1. First Main Shock

The first in a series of major earthquakes occurred on December 16, 1811, at about 2:15 a.m. near present-day Blytheville, Arkansas. Estimated at magnitude (M) 7.7, this event awakened people across much of what is now the eastern United States, up to about 900 miles away, including President James Madison in Washington, D.C. There were several large aftershocks that day, including the largest “Dawn” aftershock at about 7:15 a.m., estimated at M6.8–7.0.

2. First Steamboat Voyage Interrupted

The maiden voyage of the first steamboat on the Mississippi River, the New Orleans, began in September 1811 on the Ohio River at Pittsburgh, Pennsylvania. It reached the Mississippi River the week before the December 16 earthquake. One morning the passengers woke to find that the island they had moored to the night before had disappeared because of liquefaction due to frequent earthquake shaking.

3. Riverbanks Caved, Widespread Flooding

According to reports of boat captains and others on the Mississippi River in December 1811, the earthquakes caused the river to fill with debris as the tree-covered river banks caved into the river. Other reports said the river flooded large tracks of land, created temporary waterfalls, and even ran backward in some places. Today, similar bank failure, flooding, and river debris would affect bridge abutments, farmland and industries along the river, and barge traffic.

4. Second Main shock Earthquake

The second of the three major earthquakes, estimated at M7.5, occurred on January 23, 1812, at about 9:15 a.m. Similar to the December 1811 earthquake, this event was widely felt throughout the region. Damage occurred in an area of approximately 232,000 square miles.

5. Only 5,700 People Lived in St. Louis in 1811

In 1811, the population of what is now the central United States was very sparse. Still considered to be the western frontier, there were only about 5,700 people in the St. Louis area at the time. Most historical reports (journals, letters, and newspapers) describing the earthquake shaking and its effects were written by people located east of the Mississippi River. Today, about 11–12 million people live in St. Louis–Memphis region.

6. Landslides

The New Madrid earthquakes caused extensive landsliding along about 125 miles of the bluffs on the east side of the Mississippi River north of Memphis. A repeat of these earthquake-triggered landslides today would affect eastern access to the Mississippi River, pipelines, and transmission towers in this area.

7. Third Main shock Earthquake

The third and probably the most widely felt of the three major New Madrid earthquakes occurred on February 7, 1812, at about 3:00 a.m. That day, there were several destructive shocks, the last and largest estimated at M7.7. As a result, the town of New Madrid, Missouri, was severely damaged.

8. Homemade Seismograph Records Quakes in Cincinnati

Dr. Daniel Drake, a physician in Cincinnati, Ohio, kept a detailed account of the New Madrid earthquake sequence and cataloged thousands of earthquakes. He reportedly used a pendulum-type seismograph that “marked the direction of undulation from south-southwest to north-northeast.”

9. Widespread Ground Failure and Sand Blows

Important surface-deformation features caused by the 1811–1812 earthquakes are the Reelfoot fault scarp (an uplift of the land surface above the fault); Reelfoot Lake in western Tennessee; a large sand-blow liquefaction field in Arkansas, Missouri, and Tennessee; and the Big Lake and St. Francis sunken lands in northeastern Arkansas. The New Madrid earthquakes covered the region with sand blows, thousands of which remain today and can be seen as large, light-colored sandy patches in agricultural fields.

10. Reelfoot Lake Forms

As a result of land subsidence and uplift from the February 7, 1812, earthquake, water from the Mississippi River flooded the area and formed Reelfoot Lake. This event added to reports of the Mississippi River flowing backward.

11. Thousands of Aftershocks

By March 15, 1812, an estimated 2,000 aftershocks had been felt, but it is likely that tens of thousands more were not felt or otherwise noted. At the time, many people were afraid to enter their homes for fear they would collapse. Damage from the largest of these shocks was reported from about 300 miles away. Today, a robust aftershock sequence like this would impede rescue and rebuilding efforts throughout the region.

