Site Map

Sinking sand, sand boils, LIQUEFACTION

Put some sand in a bucket. Place a small block on the top of the sand, and gradually saturate the sand with water. The block still rests on the surface. Now, strike the bucket sharply, and watch the block of wood quickly drop. This is called "liquefaction". The same can happen to houses built, as the Bible says, "like a foolish man who built his house on sand." (Matt. 7:26)

Sandblows at New Madrid MO

“Imagine a cube full of sand and water. If you press it in from both sides (compressing and releasing it, then compressing it again), you build up what’s called pore-water pressure,” said Martitia Tuttle, geologist and consultant for the U.S. Geological Survey. “It’s like shaking a coke can — when the pressure builds up and you release it, the fluid comes shooting to the surface.” 

These fountains of water can sometimes shoot as high as 30 feet into the air, according to Tuttle.


A sandboil / sandblow. Paleoseismologists can learn a great deal by studying them.

Sand blow diagram: USGS The USGS used the above illustration to diagram a sand blow.

The Enigma paper says Bootheel sand blows are commonly 1.0-1.5 meters thick, and 10-30 meters in diameter.

Portions of roads in the Bootheel could disappear entirely as the soft, sandy soil shakes and the heavy roads sink. Southeast Missourian.


Memphis' Pyramid

Memphis' city and county government in 1991 built a 32-story pyramid on its riverfront. It could hold 20,000 spectators, is the size of six football fields, is slightly taller than the Statue of Liberty, and held great promise as sports arena/convention center. In 1992 the city started considering earthquake construction standards. The pyramid sits smack dab on a Mississippi River sandbar. Guess what could happen to 20,000 spectators in the event of a major quake. It is being remodeled into a huge sporting goods store.


Find a sandboil

Author David Stewart says the Bootheel Lineament, now called Bootheel Fault, runs from 9 miles west of New Madrid to 4 miles west of Blytheville. He said in "Damage and Losses from future New Madrid Earthquakes" (1994) that this line, although not where the primary shaking happens nowadays, has some of the most sandboils (past activity) of this area.

Open Google earth, and find a couple of sandboils 2.3 miles south of Catron, Mo., which is west of New Madrid. They have been partially farmed over. See how many you can find. See examples below. Caution: some of the white puffiness will be clouds. Others may be "borrow pits" along highways, made during highway construction.

What is liquefaction?

 A sandblow / sandboil just west of New Madrid, from Google Earth.

liquefaction diagram explanation

"Geology students called it 'the Beach'. Located in Pemiscot County, Mo., near the town of Deering, the 136-acre strand is purported to be the largest sand boil in the world, a vestige of the New Madrid earthquakes of 1811-1812." Riverfront Times. The Google Earth image below may be "the beach". This image is near Deering. A similar sandblow is "Daytona Beach" SW of Memphis along the Marianna fault. Some of the sand has been scooped out and such areas turned into rice production. One half-acre plot we found is not noticeably different from surroundings except it has erosion control grasses on it (nothing else will grow) and on close examination is covered with whitish sand.

Construction has gone on in the Bootheel, sometimes oblivious to soil problems. Consider the industrial park at Marston (below)., 

The Noranda Aluminum plant and Associated Electric power plant are together, in mid-right of above photo. Note the sandy soil elements in bottom center of picture.  Town of Marston is at left. Marston rest area of I-55 is at top. New Madrid is just upriver. The power plant with its 800 foot smokestack, (just north of the aluminum plant) (pic is near ground zero of the largest quake ever to hit the U.S.  See the red box 5 in image below.

Mississippi River ran backwards 1812

"MilePost 44 fissure" - Author David Stewart in "New Madrid Fault Finders' Guide" says the sandy soil in mid pic ends in a sand blow beneath the interstate. He says I-55 and the railroad would both fail here during a heavy quake. 

Brown soil in lower right is Mississippi River levee. Rail line runs alongside it, bringing a train load of coal per day to the nearby power plant. This is 1.5 miles north of Marston rest area. Howardville is two miles beyond top of pic.



Sikeston Municipal power plant just NW of town is in lower right. Smokestack in middle of complex does not show up properly on Google Earth. Note the oval of discolored soil just north of the plant that seems to run underneath the power station. A brighter patch (not shown) similar to "the beach" above, but smaller, is a few miles north of power plant.

