What comes next
The current analysis favors a globe in every fully checked comparison. The next work has two purposes: complete the six unfinished comparisons, then establish more precisely what the measurements reveal and how reliably they predict new data. The verified status gives the published totals; the study index explains the completed work.
An IMU (inertial measurement unit) records turning and acceleration. The archive analysis uses its recorded increments together with GPS receiver positions.
First: resolve the six unfinished stretches
| Planned work | Question it answers | Required result |
|---|---|---|
| Timing-boundary diagnostic, partial | Is the stopping check rejecting a settled fit because time alignment crosses an IMU packet boundary? | Validate accounting before continuation; both sides must pass fresh checks with all nuisance directions included, while completed controls remain stable. |
| Compare each primary fit with its repeat | Do separate starts reach sufficiently similar costs and reconstructed tracks? | Every required pair must agree; a passing local diagnostic alone cannot resolve a stretch. |
| Focused timing profiles where disagreement remains | Are competing time alignments genuinely different fitting solutions? | Reoptimize the same permitted calibration parameters across a separately frozen, bounded profile. |
The measurements, selected intervals and physical error bounds stay fixed during numerical recovery. Earlier failures remain in the record. Each new numerical study gets its own finite limits before it starts.
Then: strengthen the interpretation
| Analysis | What it adds |
|---|---|
| Separate gyro, acceleration and coordinate contributions | Isolates what the central gyro observations add, using matched tests of gyro input, acceleration input, horizontal-only receiver constraints and coordinate/height assumptions. |
| Residual correlation and calibrated uncertainty | Checks how errors persist over time and how that affects confidence in a model comparison. |
| Prediction on unused observations | Fits instrument parameters on one part of a recording and tests the resulting prediction on a separate part. |
| Shared-flight and instrument grouping | Accounts for simultaneous recorders and repeated stretches from the same flight instead of treating them as independent replications. |
| Additional independent-flight checks | Tests whether the finding transfers across distinct recordings and installations. |
Existing timing, calibration and prediction studies provide development evidence; they do not complete every check in this table. Exact device documentation remains useful supporting work. Its absence does not suspend analysis of survey-grade data.
Measure the rotation field freely
The present model comparison tests a specified background angular rate against zero. The next rotation-field study will estimate the retained field's size and direction from the observations, with aircraft motion, travel over the surface and instrument errors handled jointly.
It will ask whether one geographic pattern explains different headings, latitudes, altitudes, flights and instruments. Longer records may help separate a geographic field from an offset fixed to an instrument. First, the available track geometry must show that those quantities can be separated. An unconstrained amount removed onboard cannot be recovered independently from the field that remains in the log.
Solar versus sidereal rate
A 24-hour rate is 15.0000 degrees/hour; a sidereal-day rate is about 15.0411 degrees/hour. The difference is only 0.0411 degrees/hour, or about 0.27%. The current short-stretch comparison fixes the background rate, so it does not measure this distinction.
A dedicated test would use longer supported records, estimate the rate freely, and show that timing, gain, instrument offset and shared processing cannot account for the difference. Adding minutes from overlapping recorders does not create independent observing time. If the precision is insufficient, the result must report that limit rather than assign a day length.
Measuring a field and its rate describes the observed angular motion. Identifying its ultimate physical cause requires predictions that distinguish the proposed explanations; this analysis does not infer a cosmology from the signal alone.
Separate ground experiment
The stationary experiment will test consumer IMUs on solid supports with repeated 180° reversals. It needs physical noise, temperature and alignment checks before geographic pooling or detection claims. The NASA results do not establish consumer-device sensitivity.
Study index · Measurement method · Detailed technical roadmap.