
construct and defend a scientific explanation, with calculations from real data, of why the New Madrid Seismic Zone produces earthquakes in the middle of a plate and what that means for people.
Instruction
The question. Why does Missouri have earthquakes when it sits in the middle of the North American plate, how does the New Madrid Seismic Zone move, and what does that mean for the people who live near it? Your explanation ties together the whole unit, and every part must rest on evidence.
What to include.
- The plates move. Evidence from the sea floor (magnetic stripes, ocean-floor ages that never exceed about 180 million years) and from GPS: Missouri moves about 14.5 mm/yr with its plate while barely deforming, and Neah Bay on a plate boundary is shoved 14.6 mm/yr. Explain why continental crust can be billions of years old while ocean crust cannot.
- The engine. A model of the interior from seismic evidence (the S-wave shadow zone and the liquid outer core), and mantle convection with sinking slabs as the driver.
- The history. How Missouri’s crust was assembled, how the Reelfoot Rift opened about 565 million years ago and failed, and why its buried faults are the weak spot that gives way today.
- The motion today. The Reelfoot scarp slip rate, the GPS rate that is 30 times smaller, the sand-blow record of large earthquakes, and the Gutenberg-Richter pattern in the USGS counts.
- What it means for people. Use the USGS 50-year probabilities to say what Missourians should expect, and name one thing that forecast cannot tell them.
Three calculations. Choose three, and show every step:
- a slip rate from an offset and an age (13.7 m in 2,300 years gives 6.0 mm/yr);
- an earthquake distance from an S-P lag (a 7.5 s lag gives 8.4 × 7.5 = 63 km);
- a magnitude comparison (the M7.5 of December 1811 against the M4.0 at Cooter is 103.5 ≈ 3,200 times the amplitude and 105.25 ≈ 178,000 times the energy);
- a Gutenberg-Richter estimate from the fitted line log10N = 5.52 − 1.19M (for M = 4.0, N ≈ 100.76 ≈ 6 in 25 years, against 6 observed).
A worked paragraph. Claim: “The New Madrid Seismic Zone breaks because the slowly squeezed North American plate gives way along the buried faults of an ancient failed rift, but the rate of loading there is very slow and may change over time.” Evidence: “The Reelfoot scarp records 13.7 m of slip in about 2,300 years, about 6 mm per year, while GPS stations across the zone move together to within 0.2 mm per year.” Reasoning: “Because a fault slips in an earthquake only when stored strain overcomes friction, a fault that GPS shows is barely loading must be spending strain stored long ago, or slowing down, so its future cannot be read off a simple repeat schedule.”
Care with people. When you write about earthquakes that killed and injured people, or false forecasts that frightened them, name the human cost accurately and without drama. Your explanation is part of how communities prepare.
Limits. Say what is uncertain: why GPS and geology disagree, whether the recent cluster of large earthquakes is ending, the dating uncertainties in the sand blows, and how small counts of large earthquakes make the upper end of the Gutenberg-Richter line uncertain.
📖 Read for this lesson
The reading opens over the lesson; close it to come back to exactly where you were. The checks below draw on it.
Formative check
Work through these before moving on. They are not graded — they tell you whether the standard below has landed yet.
Using log₁₀N = 5.52 − 1.19M for 25-year New Madrid counts, about how many M ≥ 3.0 earthquakes does the line predict?
The New Madrid Seismic Zone follows the buried faults of the Reelfoot Rift, a failed rift that opened as the supercontinent Rodinia broke apart.
Write the ‘motion today’ paragraph of your explanation in claim-evidence-reasoning form, using the Reelfoot scarp slip rate, the GPS rate and one other calculation.
How confident are you that you can construct and defend an explanation of the New Madrid Seismic Zone with calculations from real data?
Unit 3 checkpoint
Do this before you start the task. It covers the whole unit and tells you what to review first.
Where is the youngest ocean floor?
Magnetic stripes on the sea floor are symmetric on both sides of mid-ocean ridges.
An S-P lag of 20 s means the earthquake is about km away (d = 8.4 × lag).
What evidence shows that the outer core is liquid?
The mantle is liquid rock, which is why it convects.
An M6.0 compared with an M4.0 releases about how many times more energy?
The Reelfoot scarp’s 13.7 m of slip in about 2,300 years gives a slip rate of about mm/yr.
The b-value of about 1.2 in the USGS New Madrid counts means:
What formed the granite of the St. Francois Mountains?
The Reelfoot Rift, which the New Madrid Seismic Zone follows, opened when the supercontinent Rodinia began to break apart.
Submit: Why Missouri Shakes in the Middle of a Plate (1,000-1,300 words) + three-minute defense
Task: Write a scientific explanation of the New Madrid Seismic Zone. Include (1) evidence that plates move and why ocean crust is younger than continental crust, with the GPS comparison of Missouri and Neah Bay; (2) a model of Earth’s interior and mantle convection built from seismic evidence; (3) a labeled model of how Missouri’s crust and the Reelfoot Rift formed and why the rift is active today; (4) at least three calculations from real data in this unit; (5) what the USGS 50-year probabilities mean for Missourians, and their limits. Hand-drawn models are fully acceptable. Cite Earle’s Physical Geology and the USGS and research sources you use.
Submission workflow
- Stage 1 — Plan. Your central claim, an evidence table (observation, source, what it shows), your three calculations worked out, and a sketch of your plate-boundary model.
- Stage 2 — Checked draft. Recheck every calculation against the worked values in the unit readings, check that each section links evidence to the claim with reasoning, and mark any number without a source. A family member may read the draft and tell you where it is unclear. Submit your check notes and your revision plan.
