Exploratory regional extension to the Uruana study, 3 November 2026

Before computing new geographic transplant results: use every previously eligible sector (14 complete 2011 pairs and eligible grid-origin 2008 observations). Add a common uniform northward slope to both years, then allow separate uniform shifts, reporting open feasibility intervals. The v7 gradient-window CSV, including the native results and controls, was inspected before expanding the analysis. No independent discovery is claimed.

Next transplant each sector's own scalar residuals and footprint using the published v7 code for ellipsoidal transport and IGRF synthesis. Keep original flight heights, orientation, 14 target/northern pairs, bands, channels, five widths and epochs. Only the location of each background varies. A match requires every observed sign criterion in both epochs. First use a global 2-degree cell-center grid, then 1-degree cell centers if computationally feasible. Score the less restrictive two-survey condition without requiring a specific dipole sign. Report every sector with any matches, global maximum signed margin by sector, inside/outside 28 microtesla frequencies as descriptive grid areas, plus native and independently computed checks. A grid zero is finite-resolution and never proves continuous geographic impossibility. No counts of occurrences of Uruana-like geology are inferred.

Sensitivity specified after seeing the 2-degree main result: at each tested field, relax one 2011 line and/or one 2008 corridor width. Remove whichever complete flight line and/or width is most favorable to each sector at that location, keeping all other constraints and both 2008 channels. Count matches separately. Compare the best two rivals under the full criterion and clarify the dependence on the all-lines rule. This is a post hoc sensitivity, not the headline test.
After seeing the 1-degree grid: resolve the strongest rival and all other sector maxima in two local stages, +/-2 degrees in 0.1 degree steps then +/-0.15 in 0.01 degree steps around the best refined center. This tests near-cell misses around the winning 1-degree site only. Cross-check the refined worst-constraint slope with independent ppigrf at all 23 solution sites. Check native slopes against the older per-line and per-band tables for every sector.