12. It Had Happened Before 1811

Research indicates that the 1811–1812 New Madrid earthquake sequence was preceded by at least two similar earthquake sequences in about 1450 AD and 900 AD and that over the past 5,000 years or so, other large earthquakes have also occurred in the region surrounding the main area of the currently active New Madrid seismic zone.

13. Archaeological Age Constraints on Earthquake Timing

Geologists and archaeologists use Native American artifacts buried in the upper Mississippi River Valley near earthquake deformation features, such as liquefaction sand blows, to help determine when large prehistoric earthquakes occurred.

14. First Disaster Relief Act

In 1814, Missouri Territorial Governor William Clark (of the Lewis & Clark Expedition) asked for Federal relief for the “inhabitants of New Madrid County.” In response, Congress allocated $50,000 for recovery in 1815. This was the first disaster relief act in the United States. Today, a repeat of the New Madrid earthquake sequence would cause billions of dollars in damage, especially to older unreinforced brick buildings.

15. More Quakes – M6.0 in 1843, M 6.6 in 1895

The strongest earthquakes to occur in the NMSZ since the 1811–1812 sequence happened on January 4, 1843, and October 31, 1895. The 1843 earthquake, estimated at M6.0 and located near Marked Tree, Arkansas, caused damage in Memphis. The 1895 earthquake, located near Charleston, Missouri, estimated at M6.3−6.6, caused extensive damage to chimneys, walls, and windows in homes, churches, and schools in the area; and caused damage to almost all the commercial buildings in downtown Charleston.

16. 1909 – Seismograph Installed at Saint Louis University

The first seismograph in the region (called a Wiechert) was installed at Saint Louis University in 1909. This instrument recorded its first earthquake on October 9 of that year.

17. Growth of Seismograph Network

In the 1930s, a seismograph was installed in Cape Girardeau, Missouri. It is the oldest continuously recording seismic station specifically installed to study the NMSZ. In 1974, installation began on a dense network of 50 seismic stations in and around the NMSZ. Today there are about 270 seismograph stations in the New Madrid cooperative network in the area between St. Louis and Memphis.

18. Another Quake – M 5.0 in 1976

An M5.0 earthquake occurred in Poinsett County, Arkansas, at the southern end of the NMSZ, on March 24, 1976. It was felt over 174,000 square miles and caused power outages, downed telephone lines, broken windows, and roof damage.

19. Will a Damaging Earthquake Occur Again?

There is a 7–10% chance or probability that a quake of magnitude 7.0 or greater will occur in the New Madrid region within the next 50 years. The probability of an earthquake of magnitude 6.0 or greater occurring during the same time period is somewhere between 28 and 46%.

20. Most Seismically Active Area East of the Rockies

The NMSZ is the most seismically active area of the United States east of the Rocky Mountains, according to information from the last 40 years of instrumental recordings. It experiences an average of 200 earthquakes (greater than M1.5) every year—most, however, are too small to be felt.

Disclaimer: The suggestions and illustrations included in this website are intended to improve earthquake awareness and preparedness; however, they do not guarantee the safety of an individual or a structure. The writers, contributors, and sponsors do not assume liability for any injury, death, property damage, or other effect of an earthquake.

spacer

spacer

The 2023 National Seismic Hazard Model – What’s Shaking? ACTIVE

By Earthquake Hazards Program
June 15, 2022

• Overview

No one can predict earthquakes. But existing faults and past earthquakes give us information about future earthquakes, and geology tells us how the ground shakes during an earthquake.

Faults, earthquakes and geology – are the ingredients that go into making a seismic hazard model. Periodically since 1976 the U.S. Geological Survey (USGS) uses the most current data and state-of-the-art forecasting methods to create an updated National Seismic Hazard Model (NSHM). Additionally, engineering analyses inform us about how built structures could react to the ground shaking forecasted by the model. The model and related maps have a surprising variety of uses, from guiding engineering design of structures to helping determine earthquake insurance rates, and informing government officials, emergency managers, and public officials about potential risks to their communities. The distinction between hazard and risk is that hazard refers to natural phenomena from earthquakes (e.g., ground shaking), whereas risk refers to consequences of the hazard (e.g., dangers the hazard poses to life and property). The USGS is currently working on the updated hazard model for the 50 states targeted for completion and release in 2023. Let’s look at what goes into the NSHM in more detail to better appreciate the improvements and progress for the 2023 model.