- Stage 3 — Final explanation. 1,000-1,300 words with the model, three calculations, sources cited and a limits paragraph.
- Stage 4 — Defense. A three-minute recording, followed by a written answer to one question about the limits of your evidence or your model.
How it is scored
| Criterion | What proficient looks like |
|---|---|
| Evidence for plate motion and crustal ages (9-12.ESS1.C.1) | Uses sea-floor and GPS evidence with specific numbers (Missouri against a plate-boundary station), and explains why ocean crust is young and continental crust old. |
| Interior and convection (9-12.ESS2.A.3) | Explains the layered interior from seismic evidence and models convection as the cycling of matter that drives the plates. |
| Crustal-history model across scales (9-12.ESS2.A.1, CCC.7) | Model shows volcanism, rifting, collision and today’s compression of the Reelfoot Rift, with processes at scales from seconds to hundreds of millions of years. |
| Quantitative reasoning (SEP.5) | Three calculations correct, with units and assumptions stated. |
| Explanation (SEP.6, WHST.9-10.2) | Clear mechanism linking slow loading in the middle of a plate to sudden fault slip; limits and human impact handled accurately. |
| Use of sources (RST.9-10.1) | Specific evidence cited from Earle, USGS data and the New Madrid research. |
The Reelfoot Scarp Slip Rate, GPS That Sees Almost Nothing, the Dated Sand Blows, and 25 Years of New Madrid Earthquakes Counted
University of Arkansas news release on Guccione, Mueller and others (Science 1999); Purdue University news release and abstract, Calais and Stein, Time-Variable Deformation in the New Madrid Seismic Zone (Science 2009); Stein, A GPS-based view of New Madrid earthquake hazard (USGS NSHM CEUS workshop, 2012); Tuttle and others 2002 (BSSA 92:2080) via USGS Publications Warehouse; USGS ComCat counts for 35-38 N, 88-91 W, 2000-2024; USGS Circular 1083 (1993); WTVG 13abc, Finding fault (December 4, 2020); all retrieved November 6, 2026 · SLS original (cites USGS public-domain data and publications and published research and news reports as facts; brief quotations only)
Why this reading. The lab asks you to rebut a popular New Madrid claim with at least two calculations. Here are the sources behind the lesson’s numbers: the Reelfoot scarp slip rate, the GPS result that disagrees with it, the dated sand blows, the catalog counts re-run on November 6, 2026 with the least-squares fit reproduced, and the 1990 forecast that the claims echo. The USGS 50-year probabilities are in the reading on the New Madrid USGS sources.
1. The Reelfoot scarp: a dated offset
A 1999 University of Arkansas news release on work by Margaret Guccione, Karl Mueller and colleagues (published in Science, 1999) describes trenches across the Reelfoot scarp, “a geologic fold that crosses an abandoned Mississippi River channel near Tiptonville, Tenn.” The team measured the widths of folded zones (“kink bands”) and dated the folded sediment, and found a slip rate of “about six millimeters per year over the past 2,300 years, totaling nearly 45 feet”; extending to the whole zone gave about 4.8 to 6.1 mm/yr. Check: 45 ft × 0.3048 = 13.7 m; 13,716 mm / 2,300 yr = 5.96 mm/yr; the lesson rounds to 6. Inverse: 13.7 m / 0.0060 m/yr = 2,283 years.
2. GPS: almost no motion
A 2009 Purdue University news release on Eric Calais and Seth Stein’s Science paper “Time-Variable Deformation in the New Madrid Seismic Zone” reports that “the ground surrounding the fault system is moving at a rate of less than 0.2 millimeters per year and there is likely no motion,” from nine GPS antennas in the zone. The paper’s abstract, printed in the release, says the zone “is currently deforming too slowly – if at all – to account for large earthquakes in the region over the past 5,000 years.” A 2012 presentation by Stein to a USGS hazard-mapping workshop adds that the recent earthquakes correspond to slip of at least 1-2 mm/yr over the past ~2,000 years, that at the GPS rate a magnitude 7 would be at least 10,000 years away, and that the lack of significant fault topography implies the recent pulse of activity is only a few thousand years old. Lesson 3.1’s own check: PTGV and RLAP differ by 0.29 ± 0.22 mm/yr.
The budget, worked. 6 mm/yr ÷ 0.2 mm/yr = 30, so the geological rate is at least 30 times the GPS rate. Slip stored at 6 mm/yr in a 500-year cycle: 6 × 500 = 3,000 mm = 3 m. Time to store 3 m at 0.2 mm/yr: 3,000 / 0.2 = 15,000 years, consistent with Stein’s “at least 10,000 years.”
3. The sand-blow record
Tuttle and others (2002), Bulletin of the Seismological Society of America 92:2080, abstract on the USGS Publications Warehouse: the fault system “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,” and “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,” from liquefaction features “at more than 250 sites.” Check with the central dates: 1450 − 900 = 550 years; 1811 − 1450 = 361 years; two intervals over 911 years average 456 years. The published 500 allows for the dating uncertainties.
