Facts about the New Madrid Seismic Zone
The New Madrid Seismic Zone (NMSZ) is the most active seismic area in the United States east of the Rocky Mountains. The NMSZ is located in southeastern Missouri, northeastern Arkansas, western Tennessee, western Kentucky and southern Illinois. Southwestern Indiana and northwestern Mississippi are also close enough to receive significant shaking from large earthquakes occurring in the NMSZ.
The active faults in the NMSZ are poorly understood because they are not expressed at the ground surface where they can be easily studied. The faults are hidden beneath 100- to 200-foot thick layers of soft river deposited soils called alluvium. Fault scarps and traces in the soft alluvium erode in a very short time or may be rapidly covered by new deposits thereby quickly hiding evidence of earthquake fault lines. Faults in places like California, where rocks are at or near the ground surface, are much easier to study because the faults are readily found, seen, measured and analyzed.
Location of earthquake epicenters in and near the New Madrid Seismic Zone (circles scaled according to magnitude.)
Knowledge about some of the NMSZ faults is obtained by seismograph recordings of the frequent small earthquakes. St. Louis University, University of Memphis, the U.S. Geological Survey and University of Kentucky operate more than 30 seismograph stations to monitor earthquake activity in the NMSZ and Central U.S.
Microseismic earthquakes (magnitude less than 1.0 to about 2.0), measured by seismographs but not felt by humans, occur on average every other day in the NMSZ (more than 200 per year).
The measured locations of microseismic earthquakes show some trends that have been used to identify active faults in the NMSZ. The trends indicate a four-segment, zig-zag fault system with a total length of about 125 miles stretching from Marked Tree, Arkansas northeastward through Missouri, Tennessee and Kentucky to Cairo, Illinois.
Active faults that have generated dangerous earthquakes in historic times or the recent geologic past (the last 10,000 years) are not always microseismically active. In fact, in some settings these quiet faults are considered the most dangerous ones because high built up stress has locked the two sides of the fault together thereby preventing the microseismic earthquakes. This is thought to happen as a prelude to a major rupture of the fault. It is not known if faults of this type exist in the NMSZ. If they do exist there is no easy way to locate them.
A series of three to five major earthquakes (believed to have been magnitude 7.0 or larger earthquakes) occurred in the NMSZ in the two month period between Dec. 16, 1811 and Feb. 7, 1812. Several thousand additional “smaller” earthquakes occurred during the three month period from Dec. 16, 1811 to March 16, 1812. These included 15 earthquakes believed to have been magnitude 6.5 to 8.0 (the size range of the 1989 San Francisco, 1994 Los Angeles and 1995 Kobe, Japan earthquakes) and 189 earthquakes magnitude 5.0 to 6.5. Two thousand were felt by people, indicated by crude seismograph instruments and recorded in personal journals at Louisville, Kentucky and Cincinnati, Ohio, which are respectively 250 and 350 miles away. The New Madrid area was at the very frontier of European settlement at this time so there were very few non-native Americans living in the immediate area to report the earthquakes.
Due to the harder, colder, drier and less fractured nature of the rocks in the earth’s crust in the central United States, earthquakes in this region shake and damage an area approximately 20 times larger than earthquakes in California and most other active seismic areas. Even though large earthquakes occur much less frequently in the NMSZ than in California, the long term average quake threat, in terms of square miles affected per century, is about the same because of the approximately 20 times larger area affected in the central United States.
The frequency of large earthquakes in the NMSZ is still being debated. Several methods have been used to make estimates. Paleoseismology techniques are used to recognize evidence of large prehistoric earthquakes preserved in geologic materials (soil and rock). From the approximate locations, dates and magnitudes the long term average recurrence interval can be calculated. Currently, paleoseismologists infer two or more large earthquakes (magnitude 7 or larger) have occurred in the last 2,000 years or less giving recurrence interval estimates of 300 to 1,000 years for the large quakes. Probability models extrapolate the 200 years (approx.) of recorded history or 100 years (approx.) of instrumental recordings to estimate frequency. Probability estimates are given in the following table.
The NMSZ appears to be about 30 years overdue for a magnitude 6.3 quake because the last quake of this size occurred 100 hundred years ago at Charleston, Missouri, on Oct. 31, 1895 (it was a magnitude 6.7). A magnitude 6.3 quake near Lepanto, Arkansas, on Jan. 5, 1843, was the next prior earthquake of this magnitude. About 75 percent of the estimated recurrence time for a magnitude 7.6 earthquake has elapsed since the last quake of this size occurred in 1812.
