Compiled: 31 October 2026 Prepared for: Toby Mikle Source batch: 29 New Hypotheses — Unconventional Sources (31 Oct 2026), entries H15, H12, H18 Status: three tests run; one partly supported, one conditionally supported, one refuted
Three hypotheses from the newest bundle were selected because each makes a number that can be checked against published measurements without a new instrument. Each was turned into a model with explicit parameters, run as a Monte Carlo over the published ranges, and judged against its own falsifier. The code, the parameter registry and the results are in this folder.
The short answer: the Martian claim survives, the quartz claim survives only in the strongest ruptures, and the Titan claim does not survive at all.
| Test | Hypothesis | Decisive number | Verdict |
|---|---|---|---|
| 01 | H15 Quartz Lightning Rod | 0.57 MV/m median; 17 MPa needed | Conditional |
| 02 | H12 Mars Dust Battery | 4.0 × 10⁻⁷ W/m² median | Partly supported |
| 03 | H18 Titan Dune Capacitor | 0.05 s retention | Refuted |
Piezoelectric quartz veins in fault zones generate charge during rupture. The charge concentrates at vein tips and triggers corona discharges in the surrounding air, producing earthquake lights.
If a rupture drops the shear stress on a quartz vein by Δτ, the piezoelectric polarization is P = d26 Δτ. The bound charge on the vein faces is σ_b = P, and the field in the air just outside a charged face is the sheet field
E_air = g · σ_b / (2 ε₀) = g · d26 · Δτ / (2 ε₀),where g is a geometric concentration factor (g = 1 for a flat face, up to ~3 for a sharp outcrop). The hypothesis is met if E_air reaches the breakdown field of air, 3 MV/m.
| Quantity | Value used | Source |
|---|---|---|
| d26, quartz | 3.1 pC/N (range 2.3–4.6) | Bechmann (1958); class-32 relation d26 = −2d11 |
| Stress drop | log-normal, median 3 MPa, σ_ln = 0.8 | Allmann & Shearer (2009) |
| Breakdown, air at 1 atm | 3.0 MV/m (2.5–3.5) | standard |
| Geometry factor g | 1 (flat) to 3 (sharp outcrop) | bounding assumption |
The d26 range carries a real ambiguity: the class-32 piezoelectric matrix gives |d26| = 2|d11| = 4.6 pC/N, while thickness-shear values near 3.1 pC/N are also in use. The test runs the whole 2.3–4.6 pC/N range rather than choosing.
The mechanism is not generic. For the median earthquake it fails by a factor of five. It works only where a rupture is strong (Δτ ≳ 6–17 MPa, depending on outcrop sharpness) and the vein tip is sharp. That combination is rare but not unphysical: the stress-drop tail above 10 MPa is populated, and quartz veins do outcrop as ridges and slickensided faces.
Two things push the result down, and both are conservative in the right direction. First, the charge is bound, not mobile: unlike a lightning rod, a quartz vein cannot redistribute charge to a tip, so g cannot be arbitrarily large. Second, the surrounding rock and the finite vein thickness reduce the external field relative to the sheet estimate. The g = 3 case should therefore be read as an upper bound.
A field or corona measurement at a real quartz outcrop during a M ≳ 5 rupture, or a laboratory shear test on a quartz vein analog at 10–30 MPa with a field probe at the tip. The hypothesis predicts a threshold, so a null result at 30 MPa would kill it cleanly.
Scope note: this test checks the mechanism, not the hypothesis’s statistical half — the claim that light flashes correlate with quartz-vein abundance and slip rate. That needs an earthquake-light catalog, which is not machine-readable here.
A grounded conductor in a Martian dust devil can harvest triboelectric charge. The current scales with saltation flux and grain size, producing >10 nW m⁻² during a vortex passage.
Treat the saltation layer as a weakly conducting slab of height h with conductivity σ and vertical field E. Charge is supplied by the grains at j_charge = χΦ (charge-to-mass ratio times mass flux) and carried by the plasma at j_cond = σE. In steady state the smaller current limits what a grounded plate can draw:
j = min(χ Φ, σ E), P/A = η · j · E · h.Kok & Renno (2009) showed that the field in Martian saltation self-limits at 15–20 kV/m: field ionisation raises the conductivity, which bleeds the charge away. That number, not the Paschen limit of 43 kV/m, is the ceiling. The hypothesis is met if P/A > 10 nW/m² (and its own falsifier asks for >1 nA).
| Quantity | Value used | Source |
|---|---|---|
| Field E | 15–25 kV/m | Kok & Renno (2009) |
| Conductivity σ | 10⁻¹³–10⁻¹⁰ S/m, nominal 10⁻¹² | Kok & Renno (2009); Farrell et al. (2021) |
| Layer height h | 0.1–0.5 m (z_cat = 30 cm) | Kok & Renno (2009) |
| Charge-to-mass χ | 10⁻⁶–10⁻⁴ C/kg | laboratory saltation range |
| Mass flux Φ | 10⁻⁶–10⁻³ kg m⁻² s⁻¹ | saltation range |
| Load fraction η | 0.10–0.40 | matched load |
The power-density half of the claim passes with a wide margin. The current half does not, at any benchtop scale: the hypothesis’s own 1 nA detection threshold requires a collector of several square metres. That is a real engineering constraint, not a technicality — a 1 cm² plate sees picoamps, which is at the edge of low-leakage electrometry.
