Is Mercury's Polar Ice Symmetric? A Pre-Registered Quantitative Framework for Testing the North–South Composition Assumption with BepiColombo's MGNS Instrument

This archive contains the complete data, code, and Monte Carlo outputs supporting a pre-registered, quantitative framework predicting what BepiColombo's Mercury Gamma-ray and Neutron Spectrometer (MGNS) should detect at Mercury's north and south poles. MESSENGER's Neutron Spectrometer found strong evidence for water ice at Mercury's north pole, but its eccentric orbit never sampled the south pole — every existing estimate of Mercury's total polar ice inventory therefore rests on the untested assumption that both poles share the same composition. This framework states specific, falsifiable numerical predictions for both poles ahead of BepiColombo's science operations phase, so they can be independently checked once the spacecraft's own measurements are returned. The archive includes: the full MESSENGER-era evidence base with sourcing and uncertainty-type annotations; a pooled Monte Carlo ensemble synthesizing independent ice-thickness estimates; an MGNS forward-response model built from the instrument team's own published sensitivity data; a full joint uncertainty budget propagating six independent uncertainty sources; a sensitivity analysis identifying which uncertainty source actually dominates the north-pole prediction; and a three-hypothesis Bayes-factor framework comparing competing explanations for the south pole's ice origin (symmetric accretion, ballistic delivery, and atmosphere-mediated delivery), built on real, sourced delivery-mechanism models. A Jupyter notebook (BepiColombo_Crosscheck.ipynb) reproduces the full pipeline end to end from a fixed random seed. Exact reported values are given in the associated manuscript. This deposit exists to make every intermediate step of that manuscript's analysis independently checkable and exactly reproducible.