes: Science, Hazard, and Policy Issues, Geological Society of America Special Paper 425, Edited by S. Stein and S. Mazzotti, 2007.

Kafka, A.L., Public Misconceptions About Faults and Earthquakes in the Eastern United States: Is it our own fault?, Seismological Research Letters, 71(3), 311-312, 2000.

Kafka, A.L. and Walcott, J.R., How Well Does the Spatial Distribution of Smaller Earthquakes Forecast the Locations of Larger Earthquakes in the Northeastern United States?, Seismological Research Letters, 69(5), 428-439, 1998.

Kafka, A.L. and P.E. Miller, Seismicity in the Area Surrounding two Mesozoic Rift Basins in the Northeastern United States, Seismological Research Letters, 67(3), 69-86, 1996.

Kafka, A.L., M.A. Winslow and N.L. Barstow, Earthquake Activity in the Greater New York City Area: A Faultfinder's Guide, in Field Trip Guidebook, 61st Annual Meeting, N.Y. State Geological Association, D. Weiss, ed., p. 177-205, 1989.

Kafka, A.L., E.A. Schlesinger-Miller and N.L. Barstow, Earthquake Activity in the Greater New York City Area: Magnitudes, Seismicity and Geologic Structures, Bulletin of the Seismological Society of America, 75(5), 1285-1300, 1985.

Raymo, C. and M.E. Raymo, Written in Stone: A Geological and Natural History of the Northeastern United States, The Globe Pequot Press, 1989.

Reiter, L., Earthquake Hazard Analysis: Issues and Insights, Columbia University Press, 1994.

Seih, K. and S. LeVay, The Earth in Turmoil: Earthquakes, Volcanoes, and Their Impact on Humankind, W.H. Freeman and Company, 1998.

Revised: September 16, 2008