T03 — Tides only tip critical faults
Supplied working report; original verdicts not validated.
Review status: Held for correction. This is an archival source note, not a passed proof.
README.md
# T03 — Tides only tip critical faults
**Hypothesis.** Vol I, theory 11, *Tides Only Tip Critical Faults* — "Earthquakes
respond to tidal stress only when fault pore pressure is within about ten per cent
of failure, so the appearance of tidal sensitivity is a precursor rather than a
permanent background property."
Source: <https://incoprea.com/research-library/original-theories-20261030/theories/tidal-triggering-pore-pressure.html>
**Verdict: not supported.** No catalogue shows phase locking that survives
accounting for aftershock clustering, and the pore-pressure contrast the
hypothesis requires — locking in high-pressure settings, none elsewhere — does
not appear. The test is negative, not refuting: a phase-only test is a blunt
instrument for a stress-state claim.
## Data
USGS ComCat FDSN event web service, retrieved 30 October 2026, one request per
calendar year (the service caps responses at 20,000 events):
| Catalogue | Box (lat, lon) | Magnitude | Period | Events |
| --- | --- | --- | --- | --- |
| Oklahoma | 34.5–37.2 N, 103–94.4 W | ≥ 2.5 | 2010–2025 | 9,722 |
| The Geysers | 38.7–39.0 N, 122.9–122.6 W | ≥ 1.2 | 2000–2025 | 51,396 |
| Iceland | 63–67 N, 25–13 W | ≥ 3.0 | 1976–2025 | 1,006 |
| Japan | 30–46 N, 128–146 E | ≥ 4.5 | 1976–2025 | 16,220 |
| Australia | 40–10 S, 112–154 E | ≥ 3.5 | 1976–2025 | 3,350 |
| Global | whole globe | ≥ 5.5 | 1976–2025 | 23,730 |
Oklahoma and The Geysers are the high-pore-pressure end (wastewater injection;
geothermal reservoir), Australia is the low end (stable continental interior).
## Method
For each event the phase of the lunar semidiurnal tide (M2) at the epicentre is
computed from the Moon's right ascension (Meeus low-precision lunar theory) and
Greenwich mean sidereal time:
`H = GMST + longitude − α_moon`, `phase = 2H mod 360°`
A uniform distribution of phase means no tidal control. Schuster's test statistic
is `R = |Σ e^{iφ}|`, with `p = exp(−R²/n)`. Because aftershocks are not
independent, the same test is repeated after thinning each catalogue to at most
one event per 30 days (largest event kept), which removes sequence dependence.
## Results
Raw catalogues:
| Catalogue | n | R/n | Schuster p |
| --- | --- | --- | --- |
| The Geysers | 51,396 | 0.017 | 2.5e−7 |
| Japan | 16,220 | 0.018 | 0.006 |
| Oklahoma | 9,722 | 0.013 | 0.18 |
| Iceland | 1,006 | 0.033 | 0.33 |
| Global | 23,730 | 0.008 | 0.25 |
| Australia | 3,350 | 0.006 | 0.89 |
After thinning to one event per 30 days, nothing survives:
| Catalogue | n | R/n | Schuster p |
| --- | --- | --- | --- |
| Iceland | 208 | 0.127 | 0.034 |
| The Geysers | 317 | 0.034 | 0.69 |
| Japan | 609 | 0.014 | 0.89 |
| Oklahoma | 194 | 0.039 | 0.75 |
| Global | 609 | 0.028 | 0.62 |
| Australia | 585 | 0.039 | 0.41 |
Bonferroni threshold for six tests is p < 0.0083; nothing reaches it.
**P1 — locking before moderate induced events, dropping after: not supported.**
With moderate Oklahoma events (M ≥ 4) at least 120 days apart, the 60-day pre and
post windows give R/n = 0.027 (p = 7e−10) before and 0.012 (p = 0.010) after. The
direction matches the prediction, but both windows are dominated by aftershocks of
the very events that define them, so the numbers are not interpretable as a
precursor signal.
**P2 — effect absent away from critical pore pressure: not supported.** Oklahoma,
the clearest high-pore-pressure induced sequence, shows no more locking than
Australia, the stable continental interior (p = 0.75 versus 0.41 after thinning).
The Geysers' apparent signal is an artefact of clustering.
**P3 — a sharp transition within about ten per cent of failure: not tested.**
No pore-pressure proxy in the catalogue can place a fault within a stated
fraction of failure stress, so this prediction is not identifiable with these
data.
## What this does and does not show
It shows that a simple phase test applied to real catalogues produces apparent
tidal triggering that dissolves once aftershock dependence is removed; any claim
of tidal sensitivity from catalogue phase statistics has to survive that check.
It does not refute tidal triggering: the expected effect is about one per cent,
fault orientation is ignored here, and a proper test projects the tidal stress
tensor onto focal mechanisms. The hypothesis's specific pore-pressure contrast
was not confirmed.
## Figures
- `figures/fig1_phase_concentration.svg` — R/n by catalogue, raw.
- `figures/fig2_phase_histogram.svg` — Oklahoma phase histogram.
## Reproduce
```bash
cd work
PYTHONPATH=. [local provenance path] \
t03_tidal_triggering.py
```