The Sikeston City power plant straddles a large sand fissure. The Bootheel is littered with sand boils. A cable TV tower alongside I-55 near Matthews is on a sand blow crater. And in Malden, an electric substation was built over a sand boil. sand boil pics. Some of the property was likely purchased because it was inexpensive... because it would not grow crops. (information from Knox and Stewart, "The New Madrid Fault Finders Guide")

Photo - multiple sandboils in India. | Liquefaction info. |  Slideshow 1 - PDF  | Slideshow 2 - PDF

liquefaction diagram explanation

US 412 heads for Kennett to the left. I-155 toward Dyersburg on the right, from Hayti-Caruthersville intersection of I-55. Note the sandy spots of soil in lower left of picture. These may indicate previous sand blows. Consider what sandy soil and liquefaction could do to heavy interstate bridges and pavement during a serious earthquake. I-55 is the main connection between St. Louis and Memphis and between Chicago and New Orleans.

Liquefaction features documented at 112 sites indicate that very large earthquakes have struck the New Madrid region in

  • A.D. 900 + 100

  • A.D. 1450-1470

  • and the Western Lowlands [Poplar Bluff MO region] in A.D. 1380 + 70 yr.

A large sand blow exposed along the Current River in the Western Lowlands is suggestive of a large earthquake about A.D. 1380. It may be related to others on the St. Francis River.

The area and size distributions of related liquefaction features suggest that the A.D. 900 event was quite similar to the 1811-1812 earthquake sequence. The A.D. 900 event induced liquefaction over a large area and may have been of M > 7.4. If the features were formed as a result of a single earthquake, the areal and size distributions of liquefaction features suggest that the earthquake may have been centered near Blytheville, Arkansas.

Less is known about the A.D. 1450-1470 event but it, too, may have been very large. ...strong evidence that this event occurred during the Late Mississippian cultural period (A.D. 1400-1670). The areal and size distribution of liquefaction features suggest that the earthquake may have been centered near Blytheville, Arkansas, and been of M > 7.2

It appears that very large earthquakes have occurred in the NMSZ every 200 to 800 yr during the past 1200 yr.

In addition, we have evidence for at least three earlier events between 4040 B.C. and A.D. 780. Prehistoric liquefaction features found as far away as southern Illinois may be related to New Madrid seismicity.

from Tuttle, Schweig 1997

A geologic fault is defined as a break in earth materials (rocks or soil) with relative displacement between both sides.

Primary faults cause earthquakes. Secondary faults are caused by earthquakes. There are numerous secondary faults in the NMSZ that can be seen, but no visible primary faults.

There is a lineament (now called fault) through the Bootheel of Missouri visible by high resolution satellite imagery (Schweig & Jibson, 1989). This lineament is almost a perfect, straight line dividing differences in surface soils and topography from four miles west of Blytheville to nine miles west of New Madrid. Some of the largest seismic sand boils are along this line.

"Damages and Losses from Future Earthquakes" by David M. Stewart, Southeast Missouri State University, 1993, Care Publications. ISBN 0934426538

Tuttle and USGS geologists found that recently excavated sandblasts predated the famous New Madrid quake by hundreds of years. Using radioactive carbon dating techniques, they estimated the ages of hickory nuts, corn kernels or other organic material they found in sediment above and below the blasts.

Tuttle's most recent results, published in 2002, suggest that other massive New Madrid earthquakes happened about A.D. 900 and 1450. Some preliminary evidence points to another quake about A.D. 300, she said. If the pattern holds--one huge quake every 500 or 600 years--the New Madrid zone would see another such event within the next few centuries.

The fault system responsible for New Madrid seismicity has generated temporally clustered very large earthquakes in A.D. 900 ± 100 years and A.D. 1450 ± 150 years as well as in 1811–1812. Given the uncertainties in dating liquefaction features, the time between the past three New Madrid events may be as short as 200 years and as long as 800 years, with an average of 500 years.

... It appears that fault rupture was complex and that the central branch of the seismic zone produced very large earthquakes during the A.D. 900 and A.D. 1450 events as well as in 1811–1812. On the basis of a minimum recurrence rate of 200 years, we are now entering the period during which the next 1811–1812-type event could occur.

Guccione and others, GSA Bulletin; March 2005; v. 117; no. 3-4; p. 319-333; DOI: 10.1130/B25435.1

The sands, silts, and clays are an average 477 meters deep at New Madrid, deepening to 987 meters near Memphis. The numbers are from a study by Roy Van Arsdale and Robin TenBrink.

5,000 and 7,000 years ago - "Daytona Beach" sandblow
now growing cotton on white sand SW of Memphis.