Sources/Usage: Public Domain.

A map from the 2018 United States Seismic Hazard Model showing highest hazard in red and lowest hazard in gray. Damage from various earthquakes around the U.S. is shown in the photographs. Larger versions of each photo can be found at: 2001 M6.8 Nisqually, Washington, 2014 M6.0 South Napa, California, 1994 M6.7 Northridge, California, 1975 M7.7 Kalapana, Hawaii, 2011 M5.8 Mineral, Virginia, and 2018 M7.1 Anchorage, Alaska.

Faults and Earthquakes

Existing faults and locations of past earthquakes are the starting point for a seismic hazard model.  The location of faults is known through observations of topographic and landscape disruptions caused by movement on said faults. Earthquakes on these faults disrupt the surface features, and the date the earthquake occurred can be estimated by the age of the disrupted features. For example, if we know that a certain sediment deposit is 1,000 years old, and we see that this sediment has been cut and moved by the fault, we know that an earthquake on the fault reached the surface in the last 1,000 years. These observations can be made in the field or using remotely sensed imagery, like aerial imagery and other techniques. Unfortunately, earthquakes don’t just happen on mapped faults – they occur all the time in places where there is no fault on the map – but all those earthquakes are also taken into account in hazard models, as indicators of where and how often earthquakes could occur away from known faults.

In addition to the fault locations, we need to know how active each fault is, which reflects how frequently a fault has earthquakes and how big those earthquakes may be. These activity rates can be measured both by geologists, with knowledge of how much displacement has occurred along a fault over a given length of time (like our previous example of a displaced sediment layer with a known age), and by geodesists. For example, geodesists use high-precision Global Navigation Satellite System (GNSS) instruments distributed across the western United States to determine the velocities of the Earth’s surface relative to the stable and very slowly deforming eastern United States. Although earthquakes are rare in the central and eastern United States, they do occur and can cause damage (for example, the 2011 M5.8 earthquake in Mineral, Virginia).

Sources/Usage: Public Domain.

Combining fault locations and activity rates allows us to calculate the frequency of rupture along a given fault. The yearly frequencies of rupture are very low, so this is translated into a percent chance of a fault rupturing over a longer given time scale (for example, 10% in 50 years). This process of calculating times between earthquakes (known as recurrence) also considers the possibility of earthquakes jumping from one fault to another, as has been observed in past earthquakes such as the 1992 M7.3 Landers earthquake in the Mojave Desert of southern California. While the chances of earthquakes on a given fault are very small over short time scales like years to decades, it is important to remember that the chance of an earthquake is generally higher in California, especially near the San Andreas fault, than in areas with fewer active faults, such as Colorado.

Ground Motions

Once we’ve gathered all the best data for fault locations and how often they generate earthquakes, we want to know the level of shaking that each earthquake is likely to produce.

The ground shaking caused by fault rupture can be measured worldwide by very sensitive instruments called seismometers, but it is nearest the fault where shaking has the greatest human impact.  The shaking intensity at a particular location is primarily related to: 1) the magnitude (or size) of the earthquake, 2) the distance from the fault, 3) the type of soil or rock present, and 4) the depth of bedrock. All these factors combine and trade off to generate combinations of very sudden jolts of shaking (high-frequency waves) and much slower, rocking or rolling waves (low-frequency waves). High-frequency waves can rattle short buildings (e.g., 1 to 2 stories), dams, and nuclear power plants, crack stiff walls, and topple or slide building contents that are not anchored to the structure, such as appliances or furniture. Low-frequency waves can especially affect tall buildings or long bridges, causing them to sway. Localized areas of structural damage can be caused by soft and deep soils or sudden changes in geology that amplify the shaking intensity.

