One idea in one sentence.
Horizon activity (growth/mergers) seeds tiny, conserved source terms that route energy into three channels — vacuum (looks like dark‑energy drift), dark matter (as soliton‑like grains), and gravitational waves — with the split controlled by simple microphysics.
What this isn’t.
Not a fifth force; not a particle zoo; not a rewrite of GR/QFT. When horizon sourcing is gated down (or negligible), the model reverts to standard ΛCDM with cold dark matter.
Active horizons, not violations.
We use an exterior source proportional to horizon‑entropy production; local covariant conservation is enforced, and causality is retarded. No energy exits from inside horizons.
Grains = dark matter.
The sourced dark matter is modeled as classical solitons from an SU(2) Skyrme EFT. One micro length scale (R_*) fixes the mass (m) and the velocity‑dependent
Drift comes from demographics.
A small vacuum term drifts with the integrated build‑up of horizon area (black‑hole formation and mergers). If data prefer no drift, set that channel to zero and recover ΛCDM.
- Grain mass (m) — fixed by the soliton solution (calibrated at (m \approx 6\ \mathrm{GeV})).
- **Grain–grain scattering
$(\sigma_T/m)(v)$ — fixed by (R_*) and effective‑range/finite‑size physics; anchored at dwarfs, falls with (v); cluster‑safe by construction. - Optional “twist” for ringdown — a small, environment/boundary‑condition parameter that could induce sub‑percent, mass‑scaled offsets in QNM tones. Optional and set to 0 if null; not required by the rest of the framework.
All three can scale with the same micro scale (R_*) on dimensional grounds, but only (1)–(2) are presently derived from the Skyrme sector.
- CDM‑limit by default. Structure formation behaves like CDM; the SIDM curve is mild and velocity‑dependent; cluster bounds are respected.
- Dark energy close to (w=-1). Late‑time (w) is (-1) up to percent‑level drift at most, and can be dialed exactly to (-1) if data prefer.
- Causality/locality. Sourcing is retarded; a topological/4‑form bookkeeping option can enforce global budget updates without superluminal propagation.
- Tiny shedding fraction. The sourced DM fraction is minute; there is no late‑time flood.
- Why DM “interacts a bit but not too much.” A single micro scale (R_*) sets both the mass and a gentle, falling ((\sigma_T/m)(v)).
- Why DE looks nearly constant yet may drift slightly. Drift tracks real BH demographics and shuts off smoothly when sourcing subsides.
- Why some BH observables might be nudged. An optional sub‑percent, mass‑scaled ringdown shift is allowed but not required.
-
SIDM curve from dwarfs to clusters (no per‑halo knobs). One calibrated curve should fit dwarf cores and remain within cluster limits. If an extreme system strains the fit, check systematics before declaring failure.
-
Background drift vs BH growth. Any departure from ΛCDM should correlate with the integrated horizon‑area history. If data show no drift, set the vacuum channel to zero.
-
PTA‑band correlation test. Regions/epochs with more horizon activity should weakly up‑weight nano‑Hz background power (a correlation test, not an amplitude claim).
-
Ringdown tones (optional). Look for sub‑percent, mass‑scaled offsets. If stacked high‑SNR events show none, set the twist parameter to 0 and proceed.
-
Polarization–flux–color “triplet” in specific transients (hypothesis). In episodes dominated by magnetic/topological relaxation (not compression), expect brightening + hardening with a drop in linear polarization within the same state. This is regime‑bound and should be pre‑registered as a sign test.
- Tier 0 — Null‑compatible. If SN/BAO/structure data favor strict ΛCDM and ringdown shifts are null, set vacuum drift and twist to zero. The framework reduces to ΛCDM + CDM, with soliton DM microphysics still allowed.
- Tier 1 — Parameter revision. If the sign of a regime‑bound observable (e.g., polarization triplet) or a PTA correlation disagrees, revisit gating (which episodes dominate) and the partition split, not the whole framework.
- Tier 2 — Hard no‑go. Violations of conservation/causality; a required, large late‑time energy injection that cannot be gated down; or cluster‑scale self‑interactions that grossly overshoot bounds and cannot be mitigated by velocity dependence.
(Ringdown alone is not a kill switch: a null simply sets the twist (\approx 0).)
Over extremely long times, horizons bleed mass and the universe trends toward heat death. In this picture that phase releases the last residuals and could “reset” conditions for a next cycle. This cycle view is philosophical; it is not needed for near‑term tests. However it is used to justify meeting the CMB and BAO constraints for early universe creation and if constraints hold across the microphysics -> macrophysics observables then it will need to be revisited.