
calculate a slip rate from a dated offset, fit the Gutenberg-Richter relation to earthquake counts, and use both to argue for or against claims about earthquakes on a major fault.
Phase 2: Investigate
Do Now. A fence built across the fault in 1900 is found offset by 4.2 m of steady creep by 2020. Without a calculator, estimate the creep rate in mm per year. Then check it.
What is the fence's average creep rate?
A human impact first. Earthquakes are not only data. Large earthquakes on major faults have destroyed cities and taken thousands of lives. Every number below stands for shaking that people live through. We study the numbers so that communities, and engineers like our client’s, can prepare.
Slip rate from the offset channel. Geologists dug trenches across the fault at the stream channel and dated buried channel deposits with radiocarbon. The channel has been offset about 130 m in roughly 3,700 years. The average slip rate is 130 m / 3,700 yr = 0.0351 m/yr, about 35 mm per year. That matches the rate GPS measures across the fault today, and it is about 70% of the roughly 50 mm per year of total motion between the two plates; other faults nearby take up the rest. Two independent methods, one over thousands of years and one over a few years, agree: strong evidence.
Steady motion, sudden slip. The plates move steadily, but on a locked stretch of the fault the rocks do not creep; they stick. Strain builds in the rocks on both sides until the fault ruptures in a large earthquake. In a large earthquake on a fault like this, a rupture can run for hundreds of kilometers and offset streams by several meters at once. If each large earthquake slips about 9 m, the plates would need 9 / 0.035 ≈ 260 years to store that much slip. Trenches often show that large earthquakes came more often, with less slip each time. Real faults are less regular than a simple model, which is why scientists forecast probabilities, not dates.
Many small, few large: Gutenberg-Richter. The table below gives earthquake counts for a 25-year regional catalog around a major strike-slip fault system (sample data, modeled on the pattern in real regional catalogs). Each column counts the events at or above a magnitude.
| Magnitude ≥ | 3.0 | 3.5 | 4.0 | 4.5 | 5.0 | 5.5 | 6.0 | 6.5 | 7.0 |
|---|---|---|---|---|---|---|---|---|---|
| Count in 25 years (sample) | 11,823 | 3,867 | 1,170 | 329 | 91 | 29 | 10 | 5 | 2 |
| log10(count) |
Each whole step in magnitude cuts the count by about ten: 11,823 → 1,170 → 91 → 10. Plot log10(count) against magnitude and the points fall close to a line. A least-squares fit to magnitudes 3.0–6.0 gives log10N ≈ 7.21 − 1.04M, where N is the 25-year count. The slope, called the b-value, is about 1, as it is in most regions.
Worked example 1: extending the line. For M = 7: 7.21 − 1.04 × 7 = −0.07, and 10−0.07 ≈ 0.85. The line predicts less than one M7+ earthquake in 25 years, roughly one every 30 years. The sample catalog recorded two. With counts that small, chance matters, so the fit is a forecast of an average rate, not a schedule. The upper end of the line is its weakest part, because it rests on the fewest events.
Worked example 2: why many small quakes do not prevent a big one. Each magnitude step releases about 32 times more energy (101.5). An M7 releases 101.5 × 2 = 1,000 times the energy of an M5. It would take about a thousand M5 earthquakes to release the energy of one M7, and the sample region had only 91 events of M5 or more in 25 years.
Forecasts, not predictions. Official forecasts combine slip rates, fault maps and earthquake statistics. They are stated as probabilities, for example “a 60% chance of at least one magnitude 6.7 or larger earthquake in this region in the next 30 years” (a sample statement). Your argument today will test popular claims against numbers like these.
Work with the sample catalog. Show each step.
Argue it out.
Exchange your rebuttal with a partner. Respond in writing (or on a recording) to their strongest point. Then write two sentences rating the reliability of each kind of evidence you used: the dated offset, the GPS rate and the 25-year catalog. Name one weakness of each.
The fault moves about 35 mm a year but releases it in jumps. Next phase we model how this boundary was built.
How confident are you that you can I can calculate slip rates and use earthquake statistics to argue about earthquake claims.
Exit ticket: the last large earthquake on the crossing's stretch of fault was 160 years ago (sample site history). At 35 mm/yr, how much slip has been stored since then? Show the calculation.
Timing: Poly ran this argumentation lab over two 50-minute periods. Day 1: slip rates and the CODAP fit; day 2: the claims, rebuttals and peer responses.
Data note: the 25-year counts are presented as sample data modeled on a real regional catalog; they keep a realistic b-value near 1. The 130 m in about 3,700 years offset reflects real trench studies of a stream offset by a major strike-slip fault. Keep the fault and the region generic.
Listen for: “overdue means it will happen soon”; treating a rate as a schedule; confusing magnitude steps with factors of 2.
Support / stretch: give a log table or a calculator sequence for 10x. Stretch: compute the expected 25-year count of M ≥ 6.7 events from the line and turn it into a 30-year probability with P = 1 − e−rate × time.
Answer note (exit ticket): 0.035 × 160 = 5.6 m. Students use this number in 3.07.