H²(G_K, μₙ) is the n-torsion of the cohomological Brauer group #
Let K be a field, Kˢ a separable closure, G_K = AbsoluteGaloisGroup K, and n a natural
number invertible in K. The inclusion μₙ ⊆ (Kˢ)ˣ induces
H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ),
and this map is injective with image the n-torsion of H²(G_K, (Kˢ)ˣ). Both facts are read off
the long exact sequence of the Kummer sequence 1 → μₙ → (Kˢ)ˣ → (Kˢ)ˣ → 1
(TauCeti.kummerShortExact):
H¹(G_K, (Kˢ)ˣ) →δ¹→ H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ) →n→ H²(G_K, (Kˢ)ˣ).
- Injectivity is exactness at
H²(G_K, μₙ)together with Hilbert 90,H¹(G_K, (Kˢ)ˣ) = 0(TauCeti.subsingleton_H1_unitsCoeff). - The image is exactness at
H²(G_K, (Kˢ)ˣ), once the map induced by then-th power map of(Kˢ)ˣis recognised as multiplication bynonH²(TauCeti.explicitCoeff2_kummerShortExact_proj).
The argument runs in the explicit low-degree model, where the long exact sequence lives, and the
result is transported to Mathlib's continuousCohomology 2 through the comparison
TauCeti.ContCohomology.explicitH2AddEquivContinuousCohomology and its naturality in coefficient
maps, TauCeti.ContCohomology.explicitH2AddEquivContinuousCohomology_coeffMap.
This is how the n-torsion subgroup of the Brauer group is seen cohomologically: for a local field
the local invariant identifies the n-torsion of H²(G_K, (Kˢ)ˣ) with (1/n)ℤ/ℤ, and composing
with the injection here gives H²(G_K, μₙ) ≃ ℤ/n.
Main definitions #
TauCeti.kummerCoeffToUnits: the inclusionμₙ ⊆ (Kˢ)ˣas a morphism of canonical coefficient objects.TauCeti.h2KummerToUnits: the induced mapH²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ)on Mathlib's continuous cohomology.
Main results #
TauCeti.explicitCoeff2_kummerShortExact_incl_injectiveandTauCeti.mem_range_explicitCoeff2_kummerShortExact_incl_iff: injectivity and the image, on the explicit model.TauCeti.explicitCoeff2_kummerShortExact_restrict_incl_injectiveandTauCeti.mem_range_explicitCoeff2_kummerShortExact_restrict_incl_iff: the corresponding statements for subgroups; injectivity requires the subgroup to be closed.TauCeti.explicitCor2_kummerCoeff_bijective_of_unitsCoeff_bijective: bijectivity of corestriction on roots-of-unity coefficients, assuming bijectivity on units coefficients.TauCeti.h2KummerToUnits_injective:H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ)is injective.TauCeti.h2KummerToUnits_range: its image is then-torsion.
References #
- J.-P. Serre, Galois Cohomology, Chapter II, §5.2, Theorem 2 and its proof.
- J. Neukirch, A. Schmidt, K. Wingberg, Cohomology of Number Fields, 2nd ed., (6.2.1) and the exact sequence following it.
The explicit model #
The map on explicit H² induced by the n-th power map of (Kˢ)ˣ is multiplication by n:
in additive notation the power map is n • ·, and a coefficient map acts on cocycles by
postcomposition.
H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ) is injective, on the explicit model.
The image of H²(G_K, μₙ) in H²(G_K, (Kˢ)ˣ) is the n-torsion, on the explicit
model.
Subgroups #
The Kummer coefficient inclusion is injective on H² of every closed subgroup of G_K.
The image of the Kummer coefficient inclusion on H² of any subgroup of G_K is precisely
its n-torsion. Unlike injectivity, this image statement does not require closedness.
If corestriction on H² with units coefficients is bijective, so is corestriction
with roots-of-unity coefficients, for every exponent invertible in the field.
The canonical object #
The inclusion μₙ ⊆ (Kˢ)ˣ as a morphism of canonical coefficient objects over G_K.
Equations
Instances For
The morphism kummerCoeffToUnits is the inclusion μₙ ⊆ (Kˢ)ˣ on elements.
The map H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ) induced by the inclusion μₙ ⊆ (Kˢ)ˣ, on Mathlib's
continuous cohomology. It is how the n-torsion subgroup of H² sits inside the cohomological
Brauer group: injective (TauCeti.h2KummerToUnits_injective) with image the n-torsion
(TauCeti.h2KummerToUnits_range) when n is invertible in K.
Equations
Instances For
The defining equation of h2KummerToUnits: it is the coefficient map of the inclusion
kummerCoeffToUnits K n.
The comparison with the explicit model carries the explicit coefficient map of μₙ ⊆ (Kˢ)ˣ to
h2KummerToUnits.
H²(G_K, μₙ) → H²(G_K, (Kˢ)ˣ) is injective for n invertible in K.
The image of H²(G_K, μₙ) in H²(G_K, (Kˢ)ˣ) is the n-torsion for n invertible in
K.