4. Table 1. The catalog counts
USGS ComCat via the FDSN event service, earthquakes only, 35-38°N, 88-91°W, January 1, 2000 through December 31, 2024, queried November 6, 2026. Counts at M ≥ 1.0 (6,288) and 1.5 (4,314) are left out of the fit because the network does not record every earthquake that small everywhere in the box.
| Magnitude ≥ | 2.0 | 2.5 | 3.0 | 3.5 | 4.0 | 4.5 |
|---|---|---|---|---|---|---|
| Count | 1,420 | 397 | 69 | 26 | 6 | 0 |
| log10(count) | 3.152 | 2.599 | 1.839 | 1.415 | 0.778 | — |
The six events of magnitude 4.0 or more: 2003-04-30 M4.0 near Blytheville, Arkansas; 2003-06-06 M4.0 near Bardwell, Kentucky; 2005-02-10 M4.1 near Etowah, Arkansas; 2005-05-01 M4.2 near Manila, Arkansas; 2005-06-02 M4.0 south of Ridgely, Tennessee; 2021-11-18 M4.0 south of Williamsville, Missouri. (The 2026 Cooter M4.0 falls after this window.)
The fit reproduced. Least squares on the five points gives log10N = 5.5159 − 1.1864M, which rounds to the lesson’s 5.52 − 1.19M. Predictions from the unrounded line: M ≥ 3.0, 90.5 (observed 69); M ≥ 3.5, 23.1 (observed 26); M ≥ 4.0, 5.89 (observed 6); M ≥ 4.5, 1.50 (observed 0); M ≥ 5.0, 0.38 in 25 years, one every 65 years; M ≥ 7.0, 0.0016, one every 15,382 years. With the rounded coefficients: M ≥ 3.0 gives 101.95 = 89; M ≥ 3.5 gives 101.355 = 22.6; M ≥ 4.0 gives 100.76 = 5.75; M ≥ 7.0 gives one every 16,100 years. The b-value, 1.19, means each magnitude step cuts the count by 101.19 ≈ 15.5. The sand blows record magnitude 7-size earthquakes about every 500 years, some 30 times more often than the line predicts: small earthquakes do not forecast the big ones here.
Rates. 1,420 events of M ≥ 2.0 in 25 years = 56.8 per year = 1.09 per week. Energy. M7.5 against M4.0: 101.5 × 3.5 = 105.25 = 177,828, about 178,000; amplitude 103.5 = 3,162.
5. The 1990 forecast
In 1990 Iben Browning, a business climatologist, gave a 50-50 chance of a major earthquake centered on New Madrid around December 3, 1990, based on tidal forces when the Sun and Moon lined up. The USGS later published a 248-page study of the public response, Responses to Iben Browning’s prediction of a 1990 New Madrid, Missouri, earthquake (Circular 1083, 1993). A 2020 WTVG (13abc) retrospective reports that a scientific review committee found the reasoning unsound, that no earthquake came, and that schools and mines closed; a geologist interviewed estimated losses “probably in the millions of dollars.”
Using this against the three claims
“Overdue, so it will happen this year”: the recurrence is 200 to 800 years with an average near 500, GPS shows the zone barely loading, and the USGS gives 7 to 10 percent for a repeat of 1811-1812 in 50 years; none of that is a date. “Split the country in two”: the Reelfoot fault is a thrust that pushes one block up over another by meters in a great earthquake and about 6 mm per year at its fastest; the rift failed 515 million years ago and is being squeezed, not pulled apart. “Small earthquakes prevent big ones”: one M7.5 releases the energy of about 178,000 M4.0 events, and the zone had six M4.0 or larger in 25 years.
Sources: University of Arkansas news, Research in New Madrid Seismic Zone points to possible earthquake threat (1999) (news.uark.edu); Purdue University news release, Calais and Stein, New Madrid (March 2009) (purdue.edu); Stein, A GPS-based view of New Madrid earthquake hazard (USGS National Seismic Hazard Model CEUS workshop, 2012) (earthquake.usgs.gov); Tuttle and others 2002, The earthquake potential of the New Madrid seismic zone (USGS Publications Warehouse) (pubs.usgs.gov); USGS FDSN event web service (count queries for Table 1) (earthquake.usgs.gov); USGS Circular 1083, Responses to Iben Browning’s prediction of a 1990 New Madrid, Missouri, earthquake (pubs.usgs.gov); WTVG 13abc, Finding fault: The failed Midwest earthquake prediction, 30 years later (2020) (13abc.com)
The Sea-Floor Evidence in USGS’s Words, GPS Velocities from Missouri and a Plate Boundary, and a Dated 30-Day Earthquake Snapshot
USGS, This Dynamic Earth (Kious and Tilling), online edition; USGS ComCat event nm60620376 and FDSN counts of M2.5+ earthquakes 2026-10-06 to 2026-11-05 UTC; Nevada Geodetic Laboratory MIDAS velocity files (IGS14 and NA frames; Blewitt and others 2016); ICS chart v2023/09 (Chibanian base 0.774 Ma); all retrieved November 6, 2026 · SLS original (cites USGS public-domain text and catalog data, Nevada Geodetic Laboratory GPS velocities and the CC BY 4.0 ICS chart; brief quotations only)
Why this reading. The lesson’s evidence table asks you to rate the evidence that plates move, and its GPS paragraph compares Missouri with a station on a plate boundary. Here is the sea-floor evidence in the U.S. Geological Survey’s own words, the GPS velocities behind the lesson’s numbers (retrieved November 6, 2026), a dated table of real earthquake counts by region for the map task, and the Cooter earthquake facts the lesson opens with.
1. The phenomenon, from the USGS catalog
USGS ComCat event nm60620376: magnitude 4.0 (moment magnitude), April 23, 2026, 17:59:04 UTC (12:59 p.m. Central Daylight Time), “1 km WNW of Cooter, Missouri”, at 36.054°N, 89.826°W and 11.8 km depth. The event summary lists 913 “Did You Feel It?” reports and a maximum community intensity of 4.3. In the ZIP-code table, the farthest Missouri reports came from Monroe City and Joplin, both 432 km away; Memphis and Cape Girardeau reported intensities near 3. Distance to the nearest plate boundary: the Cayman Trough at the edge of the Caribbean plate passes near 18.5°N, 86°W, about 1,990 km from the epicenter by great-circle distance (SLS computation). The 1811-1812 magnitudes (7.5, 7.3, 7.5) are those in the USGS summary of the sequence (see the reading on the New Madrid USGS sources).