At the Burkett archeological site in southeastern Missouri [near Charleston], there is evidence for at least six earthquakes that induced liquefaction and related ground failures during the past 4,500 yr. The four earliest earthquakes occurred in 2350 B.C. ± 200 yr and may have been part of an earthquake sequence occurring over a period of weeks to months. Sand blows and related sand dikes that formed as a result of these earthquakes served as the foundation on which a Late Archaic and Woodland cultural mound was built. The fifth earthquake occurred in A.D. 300 ± 200 yr towards the end of the Middle Woodland period. The sixth and latest earthquake occurred after A.D. 1670. Given that it produced small sand blows and liquefaction-related ground failures not far from the site, the 1895 Charleston, Missouri, earthquake was more likely the cause of the small historic sand dikes at Burkett than the 1811-1812 New Madrid earthquakes.

The 2350 B.C. and A.D. 300 earthquakes are interpreted as 1811-1812-type or New Madrid events. This interpretation is based on the relatively large size of the prehistoric liquefaction features at Burkett, the close timing of the four earthquakes in 2350 B.C., and the likelihood that the 2350 B.C. and A.D. 300 earthquakes also induced liquefaction at other sites in northeastern Arkansas and southeastern Missouri.


What is liquefaction?

Below: Cross-section of a sand blow in New Madrid area, likely in Wolf River in Collierville, SE of Memphis.    USGS photo.

sand blow cross-section in New Madrid area from Alabama GSA


 Sand blow near Charleston Mo.

Deformation, liquefaction area of New Madrid region

Below: Map of larger sand boils

sand blows sand boils map New Madrid area

At three sites near Blytheville, Arkansas, in the central part of the New Madrid seismic zone, one sand-blow crater formed between A.D. 800 and 1400, two sand-blow deposits formed between A.D. 800 and 1670, and three, possibly four, sand dikes formed since 4035 B.C. - source

In Jan., 2001, a 7.9 earthquake struck Bhuj, in extreme western India. Earthquake investigators say it has similarities to the worst-case New Madrid scenario. 


Photos of New Madrid Seismic Zone are from Google Earth

An uplift dammed the Mississippi for a few hours early on the morning of Feb. 7, 1812 from about Island #10 north past island #9. Ice breaking up on the Ohio River had just allowed several boats to make it to the region the day before. Multiple eyewitness stories came from those moored near Island #9, rudely awakened after 2 a.m, and found their boat going upstream "at the speed of a fast horse" -- "had to hold my hat on" -- while trees continued collapsing into the river. The air smelled of sulfur and was filled with coal dust.

Caruthersville-Dyersburg I-155 Bridge

Little Prairie (Caruthersville) was washed away. The people had all fled.

Above: The I-155 bridge between Caruthersville MO and Dyersburg TN is in the heart of the New Madrid Seismic Region. See map below. The Blytheville Arch is SW. The Reelfoot fault / Pascola Arch is NE.

Caruthersville Dyersburg I-155 bridge

The bridge was opened in the 1970s. A 1995 seismic retrofit proposal presented expensive challenges which have not been done: 

This note concerns the Dyersburg-Caruthersville I-155 bridge retrofit proposal:

Our current information shows that bedrock is located beneath about 2800 feet of sands, gravels, and hard clay strata.

Construction of the bridge was completed in 1974. The approach spans are supported on friction pile that were driven to about 50 feet in depth. The main truss spans are supported deep caissons that bear on dense sands 50 to 90 feet below the river bottom.

There was a partially completed seismic retrofit study in 1995 that projected a cost of $32M. I wouldn't characterize that as a fully developed cost estimate since the study was not completed and there were unresolved issues identified that could result in the cost estimate going up significantly. Also, the cost of bridge work has inflated since 1995. Funding was the issue back then in not completing a retrofit.

Greg Sunde
MoDOT – Bridge Division - Email to webmaster, Dec. 4, 2006


Subsurface Conditions (from the 1995 report)

  • River Alluvium
    • Loose sand or soft to firm clay - upper 30 to 70 feet
    • Shear wave velocities 500 to 600 fps
  • Mississippi Embayment
    • Sand, gravel, and hard clay
    • Extends to bed rock at 2700 feet

Located 6 Miles From New Madrid Fault

  • Widespread liquefaction to depths of 40 feet
  • Dynamic settlement 3 to 17 inches - Lateral spreading 1 to 13 feet
  • Critical structural components have 25% to 60% of required capacity

Above illustration puts you in mid-Bootheel of Missouri, looking north at US 60 with Sikeston on the right, Poplar Bluff on the left. A section of Crowley's Ridge in the middle, runs north-south through Dexter-Bloomfield.

