Local Level Lab

Airborne measurements. Open methods.

Smaller boxes help only slightly

All 192 calculations pass their correctness checks. Every predeclared region is included, and the known generated trajectories remain inside their prediction ranges. The calculation preserves the original physical assumptions.

The ranges narrow by at most 1.26% compared with the parent calculation. Four of six candidate/processing groups meet the unchanged target; both globe groups with the specified background field still miss it. Their east and height ranges remain too wide.

This tests a generated 7.2-second trajectory, not an aircraft recording. It shows that splitting the two tested mounting-angle coordinates is insufficient. It does not change the 53-stretch findings or qualify full-flight prediction errors. No new aircraft fits or native trajectory calculations ran.

Complete comparison · Next step · Roadmap.