2. The sea-floor evidence, as USGS states it
From This Dynamic Earth, the USGS account of plate tectonics:
- Wegener’s evidence included “distinctive fossil ferns (Glossopteris) discovered in now-polar regions, and the occurrence of glacial deposits in present-day arid Africa”.
- Magnetic striping: “Alternating stripes of magnetically different rock were laid out in rows on either side of the mid-ocean ridge: one stripe with normal polarity and the adjoining stripe with reversed polarity.” Spreading was supported because “(1) at or near the crest of the ridge, the rocks are very young, and they become progressively older away from the ridge crest; (2) the youngest rocks at the ridge crest always have present-day (normal) polarity; and (3) stripes of rock parallel to the ridge crest alternated in magnetic polarity.”
- Ages: scientists dated the striping “for nearly all of the ocean floor, parts of which are as old as 180 million years.”
- Rates: “The rate of spreading along the Mid-Atlantic Ridge averages about 2.5 centimeters per year (cm/yr), or 25 km in a million years.”
3. The reversal age and the spreading-rate example
The most recent reversal (the Brunhes-Matuyama boundary) defines the base of the Chibanian age on the International Chronostratigraphic Chart (v2023/09) at 0.774 million years. The lesson’s illustrative example uses 0.78 million years and 39 km: 3,900,000 cm / 780,000 yr = 5.0 cm/yr; with 0.774 it is 5.04. Over 80 years, 5 cm/yr gives 400 cm = 4 m.
4. Table 1. GPS velocities (Nevada Geodetic Laboratory, MIDAS)
The Nevada Geodetic Laboratory processes data from thousands of permanent GPS stations and publishes a velocity for each, both in a global frame (IGS14, which measures motion against Earth as a whole) and in a frame that holds the North American plate fixed. Files midas.IGS14.txt and midas.NA.txt, retrieved November 6, 2026; method: Blewitt and others (2016). Speeds are the horizontal velocity, rounded.
| Station | Where | Data span (decimal years) | Speed against Earth as a whole, IGS14 frame (mm/yr) | Speed against the North American plate, NA frame (mm/yr) | Typical uncertainty (mm/yr) |
|---|---|---|---|---|---|
| MOBF | St. Louis County, Mo. | 2012.0-2024.7 | 15.2 toward the west | 0.92 toward the north-northwest | ±0.16 |
| MOJC | near Jefferson City, Mo. | 2009.8-2024.7 | 14.4 toward the west | 1.20 toward the north-northeast | ±0.17 |
| MOSF | near Springfield, Mo. | 2009.7-2024.7 | 14.2 toward the west | 0.85 toward the north-northeast | ±0.16 |
| MACC | St. Francois County, Mo. | 1999.8-2024.7 | 14.5 toward the west | 1.21 toward the north | ±0.14 |
| PTGV | Portageville, Mo. (Bootheel) | 1999.2-2024.7 | 14.1 toward the west | 1.29 toward the north | ±0.14 |
| RLAP | near Reelfoot Lake, Tenn., 32 km east of PTGV | 1997.8-2024.7 | 14.3 toward the west | 1.07 toward the north | ±0.17 |
| NEAH | Neah Bay, Wash. (Cascadia plate boundary) | 1996.0-2024.7 | 5.5 toward the south-southwest | 14.57 toward the northeast | ±0.17 |
| KOD5 | Kodiak, Alaska (Aleutian plate boundary) | 1996.1-2019.9 | 17.2 toward the west | 22.04 toward the north-northwest | ±0.40 |
| MKEA | Mauna Kea, Hawaii (Pacific plate) | 1996.8-2024.7 | 71.9 toward the west-northwest | not listed (station is on another plate) | ±0.17 |
Reading the table. Against Earth as a whole, the Missouri stations all move 14 to 15 mm/yr toward the west: the whole plate is moving, carrying Missouri with it. Hawaii’s Mauna Kea, on the Pacific plate, moves about 72 mm/yr toward the west-northwest, five times faster and in a different direction; the boundary between such plates is where the crust breaks. Against the North American plate, the Missouri stations keep only about 1 mm/yr, nearly all of it a shared northward drift that comes from how the “fixed North America” frame is defined for the whole continent; what matters for a fault is the difference between nearby stations. PTGV and RLAP sit on opposite sides of the New Madrid Seismic Zone, 32 km apart. Their difference is -0.17 mm/yr east and -0.23 mm/yr north, a relative speed of 0.29 mm/yr, with an uncertainty of about ±0.22 mm/yr in each direction: not distinguishable from zero. Published studies with more stations put the bound below 0.2 mm/yr (Lesson 3.3). Neah Bay and Kodiak, on plate boundaries where an ocean plate is sinking under North America, are shoved 14.6 and 22.0 mm/yr toward the continent relative to its stable interior, because the locked boundary drags the edge of the plate. That is the contrast the lesson builds on.
Worked. MACC, 14.5 mm/yr × 25 yr = 362.5 mm ≈ 0.36 m. NEAH relative to the plate, 14.6 mm/yr × 80 yr = 1,168 mm ≈ 1.2 m. PTGV-RLAP, 0.29 mm/yr × 80 yr = 23 mm, smaller than its own uncertainty.