The probability of a repeat of the 1811-12 (magnitude 7.5-8.0) earthquake is 7-10%
A magnitude 7.6 earthquake in the NMSZ will cause major damage near the fault system in the Missouri Bootheel, northeast Arkansas and western Kentucky and Tennessee. Significant damage is expected to extend north of St. Louis up the Mississippi River valley, up the Ohio and Wabash River valleys to near Owensboro, Kentucky and Indianapolis, Indiana and down the Mississippi River valley to near Greenville, Mississippi. Significant damage is also expected in about 15 additional counties each in southern Illinois, western Kentucky and Tennessee, northeastern Arkansas and northwestern Mississippi and in about five counties in southeast Missouri outside the Bootheel.
The 1811-1812 New Madrid earthquakes destroyed most or all of the simple buildings in New Madrid and Little Prairie (Caruthersville), the only two towns in the area at the time. The simple, single story log buildings were in fact a very earthquake resistant type of construction. The land in a huge area (about 5,500 square miles or about 3.5 million acres) was also damaged or destroyed by landslides, fissures, sandblows, lateral spreads, subsidence, submergence and uplift. Much of this land became unusable for the subsistence type agriculture of that day. Buildings in St. Louis, 150 miles or more to the north, were slightly damaged. The damage mostly consisted of broken or collapsed stone chimneys and broken stone buildings, a type of unreinforced masonry construction that is especially susceptible to earthquake damage.
The primary cause of damage and injuries in earthquakes is the destruction of manmade structures and their contents. Modern society is highly dependent on our manmade structures and infrastructure. Tall structures (buildings, bridges, dams, etc.), structures with large open spaces (auditoriums, classrooms, factories, hospitals, churches etc.), structures made of brittle materials (brick and unreinforced concrete), structures with complex shapes and lots of corners and unanchored building contents (bookshelves, file cabinets, storage racks, pipes, fire sprinkler systems, water heaters, mechanical equipment, etc.) are particularly susceptible to damage from shaking.
Damage from earthquakes in the NMSZ will vary depending on the earthquake magnitude, the character of the land and the degree of urbanization. The Bootheel area is dominantly rural with scattered, small to medium size towns. Damage to the land will be extensive and will have a big impact in this area because it will disrupt farming. Thick unfertile sand will cover significant areas of fields, fields leveled for irrigation will no longer be level, drainage ditches will be blocked, grain elevators will be damaged and transport of materials, supplies and crops will be difficult.
The Bootheel towns do not have many tall structures to be damaged by shaking but they do have some large open structures and many unreinforced masonry structures which are likely to be damaged. The more distant, densely populated urban area of St. Louis is not likely to have damage to the land but its huge stock of structures and their contents are likely to receive significant damage from shaking.
A 1990 Federal Emergency Management Agency (FEMA) report estimates damage and injuries in St. Louis from a 7.6 magnitude NMSZ earthquake to be $2.8 billion, 260 deaths and 1,060 serious injuries. A 1985 FEMA report makes estimates for Poplar Bluff for the same earthquake of $693 million in damages, 1 death if the quake occurs at night or 17 deaths if it occurs during the day (the difference being 15 deaths at schools during the day when schools are in session).
A 1991 Missouri State Emergency Management Agency (SEMA) and FEMA report estimates for a 7.6 magnitude NMSZ earthquake the Bootheel counties of Mississippi, New Madrid and Pemiscot would have 0.2 percent to 2 percent of their population killed, 1 percent to 10 percent of their population seriously injured, 10 percent of their buildings collapsed and 30 percent of their buildings receiving severe structural damage. For Scott, Stoddard and Dunklin counties the percentage estimates are 0.1 percent to 1 percent killed, 0.5 percent to 5 percent seriously injured, 5 percent of structures collapsed and 20 percent with severe structural damage. For Butler, Bollinger, Cape Girardeau, Perry, St. Louis (including the City), Lincoln, Pike, Marion and Lewis Counties the estimates are 0.02 percent to 0.2 percent killed, 0.1 percent to 1 percent seriously injured, 1 percent building collapses and 10 percent with severe structural damage.