Methodology

1. How this is built

The model has three layers:

The split is deliberate: the prose of this site contains not one hand-typed number. If a value looked stretched, it is enough to change the corresponding parameter in the table in chapter 6 and watch how much the result depends on it — the calculator on the home page does exactly that, live. The conclusions deliberately rest on quantities that would have to be wrong by several orders of magnitude to change anything.

2. The contrail formation criterion

The slope of the line along which exhaust mixes with the surrounding air:

The threshold temperature in saturated air (Schumann's analytical approximation to the exact solution of the criterion):

Saturation vapour pressures use the Magnus–Tetens formula, separately over supercooled water and over ice. The gap between those two curves is the region in which a contrail does not evaporate but grows.

3. The programme's mass balance

The starting point is generous to the theory: we take a real, peer-reviewed costing of geoengineering (Smith and Wagner, 3.0 Mt of SO₂ a year in year fifteen) and ask how much of it would be needed for the same effect if dispersed from cruising altitude.

Three places where this estimate is too low — that is, favourable to the theory:

The fuel cost of carrying the payload uses a rule of thumb: about 3.5 % of the mass of the extra payload is burned per flight hour, at an average sector length of 2.2 hours. That parameter is approximate; halving it halves the result and moves no conclusion.

4. The human balance — staffing ratios

We count people, not shifts and not flights. The ratios are taken from the low end of the industry ranges.

Order-of-magnitude check: 29,039 aircraft × 9 pilots gives about 261,351 pilots worldwide. Boeing forecasts demand for 660 thousand new pilots and 710 thousand technicians over twenty years^[Boeing], which at normal turnover corresponds to a base of the same order. The ratios are therefore not inflated.

Deliberately not counted: cabin crew (a separate variant), air traffic controllers, customs and border staff, insurers, auditors, makers of tanks and pumps, tanker drivers, medical staff. Every one of those groups raises the figure.

5. The secrecy decay model

The value p = 4.09·10⁻⁶ is the lowest of Grimes's three calibrations, that is, an assumption that the participants in this programme are more discreet than NSA staff, the Tuskegee doctors and FBI analysts.

The second channel: moral defection

The Grimes model describes accidental leaks, or leaks of conscience towards strangers. Here there is an extra factor that no calibration case had: the participant is harming their own family. We treat it as an independent process with an annual per-person probability q, where the number of defections per year is N·q and the chance of no defection over t years is (1−q)^N·t. We give q over a range from 1:1,000 to 1:1,000,000 a year, because it cannot be determined empirically — there is no precedent for a programme in which participants poison their own children.

6. The full list of parameters

Everything that goes into the model, with its unit and source — what each quantity means is explained in the chapters above. Entries marked assumption do not come from the literature and are stated openly.