5. Table 2. Where earthquakes happened in 30 days (USGS catalog snapshot)
USGS ComCat, earthquakes of magnitude 2.5 and above from 2026-10-06 to 2026-11-05 UTC, counted inside latitude-longitude boxes. The live map shows today’s; this snapshot will not change.
| Region (box) | Setting | M2.5+ earthquakes in 30 days |
|---|---|---|
| Whole world | — | 2,089 |
| Alaska and the Pacific Northwest (42-62°N, 170-120°W) | Subduction zones | 389 |
| Indonesia and New Guinea (12°S-8°N, 95-155°E) | Subduction zones | 185 |
| Chile and Peru (45-5°S, 80-66°W) | Subduction zone | 64 |
| Japan (30-46°N, 128-148°E) | Subduction zones | 57 |
| Central United States plains (35-49°N, 105-85°W) | Plate interior | 9 |
| Interior Brazil and Argentina (35-5°S, 65-40°W) | Plate interior | 4 |
| Northern Europe (50-70°N, 0-30°E) | Plate interior | 1 |
| Western Australia (35-15°S, 112-130°E) | Plate interior | 0 |
| Sahara (15-32°N, 15°W-30°E) | Plate interior | 0 |
The four plate-boundary boxes hold 695 of the month’s 2,089 earthquakes; the five plate-interior boxes hold 14. Earthquakes line up along boundaries. The largest in the Alaska box was a magnitude 5.5 on October 6, 92 km north-northwest of Aleneva, Alaska, 88 km deep. In the New Madrid box (35-38°N, 91-88°W) the same month had 1 earthquake of magnitude 2.5 or more (M2.6 on October 16, 9 km west-southwest of Equality, Illinois) and 16 of magnitude 1.0 or more, most of them near Tiptonville, Tennessee, and Risco and Holland, Missouri: a small, persistent cluster in the middle of the plate.
6. Table 3. The evidence table, with numbers to rate
| Observation | What it shows | The number that lets you rate it |
|---|---|---|
| Continental fit and matching fossils | The continents were joined | Glossopteris and glacial deposits in now-tropical Africa (USGS); suggestive, but no rate and no mechanism |
| Magnetic stripes | New crust forms at ridges and moves away, recording reversals | Stripes symmetrical about ridges (USGS); most recent reversal 0.774 million years ago (ICS) |
| Ocean-floor ages | Ocean crust is young and recycled | Oldest floor about 180 million years (USGS) against 1,480 million years for the St. Francois Mountains and about 4,000 million for the oldest continental rocks (Unit 2) |
| GPS | The plates move today; the middle of a plate barely deforms | Missouri 14-15 mm/yr with its plate; across the New Madrid zone 0.29 ± 0.22 mm/yr; Neah Bay 14.6 mm/yr relative to the plate |
| Earthquake locations | Boundaries are where the crust breaks, with rare exceptions | 695 of 2,089 earthquakes in four boundary boxes against 14 in five interior boxes in one month |
Sources: USGS, This Dynamic Earth: Developing the theory (pubs.usgs.gov); USGS, This Dynamic Earth: Magnetic stripes and isotopic clocks (pubs.usgs.gov); USGS, This Dynamic Earth: Understanding plate motions (pubs.usgs.gov); USGS, This Dynamic Earth: Historical perspective (pubs.usgs.gov); USGS event page, M 4.0 – 1 km WNW of Cooter, Missouri (nm60620376) (earthquake.usgs.gov); USGS Did You Feel It? ZIP-code table for nm60620376 (earthquake.usgs.gov); USGS FDSN event web service (count queries for Table 2) (earthquake.usgs.gov); Nevada Geodetic Laboratory, MIDAS velocities, North America frame (geodesy.unr.edu); Nevada Geodetic Laboratory, MIDAS velocities, IGS14 frame (geodesy.unr.edu); Nevada Geodetic Laboratory, MIDAS velocity file README (geodesy.unr.edu); ICS International Chronostratigraphic Chart v2023/09 (CC BY 4.0) (stratigraphy.org)
How Missouri’s Crust Was Assembled: Two Supercontinent Cycles, the Dated Rocks, the Reelfoot Rates, and a Recorded Sand-Box Run
University of Memphis CERI, Northern Embayment Lithosphere Experiment: Tectonic setting; USGS Professional Paper 1866 (2021) as summarized in the St. Francois Mountains reading; University of Arkansas news release on Guccione, Mueller and others (1999); Stein 2012 (USGS NSHM CEUS workshop); Purdue release on Calais and Stein 2009; Nevada Geodetic Laboratory MIDAS velocities; retrieved November 6, 2026; SLS original illustrative sand-box record · SLS original (cites university, USGS public-domain and published research sources; brief quotations only)
Why this reading. Your three-panel model needs dates for each panel, and the sand-box lab needs a data set if you cannot run it. Here is the region’s history as seismologists at the University of Memphis summarize it, the dated rocks that pin each panel, the rates that connect them, and a completed sand-box record built to the lab’s procedure. Where sources give different dates, the difference is shown rather than smoothed over.
1. Two complete cycles (University of Memphis CERI)
The Center for Earthquake Research and Information’s Northern Embayment Lithosphere Experiment page titles the region’s history “Two Complete Wilson Cycles” (a supercontinent assembled and broken apart, twice) and lists the most accepted sequence:
- Granite and rhyolite basement: “The Eastern Granite-Rhyolite Province … underlies the [Mississippi Embayment], suggesting that Proterozoic basement rocks throughout the region consist primarily of granite and granite porphyry.” Rodinia was assembled during the Grenville orogeny, 1.35 to 1.0 billion years ago.