During the Civil War, the yellow on the right (between Dexter and Sikeston) was a swamp, with waist deep water, which inhibited troop movement. In a big shake, the yellow area, running north-south, would see the most liquefaction. Note location of I-55 at right.

Tiptonville Dome, Reelfoot Lake

liquefaction diagram explanation

“The ground failure that resulted from liquefaction during the New Madrid earthquakes was severe. We're talking about vertical displacement of 3 to 6 feet (1 to 2 meters), and lateral displacement up to 33 feet (10 meters),” said Tuttle. “A recurrence of that type of event would have severe consequences for engineered structures.” NASA

Above: Another sandboil, farmed over, just east of New Madrid.

In a letter published in Lorenzo Dow’s Journal, New Madrid resident Eliza Bryan wrote in 1816:

. . . the surface of hundreds of acres was, from time to time, covered over in various depths by the sand which issued from the fissures, which were made in great numbers all over this country, some of which closed up immediately after they had vomited forth their sand and water . . .

Above: Apparent sandblows straddle I-55 just south of Sikeston.

Below: apparent sandblows just east of Malden.

Next photo shows apparent sandblows just north of Malden.

Above: Big Lake National Wildlife Reserve, west of Blytheville, Ark, once a free-flowing river system, changed to a lake / swamp ecosystem by the New Madrid earthquakes of 1811-12. Big Lake today consists primarily of wooded swamps and open water with an average depth of three feet, bordered by a virgin cypress-tupelo swamp. Five large drainage ditches empty into the reserve. Blytheville Air Base is at top right.

Seismic History - Paleoseismicity

Although the New Madrid earthquakes of 1811 and 1812 are the best known seismic activity in the NMSZ, Native American traditions have recorded several previous earthquake shocks that allegedly “devastated’ the region (Fuller, 1912).

Fuller noted ground cracks with trees over 200 years old growing in them, suggesting an older series of earthquakes than the 1811-1812 sequence. Fuller also found faults extending through the Lafayette gravel in the Chickasaw Bluffs and on Crowley´s Ridge that appeared to pre-date the 1811-12 New Madrid earthquakes.

chickasaw bluffs


Search YouTube for "liquefaction"

Using paleoliquefaction and archeological evidence, Tuttle and Schweig (1995) determined that two, possibly three prehistoric earthquakes, occurred near Blytheville, AR in the past 5000 to 6000 years. One sand blow crater in the area was dated between A.D. 800 and 1400, and another between A.D. 800 and 1670.

According to Tuttle and Schweig, prehistoric liquefaction features in the Blytheville area suggest recurrence intervals of hundreds of years for earthquakes large enough (mb = 6.2) to cause liquefaction. The prehistoric liquefaction features found were equal to or larger than the liquefaction features formed by the 1811-1812 earthquakes. This suggests that the prehistoric earthquakes may have been of similar sizes and type to the 1811-1812 earthquakes.

Kelson, et al., (1994), found evidence of three prehistoric earthquake events in the last 2400 years while studying the Reelfoot scarp. Saucier (1991) also identified paleoliquefaction features in the NMSZ. Paleoliquefaction features found near Reelfoot Lake were interpreted as being formed by two large prehistoric earthquakes that occurred within the past 1500 years.

Historical Seismicity

Aftershocks strong enough to be felt continued through 1817. The earthquakes were felt over an area of 5 million km2 (Van Arsdale, 1997), more than any other earthquake in American history. The area experiencing a Modified Mercalli Intensity (MMI) (Table 3.1) of V or greater was approximately 2.5 million km2 (Nuttli, 1973).

This area of strong shaking is two to three times greater than the area affected by the 1964 Alaska earthquake and 10 times larger than the area affected by the 1906 San Francisco earthquake (Stover and Coffman, 1993). Figure 3.6 is a map showing the Modified Mercalli Intensities experienced during the December 16, 1811 earthquake.

The December 16th earthquake was so severe that it awakened people as far away as Pittsburgh, PA, and Norfolk, VA. The felt areas of the three largest earthquakes extended to the gulf coast to the south, the Atlantic coast to the southeast, and to Quebec, Canada, in the northeast. Major topographic changes, including large ground fissures, sand blows, sunk lands, caving of river banks, and disappearing islands in the Mississippi River affected an area between 78,000 and 130,000 km2 (von Hake, 1974; Stover and Coffman, 1993).

Use of material from this website: We encourage use of these pics & graphics elsewhere. No credit to this site needed. See Google Earth use policies.





Powered by