Databases of past ground motions are used to make models of future potential ground shaking. When combined with fault locations and earthquake frequencies, a hazard model forecasts the chances of low- and high-frequency ground shaking intensities at any location of interest. As described below, these hazards are fed into engineering project needs and construction code applications to help improve designs of new structures or guide the retrofitting plans of old ones.

Engineering & Risk

Engineering and risk research extends the hazard model to practical civil engineering applications and provides feedback for improving the hazard model. Traditionally, the NSHM serves as a starting point for designing earthquake-resistant structures. For buildings, structural engineers need to design beams, columns, and connections of buildings to withstand shaking during an earthquake. The building codes used for these designs are based on forecasted ground motion intensities derived from the NSHM. Besides collaborating with building code partners for designing buildings, such as the Building Seismic Safety Council (BSSC), USGS engineers and scientists also collaborate on designs for other structures, such as bridges with the American Association of State Highway and Transportation Officials (AASHTO) and railways with the American Railway Engineering and Maintenance-of-Way Association (AREMA).

Sources/Usage: Public Domain.

USGS map showing (1) the locations of major populations and (2) the chance of slight (or greater) damaging earthquake shaking occurring in 100 years.

In addition to engineering design, the USGS engineers and scientists also collaborate with partners on extending the hazard forecasts of ground shaking to risk forecasts of consequent damage to buildings and other structures. For example, USGS and the Federal Emergency Management Agency (FEMA) have forecasted seismic damage to the nation’s buildings in terms of yearly repair costs. Furthermore, the insurance industry uses the NSHM to assess seismic risk and determine insurance premiums.

What’s new for 2023?

As more observations are made, our knowledge of active faults increases. To capture all of what we know about active faults across the U.S., we are reviewing  new reports and literature published since the previous update to the NSHM fault model in 2014. Additionally, given improvements made between 2014 and now in computational power, modeling algorithms, and the availability of more data, we can incorporate more faults than in previous models – over 1,000 for the 2023 NSHM, whereas the previous model used about 640. This reduces our uncertainty on where earthquakes may happen in the future, and thereby increases the precision of the resulting seismic hazard model.

The 2023 NSHM will include recently updated ground motion models for subduction-zone faults (present in the Pacific Northwest and Alaska) as part of its updated national hazard assessment. The Hawaii portion of the NSHM has already been updated to reflect, in part, a swath of new earthquakes related to volcanic eruptions that have changed the way we understand ground shaking on the Island of Hawai‘i. The 2023 update for the Lower 48 states will also focus on better characterizing shaking within sedimentary basins, large geological depressions generally filled with soft sediments down to deep bedrock, where major urban areas are often centered. New studies in southern and central California, Seattle, Portland, Reno, and Las Vegas will be performed to determine if these basin regions need special treatment within the NSHM.

To broaden the utility of the NSHM beyond traditional design and assessment of buildings, USGS engineers and scientists are currently exploring potential applications that assess seismic risk for other critical infrastructure systems such as gas pipelines, highway bridges, and dams. Other engineering and risk research efforts related to the NSHM include: 1) developing earthquake scenarios for response planning, 2) extending the ground shaking forecast to ground failures such as soil liquefaction that can cause additional damage, and 3) studying the interactions of multiple hazards for engineering and risk assessments of infrastructure systems.

While incomplete, the preceding examples aim to illustrate the many different improvements to an already useful forecast of future earthquake shaking and its consequences in the United States.

~ written by Alex Hatem, Sean Ahdi, N. Simon Kwong, and Lisa Wald, June 15, 2022

For More Information

• Earthquake Hazards Program – Hazards
• Earthquake geology inputs for National Seismic Hazard Model (NSHM) 2023

Acknowledgements

The authors thank Allison Shumway, Rob Chase, and Andrew Makdisi for their suggestions for photographs and help with the first image. We also thank Nico Luco and Leah Salditch for their reviews that greatly improved the readability.