- “Initial opening and subsidence of the Reelfoot rift occurred in Early to Late Cambrian time (ca. 565 Ma) in response to the breakup of the supercontinent Rodinia. … Rifting was not accompanied by major intrusive or volcanic activity.”
- By Middle Cambrian time (ca. 530 Ma) an ocean, the Iapetus, lay to the south, and by Late Cambrian time (ca. 515 Ma) the rift “was part of a stable continental passive margin, covered by a thick deposit of shallow marine carbonates.”
- According to one interpretation, Cambrian faults in the rift were reactivated in the Late Paleozoic (ca. 290 Ma) as the Iapetus Ocean closed in the Alleghanian-Ouachita mountain building; other researchers place the reactivation in the mid-Cretaceous.
- The embayment was raised in the mid-Cretaceous and began to subside in the Late Cretaceous; it “is presently covered with unconsolidated Late Cretaceous and Cenozoic sediments” and “may still be subsiding but at a very low rate.”
Table 1. The dated rocks behind each panel
| Panel | Feature | Measurement | Source |
|---|---|---|---|
| ~1,470 Ma: volcanoes and granite | St. Francois Mountains | Volcanic and plutonic rocks 1.48 to 1.45 billion years old, essentially continuous magmatism for about 30 million years, linked to crustal extension, mantle upwelling and lower-crust melting | USGS Professional Paper 1866 (2021), via the reading on the St. Francois Mountains |
| ~1,470 Ma | The basement under the embayment | Granite and granite porphyry of the Eastern Granite-Rhyolite Province; CERI’s page gives 1.8-1.6 billion years for the province, older than the USGS dates for the exposed St. Francois rocks, because the province is a broad belt of different ages | University of Memphis CERI |
| ~540 Ma: the rift | Reelfoot Rift | Opened ca. 565 Ma; passive margin by ca. 515 Ma, covered by shallow-marine carbonates | University of Memphis CERI |
| ~300 Ma (optional panel) | Ouachita-Alleghanian collision | Rift faults reactivated ca. 290 Ma as the Iapetus Ocean closed | University of Memphis CERI |
| 0 Ma: today | Reelfoot scarp | About 13.7 m (45 ft) of slip in about 2,300 years, about 6 mm/yr | University of Arkansas, on Guccione, Mueller and others 1999 |
| 0 Ma | Reelfoot Lake | Formed by subsidence of 1.5 to 6 m in the 1811-1812 earthquakes | USGS summary of the 1811-1812 sequence (reading on the New Madrid USGS sources) |
| 0 Ma | GPS | Less than 0.2 mm/yr across the zone; Missouri moves about 14.5 mm/yr west with its plate | Calais and Stein 2009 (Purdue release); Nevada Geodetic Laboratory |
| Seconds | 1812 rupture | Magnitude 7.5 on the Reelfoot fault, February 7, 1812 | USGS |
2. The lesson’s rates, checked
- 6 mm/yr for 1 million years: 6 × 1,000,000 mm = 6,000,000 mm = 6 km. For the 290 million years since reactivation: 6 mm × 2.9 × 108 = 1.74 × 109 mm = 1,740 km. The Bootheel shows nothing like that, which is why Stein (2012, USGS workshop) concludes that “lack of significant fault topography” and other evidence imply the recent pulse of activity is only a few thousand years old.
- Missouri at 14.5 mm/yr: 14.5 km per million years; over 565 million years, 8,192.5 km if steady (it was not: plates change speed and direction).
- Time between panels: 1,470 − 565 = 905 million years from the St. Francois volcanoes to the rift; the rift was active for about 565 − 515 = 50 million years.
- The lesson’s illustrative mountain: 5 mm/yr uplift minus 2 mm/yr erosion = 3 mm/yr, or 3 km per million years.
Table 2. Illustrative sand-box record
Constructed teaching data, not measurements of a particular run. Clear box 40 cm long; six alternating layers of sand and flour, each about 5 mm, total 30 mm; cardboard wall pushed 1 cm at a time; side-view photographs at every 2 cm.
| Wall movement (cm) | Maximum height of the pile (mm) | Distance from wall to the first fault (mm) | Faults visible | What the photograph showed |
|---|---|---|---|---|
| 0 | 30 | — | 0 | Flat layers |
| 2 | 37 | 48 | 1 | Layers bulge up next to the wall; one break dips toward the wall; layers on the wall side ride up over the others |
| 4 | 45 | 66 | 2 | A second fault ahead of the first; the first block tilts back toward the wall |
| 6 | 52 | 85 | 3 | A wedge of stacked slices; the top surface now slopes away from the wall like a mountain front |
| 8 | 60 | 104 | 4 | Fourth fault farthest from the wall; the oldest slices nearest the wall are steepest; tapping the box flattens the crest and exposes tilted layers |
Read the record as the crust: each new fault forms ahead of the last, so the pile grows forward as well as up, doubling its height (30 to 60 mm) with 8 cm of shortening. In the Ouachita Mountains the wall is the colliding continent and the stacked slices are the folded and faulted ranges; at the Reelfoot fault, a single thrust in the squeezed old rift raises the scarp. Scale: 8 cm of shortening in about a minute against 8 km at 6 mm/yr, which is 8,000,000 mm / 6 mm per year = 1.33 million years. The model omits the erosion that would remove much of the rise over that time, the heat that lets deep rock fold rather than break, and the earthquakes in which each real fault moves.
