You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Model the crowd wave: coherence of a whole dark matter halo, not single particles #2
The witness number W is computed per particle. Wave dark matter's galaxy-wide wave is a crowd effect: enormous numbers of particles share one wave, described by a single field under its own gravity (the Schrödinger–Poisson equations). How does a crowd wave lose coherence, and does the per-particle W say anything about it?
Why it matters
The Witness Map's finding that dark matter "has no witnesses" is about single particles. A many-particle wave can behave differently: interference granules, a dense central core, and gravitational scattering among the granules themselves.
First steps
Write down the coherence time of a fuzzy dark matter halo from the literature (granule relaxation, core oscillations).
Compare it with the per-particle picture and say plainly where the map's statement stops applying.
Decide whether a crowd row belongs on the map, and how to label it so it is not confused with the per-particle rows.
Done in #10 (which says "Fixes #2"), landed on main through #22. GitHub did not apply the closing keyword because #10 merged indirectly, so I'm closing it by hand. All three first steps are covered: the coherence clocks from the literature, the comparison with the per-particle picture, and the crowd panel, labelled "not a witness number". #10's reviewer recommended publishing the panel, and it is now on the page.
Activity
zero-aska commentedon Oct 18, 2026
Done in #10 (which says "Fixes #2"), landed on main through #22. GitHub did not apply the closing keyword because #10 merged indirectly, so I'm closing it by hand. All three first steps are covered: the coherence clocks from the literature, the comparison with the per-particle picture, and the crowd panel, labelled "not a witness number". #10's reviewer recommended publishing the panel, and it is now on the page.
Generated by Claude Code