Static decay predictors are incomplete. The atmosphere varies 3× across the solar cycle; your asset lives through both ends. Reserves priced against any single percentile leave the other two unpriced. The methodology delivers all three. The M2 historical-prediction gauntlet detected the static-Fbaseline cycle-invariance limit empirically — disclosed below, not hidden.
Definitions: Fbaseline(h) is the SWADI substrate-weighted atmospheric density per altitude shell (decision #25); βrel(h) is the per-shell storm-response coefficient (decision #28); F10.7(t) and ap(t) are read at every integration step from the F2-validated GFZ-Potsdam 95-year archive (1932–2026, SHA-pinned). Three archetype windows are replayed forward from the asset's decommissioning epoch: P10 = 2008–2010 cold-sun deep solar minimum; P50 = 1976–2026 50-year nominal; P90 = 1979–1981 Cycle 21 hot-sun peak. Integrator is clamped to a 1-day max timestep so it never steps over a storm.
Deliverable: an audit JSON pinned to the snapshot SHA, plus a rendered PDF. Same form as the public artifacts in the audit charter.
| Orbit | P10 (yr) | P50 (yr) | P90 (yr) | P10/P90 ratio | FCC 25-yr? |
|---|---|---|---|---|---|
| ISS-class 410 km A/m=0.006 | 2.05 | 1.67 | 0.65 | 3.17× | yes |
| Low-LEO 450 km A/m=0.012 | 2.14 | 1.73 | 0.67 | 3.18× | yes |
| Mid-LEO 600 km A/m=0.010 | 29.65 | 13.80 | 9.71 | 3.05× | yes |
| High-LEO 800 km A/m=0.007 | >100 | >100 | >100 | 1.00× | no |
The envelope is a forward-prediction tool. As a sanity check we replay each canonical historical reentry against its own decommissioning-epoch orbit parameters and ask: does the actual lifetime fall inside the methodology's [P90, P10] envelope? Honest result, with the static-Fbaseline cycle-invariance limit visible.
| Asset | alt (km) | A/m | observed | P90 | P50 | P10 | result |
|---|---|---|---|---|---|---|---|
| Skylab | 433 | 0.0036 | 5.42 yr | 1.71 | 2.06 | 5.01 | above P10 by 0.41 yr |
| UARS | 575 | 0.0086 | 5.78 yr | 7.75 | 17.38 | 23.19 | below P90 by 1.97 yr |
| ROSAT | 565 | 0.0070 | 12.69 yr | 8.01 | 13.96 | 24.37 | inside envelope |
Replay starts from each asset's epoch_eom_iso (Skylab 1974, ROSAT 1999, UARS 2005) and integrates forward against the three archetype windows. Skylab's actual atmosphere was quieter than even the cold-sun archetype's; UARS reentered during a hot-sun phase that exceeded the methodology's hot-sun replay. ROSAT, mid-cycle, lands inside the envelope. Decision #29 articulates the v0.4.0.7 cycle-stratified Fbaseline fix; until then the limit is disclosed empirically.
Engagements are direct. Email with the asset, fleet, or jurisdiction you want assessed and a sample lands in your inbox within one day.
Engage by email →