Sources: University of Memphis CERI, NELE: Tectonic setting (Two Complete Wilson Cycles) (memphis.edu); USGS Professional Paper 1866, Petrology and geochronology of 1.48 to 1.45 Ga igneous rocks in the St. Francois Mountains terrane (2021) (usgs.gov); University of Arkansas news, Research in New Madrid Seismic Zone points to possible earthquake threat (1999) (news.uark.edu); Stein, A GPS-based view of New Madrid earthquake hazard (USGS NSHM CEUS workshop, 2012) (earthquake.usgs.gov); Purdue University news release, Calais and Stein, New Madrid (March 2009) (purdue.edu); Nevada Geodetic Laboratory, MIDAS velocities, IGS14 frame (geodesy.unr.edu)
The USGS Sources to Summarize: The New Madrid Seismic Zone, the West Coast’s ShakeAlert, and Missouri’s Warning Channels
USGS, Summary of 1811-1812 New Madrid Earthquakes Sequence (October 2, 2019); USGS Fact Sheet 2009-3071, Earthquake Hazard in the New Madrid Seismic Zone Remains a Concern (August 2009); USGS ComCat earthquake catalog query, 2025, 35.5-37.5 N, 88.5-90.5 W; USGS Fact Sheet 2014-3083, version 1.2 (February 2017), ShakeAlert: An Earthquake Early Warning System for the United States West Coast; USGS image description, ShakeAlert Earthquake Early Warning Basics; USGS news release, Entire U.S. West Coast Now Has Access to ShakeAlert (May 4, 2021); NWS, Wireless Emergency Alerts; NWS, NOAA Weather Radio All Hazards, SAME and event-code pages; all retrieved November 5, 2026 · Public domain (works of the United States government)
Why this reading. The unit asks whether a warning could ever outrun the shaking from an earthquake in the New Madrid Seismic Zone, and the performance task says to use USGS pages as your primary sources and to cite no figure without a source and an access date. Here are those sources, quoted, each with its identifier and the date it was retrieved (November 5, 2026), so the explainer can be written from inside the lesson. Read them for three things: what happened in Missouri in 1811–1812 and what USGS says about the hazard now; how the West Coast’s ShakeAlert system works (your working model); and which warning channels already reach Missouri phones and radios.
1. The 1811–1812 New Madrid earthquakes
USGS, “Summary of 1811-1812 New Madrid Earthquakes Sequence” (October 2, 2019). The sequence is named “after the Missouri town that was the largest settlement on the Mississippi River between St. Louis, Missouri and Natchez, Mississippi.” The three principal shocks: “The first principal earthquake, M7.5, occurred at about 2:15 am (local time) in northeast Arkansas on December 16, 1811. The second principal shock, M7.3, occurred in Missouri on January 23, 1812, and the third, M7.5, on February 7, 1812, along the Reelfoot fault in Missouri and Tennessee.”
How far the waves reached: damaging shaking covered an estimated 600,000 square kilometers and the earthquakes were felt over about 5,000,000 square kilometers; “The area of strong shaking associated with these shocks is two to three times as large as that of the 1964 Alaska earthquake.” The human side: “Only one life was lost in falling buildings at New Madrid, but chimneys were toppled and log cabins were thrown down as far distant as Cincinnati, Ohio, St. Louis, Missouri, and in many places in Kentucky, Missouri, and Tennessee.” Few people lived in the region in 1811; the low toll reflects that, not small shaking. The ground itself changed: subsidence formed Reelfoot Lake in Tennessee, where the land dropped 1.5 to 6 meters.
2. The hazard now
USGS Fact Sheet 2009–3071, “Earthquake Hazard in the New Madrid Seismic Zone Remains a Concern” (August 2009): “Many structures in Memphis, Tenn., St. Louis, Mo., and other communities in the central Mississippi River Valley region are vulnerable and at risk from severe ground shaking.” The zone “is the most seismically active area of the United States east of the Rockies.” On the odds: “the USGS estimates the chance of having an earthquake similar to one of the 1811–12 sequence in the next 50 years is about 7 to 10 percent, and the chance of having a magnitude 6 or larger earthquake in 50 years is 25 to 40 percent.” The fact sheet adds that roadways in the Mississippi Valley of Arkansas and Missouri, “such as Interstate 55,” could become impassable.
3. Still active: one year in the USGS earthquake catalog
A query of the USGS earthquake catalog (ComCat, FDSN event service) for 2025, in a box from 35.5°N to 37.5°N and 88.5°W to 90.5°W around the seismic zone, returned 232 earthquakes of magnitude 1.0 or larger, of which 13 were magnitude 2.5 or larger. The largest was magnitude 3.07 on September 6, 2025, located “4 km S of Puxico, Missouri.” Almost all were too small to feel. They matter here for a different reason: every one was detected and located by seismometers already operating in the region. The sensors exist; what Missouri does not have is a public system that turns the first seconds of a large earthquake into an alert.
4. ShakeAlert: how a working system does it, on the West Coast
USGS Fact Sheet 2014–3083, version 1.2 (February 2017), is titled “ShakeAlert—An Earthquake Early Warning System for the United States West Coast,” and says the system “is designed to cover the West Coast States of California, Oregon, and Washington.” Why it is possible: “Early warning is possible because information can be sent through communication systems virtually instantaneously, whereas seismic waves travel through the shallow Earth at speeds ranging from 0.5 to 3 miles per second.” The waves: “The faster but weaker P waves trip nearby sensors, causing alert signals to be sent out, giving people and automated electronic systems some time (seconds to minutes) to take protective actions before the arrival of the slower but stronger S waves and surface waves.” The fact sheet’s diagram labels the network “Sensors positioned about 6-12 miles apart.”
