Cause of ExoTiC-JEDI susceptibility to GP-induced artifacts near transit

Ascertain why the ExoTiC-JEDI reduction of JWST/NIRSpec PRISM time-series photometry for the Kepler-167e transit is more susceptible than the custom and katahdin reductions to Gaussian-process trend modeling artifacts that introduce large step-function features near the planetary transit, yielding spurious exomoon signals; identify the specific reduction steps or assumptions responsible for this discrepancy and characterize their interaction with GP trend modeling.

Background

In the authors’ matrix of analyses, the ExoTiC-JEDI reduction combined with Gaussian-process trend models yielded pronounced step-function behavior coincident with the planetary transit, driving large apparent moon signals (R_SP ~ 0.17) that the authors consider spurious. This behavior was not observed in the custom or katahdin reductions using the same GP approaches.

The authors note that the artifact could relate to the similarity between the exposure-long trend timescale and the transit duration, but explicitly state that it is unclear why ExoTiC-JEDI would be more affected than the other reductions. They also report chromatic and negative-radius moon tests indicating the signal is spurious, strengthening the motivation to identify the underlying cause in the pipeline–model interaction.