Constant color is not rejected for 72 ZTF nights of WD J1013−0427 with free shared offsets and Gaussian magnitude errors

We tested shared calibration effects in a frozen public ZTF image extraction of WD J1013−0427. The three-pixel aperture supplies 72 paired nights from 2018 to May 2026. Under independent Gaussian magnitude errors with known supplied variances, constant color gives p = 0.00498 with fixed offsets. It gives p = 0.07499 when two shared native magnitude offsets are free for the 22 image-control nights. Inversion of the fixed-slope Gaussian test gives conditional 95 percent slope sets of [0.338, 0.901] and [0.346, 1.032], respectively. A conservative bounded-offset test stops rejecting constant color when an allowed differential shift reaches 0.03089 mag. This corresponds to a bound of 0.01545 mag per band. The audit does not show that such an error is probable. On the common 61-night sample, the five- and seven-pixel extractions give empty confidence sets under this magnitude model. We supply the test derivation, numerical checks, simulation trials, code, result tables, figures, and the complete signed source publication with its images and provenance. The coverage result is conditional on fixed known errors and does not establish real survey coverage. This work gives a calibration requirement, not a new planet, event, or dust grain size.

white dwarfs · ZTF · color calibration · shared errors · confidence sets · public astronomical data · measurement errors · circumstellar debris · reproducibility