“The P wave, which travels fastest, trips sensors placed in the landscape, transmitting data to a ShakeAlert® processing center where the location, size, and estimated shaking of the earthquake are determined. If the earthquake fits the right profile a ShakeAlert® message is issued by the USGS. The message is picked up by ShakeAlert® partners which could be used to produce an alert to notify people to take a protective action such as Drop, Cover, and Hold On …” (USGS image description, “ShakeAlert Earthquake Early Warning Basics”)
This is the chain your diagram must show: P wave → sensors → data to a processing center → location, size and expected shaking computed → USGS issues a message → partners turn it into alerts on phones and into automated actions. USGS news release, May 4, 2021: public alerting “began with California in 2019 and expanded to Oregon in March 2021,” and reached Washington that day, making ShakeAlert “available to more than 50 million people in California, Oregon and Washington”; the same release puts the number of people in the United States at risk of potentially damaging shaking at more than 143 million. Alerts reach phones through “FEMA’s Wireless Emergency Alert system, third-party phone apps, and other technologies.”
What seconds buy. The 2014 fact sheet lists human responses, “Public–Citizens, including schoolchildren, Drop, Cover, and Hold On; turn off stoves; safely stop vehicles,” and automated ones, “Transportation–Automatically slow or stop trains to prevent derailment” and “Businesses–Open elevator doors, shut down production lines, secure chemicals.”
5. The warning channels that already reach Missouri
Wireless Emergency Alerts. National Weather Service: “Wireless Emergency Alerts (WEA) are emergency messages sent by authorized government alerting authorities through your mobile carrier.” “No signup is required! Alerts are sent automatically to WEA-capable phones during an emergency.” Messages are “up to 360 characters on 4G LTE networks and beyond,” and the NWS sends them for warnings including tornadoes. A Missouri phone that buzzes for a tornado warning is using the same delivery channel the West Coast uses for ShakeAlert.
NOAA Weather Radio All Hazards. “A nationwide network of radio stations broadcasting continuous weather information directly from the nearest National Weather Service office,” which also broadcasts warnings for “natural (such as earthquakes or avalanches)” hazards. It has “more than 1000 transmitters, covering all 50 states,” on seven frequencies from 162.400 to 162.550 MHz, and reception “typically extends in about a 40 mile radius from the transmitter.” The NWS list of event codes includes “Earthquake Warning” (code EQW) as operational. Outdoor warning sirens, run by cities and counties, are the third channel; they carry no text, only a sound that means “go find out.”
6. Figures you may cite, with their sources
| Figure | Value | Source |
|---|---|---|
| Magnitudes of the three principal 1811–1812 shocks | M7.5 (Dec. 16, 1811), M7.3 (Jan. 23, 1812), M7.5 (Feb. 7, 1812) | USGS summary, Oct. 2, 2019 |
| Area of damaging shaking; area where felt | 600,000 km²; 5,000,000 km² | USGS summary |
| Chance in 50 years of a repeat of an 1811–1812 earthquake; of M6 or larger | 7 to 10 percent; 25 to 40 percent | USGS FS 2009–3071 |
| Earthquakes in the zone in 2025, M1.0+ and M2.5+ | 232 and 13 (largest M3.07, near Puxico, Mo.) | USGS ComCat query, retrieved Nov. 5, 2026 |
| Speed of shaking waves in the shallow Earth | “0.5 to 3 miles per second” (about 0.8 to 4.8 km/s) | USGS FS 2014–3083 |
| ShakeAlert coverage | California (2019), Oregon (March 2021), Washington (May 4, 2021); more than 50 million people | USGS news release, May 4, 2021 |
| WEA message length | up to 360 characters on 4G LTE and later | NWS WEA page |
| NOAA Weather Radio | more than 1000 transmitters; 162.400–162.550 MHz; about a 40-mile reception radius | NWS NOAA Weather Radio pages |
Not in these sources, and therefore not citable from them: any date when earthquake early warning might come to Missouri, the magnitude threshold for a public alert, the number of sensors in the network, and the typical delivery time to a phone. If you find such a figure on another USGS or NWS page, cite that page and the date you read it, because these systems change.
Your turn
- Write a four-sentence summary of how ShakeAlert works using only section 4, then mark each sentence with the quotation it rests on.
- Using sections 1 to 3, write two sentences explaining why an earthquake early-warning system would be useful in Missouri even though large New Madrid earthquakes are rare.
- Section 5 lists three channels. For each, name the kind of wave that carries the message the last part of the way to a person (radio wave, sound wave or light from a screen).
Sources: USGS, Summary of 1811-1812 New Madrid Earthquakes Sequence (usgs.gov); USGS Fact Sheet 2009-3071, Earthquake Hazard in the New Madrid Seismic Zone Remains a Concern (pubs.usgs.gov); USGS earthquake catalog (FDSN event service) count, 2025, M1.0+, New Madrid region box (earthquake.usgs.gov); USGS Fact Sheet 2014-3083, ShakeAlert: An Earthquake Early Warning System for the United States West Coast (pubs.usgs.gov); USGS, ShakeAlert Earthquake Early Warning Basics (image description) (usgs.gov); USGS news release, Entire U.S. West Coast Now Has Access to ShakeAlert Earthquake Early Warning (May 4, 2021) (usgs.gov); National Weather Service, Wireless Emergency Alerts (weather.gov); National Weather Service, NOAA Weather Radio All Hazards (weather.gov); National Weather Service, NWR Specific Area Message Encoding (SAME) (weather.gov); National Weather Service, NWR event codes (weather.gov)