Contents of this source
Scalar descent through quasi-étale uniformization
Status and authorship.
The scalar assertions proved below are DIRECT_COROLLARY of the specified primary results. They are not an arbitrary-current descent theorem and do not settle KC12. The independent-audit agent constructed this deduction after the root requested a source audit. The root independently proposed the same Galois-saturation repair after being informed of the relevant passage in Ducros. Neither participant is an author-independent reviewer of this deduction. A separate review is required before it is used as an accepted input.
Definitions and exact external interfaces
Let be a complete nontrivially valued non-Archimedean field. In its application to this project, is the original , or a finite extension used only in scalar analytic geometry, with the given absolute value extended and never renormalized. All spaces in the descent assertion are good, Hausdorff and strictly -analytic.
The notation has the geometric meaning of Gubler–Jell–Rabinoff, Definition 5.1: on a covering by strictly analytic domains the function is a finite real linear combination of logarithms of absolute values of analytic units on those domains. This definition concerns functions on the full analytic domains. It is not a definition in terms of a selected skeleton or a chosen finite set of ordinary current tests. We use the following primary inputs, with their original hypotheses.
Gubler–Jell–Rabinoff, arXiv:2111.05741v3, Definition 5.1, Remark 5.2 and Lemma 5.4: geometric real PL functions are G-local and are preserved by analytic pullback. Proposition 5.11, printed pp. 34–35, gives descent along finite, flat, surjective morphisms of good strictly analytic spaces. Corollary 5.12, printed p. 36, gives descent after extension to the completion of an algebraic extension.
Ducros, Families of Berkovich spaces, Theorem 5.4.6 and the reduction in the proof of Proposition 9.1.1(c), printed pp. 118–120 and 177, give the following local form. A quasi-étale morphism can be covered G-locally on its source by strict affinoid domains admitting a factorization where is an analytic domain of , is a connected strict affinoid domain of , and is connected finite étale. One may use finitely many such pieces when the source is compact. The same proof explicitly passes to a connected finite Galois cover of dominating .
Ducros, Proposition 9.1.1, printed pp. 177–178: if is quasi-compact and -strict, is separated, and a coherent sheaf on is -flat with quasi-finite support, its image is a compact -strict analytic domain of . We use only the case of the structure sheaf of a strict affinoid domain of a finite étale cover, with . Such a domain maps quasi-étale, and hence flat, to the base.
Temkin, arXiv:1503.05729v2, Theorem 3.3.1 and Corollary 3.3.2, printed p. 12, together with the rigid/Berkovich translation in §3.4: a compact quasi-smooth strictly -analytic space has, after a finite extension , a surjective quasi-étale affinoid cover possessing a strictly semistable formal -model. It is this statement, retaining the chosen strictly semistable model, that is used below. The separate wording of Theorem 3.4.1 says only that the maximal affine model is semistable; that wording alone is not silently strengthened.
Gubler–Rabinoff, arXiv:2506.13548v2, Definition 7.1 and Proposition 7.2(6),(7), printed pp. 34–35: classically pluriharmonic scalars are preserved by base extension and analytic pullback. Proposition 7.8, printed p. 36: on a good, quasi-compact strictly analytic space having a strictly semistable formal model, every classically pluriharmonic scalar is geometric real PL.
Here classically pluriharmonic means that both and are classically psh in the curve-test sense of the last reference. No identification with an arbitrary continuous ordinary-current harmonic witness is included in that definition.
Galois saturation and scalar PL descent
Proposition (compact quasi-étale scalar descent).
Let be a surjective quasi-étale morphism between compact good Hausdorff strictly -analytic spaces. For every real function on , The assertion is about the given function on all analytic points. It does not assert that has an interior lift over each point of , nor that becomes finite on an ordinary neighborhood of every target point.
Proof.
Analytic pullback proves the reverse implication. Assume the left side. By HU1 and compactness, choose finitely many strict affinoid source domains covering , with factorizations as in HU1. The finite family can be obtained from finitely many source neighborhoods covered by the local affinoid pieces; thus it is an actual G-cover and not just a collection of selected points. It suffices to prove the assertion on the image of each . Fix one piece and omit its index.
Choose a connected finite Galois cover dominating , and write . Let be the inverse image of . It is a compact strict affinoid domain of . The composite is quasi-étale: analytic-domain embeddings and finite étale maps are quasi-étale, and this property is stable under composition. In particular it is flat and has zero-dimensional fibers. Ducros, Proposition 9.1.1, applied to therefore shows that It is consequently also an analytic domain of .