There is also a physical tension inside the hypothesis. It says the current “scales with saltation flux”, but the field self-limits because the conductivity rises with E. In the high-flux regime the current is set by the plasma, not by the sand; the scaling flattens. The model reproduces that: above χΦ ≈ σE the current stops growing.
A Mars-pressure saltation experiment that measures χ for the relevant grain size, and a field measurement in a Martian dust devil. The power margin is large enough that the remaining question is collection geometry, not physics.
Scope note: this test checks the energy budget, not the claim that “the discharge rate, not the dust density, is the controlling variable”. Testing that needs the time-resolved current during a vortex passage, which no published Mars dataset has.
Titan’s dune sands are charged to 10⁵–10⁶ V m⁻¹ and retain charge for months. The dunes act as giant capacitors.
A bed of permittivity ε_r and conductivity σ relaxes free charge with the dielectric time τ = ε_r ε₀ / σ. The field decays as E(t) = E₀ exp(−t/τ). The hypothesis needs τ ≳ 1 month = 2.6 × 10⁶ s at 10⁵–10⁶ V/m.
| Quantity | Value used | Source |
|---|---|---|
| Surface conductivity | 4 × 10⁻¹⁰ S/m (10⁻¹¹–10⁻⁹) | Grard et al. (2006), Huygens PWA |
| Surface permittivity | ~2 (1.8–2.6) | Huygens PWA |
| Lower-atmosphere conductivity | 10⁻¹⁴ S/m (10⁻¹⁶–10⁻¹²) | Lorenz (2021) |
| Claimed retention | 1–12 months | H18 |
There is no version of the measured Titan surface that holds a field for months. The nuance worth stating is that the Huygens value is itself contested: Lorenz (2021) argues the high conductivity and its sudden drop 11 minutes after landing were caused by the probe’s radioisotope heater units ionising the air around it. That caveat does not rescue the hypothesis. If the surface is as insulating as clean ice, the atmosphere above it still relaxes charge in minutes; to reach months the atmosphere would have to be below 10⁻¹⁸ S/m, which is far below any measured planetary atmosphere.
The capacitor analogy fails for a second reason: the dunes are not being continuously charged. Titan’s surface winds are ~1 m/s, near or below the saltation threshold, so there is no sustained charging current to replace what leaks away. A dune that both leaks in seconds and is not being charged cannot accumulate a seasonal charge.
A laboratory measurement of the conductivity of Titan dune analog (tholin-coated organic sand) at 94 K and low pressure. If it came out below 10⁻¹⁸ S/m the claim would reopen; nothing measured so far is within seven orders of that.
All three hypotheses are about the same competition: charge separation versus relaxation. The differences are in the medium.
| Setting | Conductivity (S/m) | τ = εε₀/σ | Is the field driven? | Outcome |
|---|---|---|---|---|
| Quartz vein | ~10⁻¹⁶ (dry insulator) | ~5 days | yes, during rupture | charge survives, but is too small |
| Mars saltation layer | ~10⁻¹² | ~9 s | yes, continuously | steady field at 15–20 kV/m |
| Titan dune | 4 × 10⁻¹⁰ | 0.05 s | no | charge cannot accumulate |
The Mars case passes because the charging is continuous: a 9-second relaxation time is irrelevant when grains are delivering charge every second. The Titan case fails because the claim is about storage: a 0.05-second relaxation time destroys any seasonal capacitor, and the wind is too weak to replenish it. The quartz case sits in between: the medium holds charge for years, but the piezoelectric charge density at ordinary stresses is simply too small to break down air.
That is the useful generalisation from this batch: a charge-separation hypothesis needs both a long relaxation time and a charge source matched to the observation time. H18 has neither; H15 has the first but not the second at typical stress; H12 has both because it never needs to store anything.
| File | What it is |
|---|---|
tests/01-quartz-vein/code/model.py |
H15 model: analytic field, Monte Carlo, deterministic table |
tests/02-mars-saltation/code/model.py |
H12 model: current-limited harvest, Monte Carlo |
tests/03-titan-dune/code/model.py |
H18 model: relaxation time, required conductivity |
tests/*/out/results.json |
every number quoted above |
shared/params.json |
parameter registry with source keys |
shared/make_figures.py |
rebuilds the three figures |
figures/ |
fig01 quartz, fig02 mars, fig03 titan |
literature/ |
the source papers used |