Symmetric fits to one July 2025 WASP-103 b JWST transit give a 23-second difference between two infrared bands

We test symmetric transit centers in the public July 2025 JWST NIRSpec PRISM visit of WASP-103 b. We use archive pipeline 3.0.0 and CRDS context jwst_1584.pmap. Five fixed spectral-sample sums use 77,373 integrations. Spherical transit fits give a center in the 3.8–5.0 micrometre band that is 23.15 seconds later than in the 2.0–2.8 micrometre band. Paired residual-block refits give conditional scatter of 7.71 and 10.42 seconds for 5- and 20-minute blocks. The pair was selected after inspection. Several limb parameters reach physical bounds. These results show sensitivity to wavelength, but do not establish an atmospheric cause. The white sum gives BJD TDB 2460863.34428834. Eleven initial white configurations span 4.22 seconds. Extra checks extend this range to 4.60 seconds. White residual-block scatter is 6.72–7.41 seconds. The primary center is 28.61 seconds before the Kutluay et al. (2026) Table 4 linear clock. Independently delivered integration-time tables agree with the spectral times. Their first integration starts precede exposure-header starts by 4.73–4.90 seconds. We retain the delivered times and infer no correction. This unresolved difference and model errors limit absolute physical timing. We publish reduced data, source hashes, code, all fits, block draws, figures, and reproduction instructions. No new planet or period-change rate is inferred.

WASP-103 b · JWST · NIRSpec PRISM · transit timing · wavelength sensitivity · public astronomical data · reproducible analysis