For every analytic point , the Galois group acts transitively on the points of . This is the usual finite-Galois-cover fiber property, and is also the property used in the Galois saturation in Ducros’s proof of Proposition 9.1.1(c). More explicitly, after tensoring the finite étale fiber algebra with , its geometric points form a -torsor. The absolute Galois group of commutes with this action; therefore is transitive on its orbits, equivalently on the field factors of that finite étale algebra. Each finite extension of the complete valued field has its unique extended absolute value, so these factors are exactly the analytic fiber points. Hence Indeed, any point on the right maps to the image of ; conversely, a point above is a conjugate of some chosen point of above . Both sides carry their induced analytic-domain structure in , and analytic-domain structures on a fixed subset are unique. The union in HU4 is finite. No points lying on domain boundaries have been removed.
Put on , where is restricted to . Its restriction to is geometric real PL by pullback from . Since , the function is -invariant. Transporting the presentation on by each proves that is geometric real PL on every .
For clarity, no goodness assumption is needed for the possibly non-affinoid union in HU4. Cover by strict affinoid domains . For each such , the space is strict affinoid and is finite étale and surjective. Its intersections with the finitely many domains form a finite analytic-domain cover. Restriction of the preceding PL presentations and G-locality show that is real PL. Gubler–Jell–Rabinoff, Proposition 5.11, now gives . G-locality gives , understood through this affinoid cover if necessary.
Finally, the finitely many compact analytic domains cover , because the cover and is surjective. They form a finite G-cover of . Applying G-locality once more proves HU2.
What the saturation accomplishes.
The finite étale map used for descent is , after has been recognized as an actual target analytic domain. It need not map back into the single original source piece . The equality of the scalar pullback on all its Galois conjugates supplies PL presentations on the additional pieces. Thus this proof is not a domination of the original quasi-étale cover by an ordinary finite étale cover, and it makes no assertion of descent for an arbitrary current originally given only on .
The scalar consequence of altered uniformization
Corollary (classical pluriharmonic scalars are geometric real PL in the compact strict quasi-smooth case).
Let be a compact good Hausdorff quasi-smooth strictly -analytic space. Every classically pluriharmonic real function on belongs to : In particular this holds over the original . It also yields the corresponding local assertion at a point admitting a compact strict quasi-smooth analytic neighborhood on which the scalar is classically pluriharmonic.
Proof.
Use Temkin, Corollary 3.3.2 and §3.4, to choose a finite extension and a surjective quasi-étale morphism with affinoid and possessing a strictly semistable formal model over . The original absolute value is merely extended to . The scalar is classically pluriharmonic by Gubler–Rabinoff, Proposition 7.2(6),(7). Their Proposition 7.8 shows that it is geometric real PL on . Apply HU2 over to obtain . Finally, Gubler–Jell–Rabinoff, Corollary 5.12, descends this property along the finite algebraic extension and gives HU5. All steps concern scalars and their analytic pullbacks. No current over is constructed or imported.
Boundary pressure test and the remaining KC interfaces
Let satisfy , and consider the compact annulus and its two closed subannuli The map is surjective and quasi-étale. The unit Gauss point is in the ordinary interior of and on the analytic boundary of both displayed source domains. Thus the shortcut asserting an ordinary interior lift for every point of a surjective quasi-étale compact cover is FALSE. This is the already frozen R19 UF5 example, not a new counterexample family. It does not contradict HU2: themselves are a finite target G-cover, and scalar PL presentations glue on that cover. The associated R19 UF6–UF8 current example explains why discarding the boundary is nevertheless invalid for arbitrary-current action.
The following conclusions are deliberately not part of HU5.
A finite continuous scalar with ordinary has not, by the definition alone, been identified with a classically pluriharmonic scalar. Any use of HU5 for KC harmonic witnesses must pin a valid comparison in exactly that scalar class. No general positivity-to-regularization implication is inferred from the abstracts of the cited papers.
Geometric real PL on a G-cover is not yet a single global unit-log expression on a prescribed ordinary domain. A residue/Picard argument, when applicable, must retain its own exact hypotheses and ordinary-germ quantifiers.
HU2 and HU5 construct neither pullback nor descent for arbitrary allowed positive currents along the quasi-étale map. They do not localize a previously chosen full carrier polar.
No conclusion about mixed KC harmonic generators with arbitrary companion potentials, the full carrier-CLN seminorm, or the actual signed PL relation ideal follows just from scalar recognition. No ordinary-current equality is promoted to an arbitrary action on GK delta tests.
Accordingly the original same-germ KC12 implication and the variable-PS/CTO gates retain status OPEN. The precise topological obstruction to the naive three-result concatenation is repaired for geometric scalar PL descent by HU2; it is not repaired for arbitrary-current descent by the same argument.