Value chain · CLM-0908
In plain terms
There are two different ways the scheme can fail and they are usually confused. One is that the structure has moved so far the satellite cannot tell which wave crest it is looking at. The other is that the values worked out between sensors are not accurate enough. The second gets better as the sensors get denser, but the first does not get better at all, so a well-sampled design runs into the first problem, not the second.
The claim as recorded
Interpolating pilot phase between sparse detectors is subject to two separate limits acting on two different quantities. AMBIGUITY acts on the absolute deviation of a detector from its nominal position, which must stay within half a wavelength, 61.18 mm at 2.45 GHz, for the measured phase to be unwrapped at all. ACCURACY acts on the interpolation residual at an element, which must stay within 2.955 mm for CALC-085's 0.1 dB budget. Since interpolation exists to make the residual much smaller than the deformation, the better it works the more likely ambiguity is to bind first. Ambiguity binds before accuracy whenever the interpolation attenuates the deformation by more than about 21 times at one sigma of unwrap headroom, or about 7 times at three, which corresponds to a spacing-to-correlation-length ratio below 0.338 or 0.600 respectively.
Evidence behind this: T0-derivation. CALC-102 carries the arithmetic.
Scope and limits
The one-sigma row is not a design criterion, it is the point at which unwrapping fails half the time. A real design needs three sigma or better, which moves the crossover to D/Lc = 0.600 and makes ambiguity bind sooner still. No structural correlation length for a real structure is held, per CLM-0907, so which side of the crossover a real design sits on is not determined here. What is determined is where the boundary is. THREE POINTERS CORRECTED 2026-09-13 ON VERIFICATION. The closed form is CLM-0909's, not CLM-0897's - CLM-0897 is the patent reading on reference frequency and timing and contains no closed form - and the missing structural correlation length is CLM-0907's absence, not CLM-0901, which is the Ruze tolerance record. EACH WRONG POINTER NAMED A REAL RECORD, so every one of them resolved and rendered as a link, and only a reader who followed it would have found nothing there. The substance and every figure in this record were CONFIRMED. THE CLOSED FORM THIS RECORD USES WAS NOT REGISTERED, ESTABLISHED 2026-09-14; IT IS NOW CLM-0912 AND CALC-106, AND `support` AND `assumes` SHOULD BE READ AGAINST THOSE. The earlier pointer stands here as the record of what was wrong: those two fields cite CLM-0909 for it, and CLM-0909 IS THE INTERPOLATION READING OF THE PATENT AND CONTAINS NO CLOSED FORM - measured, zero occurrences - so it was a pointer to a real record that does not hold the thing, which is the failure a letter label at least makes obvious. The mathematics existed in working files and had never been entered. It has been, and the accuracy cap below no longer rests on an unregistered derivation. One thing did change in the registration: the 1 m row of the design table was re-solved and corrected from 0.083 to 0.082 Lc, the draft figure having rounded 0.8 per cent outside the budget. Nothing this record uses depends on that row.
What it assumes
The closed form of CLM-0909 for the residual, hence a stationary field with Gaussian autocorrelation, and the same conventions as that record.
Where it comes from
This is the project's own working, shown in full:
CALC-102. Ambiguity cap sigma_s < (lambda/2)/n_sigma. Accuracy cap sigma_s < 2.955 mm / sqrt(g(D/Lc)) using CLM-0909's closed form. Equal when sqrt(g) = 2.955 mm / ((lambda/2)/n_sigma), solved by bisection.
Basis: dimensionless (spacing ratio); mm (deformation rms)
Related records
- Answers question
- QST-013 not published in this release
Cite this
energyaperture.com/c/clm-0908
CLM-0908, Energy Aperture, as of 2026-09-13.