ScalingStacks

2. Model metrics, semipositive metrics, and envelopes [0381]

Original official author HTML, exact retained edition. Historical TeX conversion verdicts remain unchanged. Cited-edition alignment and mathematical self-containment are not assessed.

Complete original source context Β· Original author HTML

2. Model metrics, semipositive metrics, and envelopes

Let XX be a proper variety over a complete non-archimedean valued field (K,||)(K,{|\phantom{a}|}).

2.1.

A model of XX is given by a proper flat scheme 𝒳{{\mathscr{X}}} over S:=Spec​K∘S:={\rm Spec}\,K^{\circ} together with an isomorphism hh between XX and the generic fiber 𝒳η{{\mathscr{X}}}_{\eta} of the SS-scheme 𝒳{{\mathscr{X}}} which we read as an identification. Given a model 𝒳{{\mathscr{X}}} of XX there is a canonical surjective reduction map red:XanβŸΆπ’³s{\mathrm{red}}\colon X^{\mathrm{an}}\longrightarrow{{\mathscr{X}}}_{s} where 𝒳s{{\mathscr{X}}}_{s} denotes the special fiber π’³βŠ—K∘K~{{\mathscr{X}}}\otimes_{K^{\circ}}\tilde{K} of 𝒳{{\mathscr{X}}} over SS.

Let LL be a line bundle on the proper variety XX. A model of (X,L)(X,L) or briefly a model of LL is given by a model (𝒳,h)({{\mathscr{X}}},h) of XX together with a line bundle β„’{\mathscr{L}} on 𝒳{{\mathscr{X}}} and an isomorphism hβ€²h^{\prime} between LL and hβˆ—β€‹(β„’|𝒳η)h^{*}({\mathscr{L}}|_{\mathscr{X}_{\eta}}) which we read as an identification.

Given a model (𝒳,β„’)({{\mathscr{X}}},{\mathscr{L}}) of (X,LβŠ—m)(X,L^{\otimes m}) for some mβˆˆβ„•>0m\in\mathbb{N}_{>0} there is a unique metric βˆ₯βˆ₯β„’{\|\ \|}_{\mathscr{L}} on LanL^{\mathrm{an}} over XanX^{\mathrm{an}} which satisfies the following: Given an open subset 𝒰{\mathscr{U}} of 𝒳{{\mathscr{X}}}, a frame tt of β„’{\mathscr{L}} over 𝒰{\mathscr{U}}, and a section ss of LL over U=Xβˆ©π’°U=X\cap{\mathscr{U}} we write sβŠ—m=h​ts^{\otimes m}=ht for some regular function hh on UU and get β€–sβ€–=|h|m\|s\|=\sqrt[m]{|h|} on Uan∩redβˆ’1​(𝒰s)U^{\mathrm{an}}\cap{\mathrm{red}}^{-1}({\mathscr{U}}_{s}). Such a metric on LanL^{\mathrm{an}} is called a model metric determined on 𝒳{{\mathscr{X}}}.

2.2.

A model metric βˆ₯⁣βˆ₯{\|\ \|} on π’ͺXan{\mathcal{O}}_{X^{\mathrm{an}}} induces a continuous function f=βˆ’log⁑‖1β€–:Xan→ℝf=-\log\|1\|\colon X^{\mathrm{an}}\to\mathbb{R}. The space of model functions

π’Ÿ(X)={f:Xan→ℝ|f=βˆ’logβˆ₯1βˆ₯Β for some model metricΒ βˆ₯βˆ₯Β onΒ π’ͺXan}{\mathscr{D}}(X)=\{f\colon X^{\mathrm{an}}\rightarrow\mathbb{R}\,|\,f=-\log\|1\|\mbox{ for some model metric }{\|\ \|}\mbox{ on }{\mathcal{O}}_{X^{\mathrm{an}}}\}

has a natural structure of a β„š\mathbb{Q}-vector space. We say that a model function f=βˆ’log⁑‖1β€–f=-\log\|1\| is determined on a model 𝒳{{\mathscr{X}}} if the model metric βˆ₯⁣βˆ₯{\|\ \|} is determined on 𝒳{{\mathscr{X}}}. A vertical divisor DD on 𝒳{{\mathscr{X}}} determines a model π’ͺ⁑(D){\mathcal{O}}(D) of π’ͺX{\mathcal{O}}_{X} and an associated model function Ο†Dβ‰”βˆ’log⁑‖1β€–π’ͺ⁑(D)\varphi_{D}\coloneqq-\log\|1\|_{{\mathcal{O}}(D)}. Such model functions are called β„€\mathbb{Z}-model functions. Let π”ž{\mathfrak{a}} denote a vertical ideal of 𝒳{{\mathscr{X}}}. Let EE denote the exeptional divisor of the blowup of 𝒳{{\mathscr{X}}} in π”ž{\mathfrak{a}}. Then log⁑|π”ž|:=Ο†E\log|{\mathfrak{a}}|:=\varphi_{E} is called the β„€\mathbb{Z}-model function defined by the vertical ideal π”ž{\mathfrak{a}}.

2.3.

Consider a model 𝒳{{\mathscr{X}}} of the proper variety XX over KK. The rational vector space space N1​(𝒳/S)β„šN^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}} is by definition the quotient of Pic​(𝒳)β„šβ‰”Pic⁑(𝒳)βŠ—β„€β„š{{\rm Pic}\,}({{\mathscr{X}}})_{\mathbb{Q}}\coloneqq{\rm Pic}({{\mathscr{X}}})\otimes_{\mathbb{Z}}\mathbb{Q} by the subspace generated by classes of line bundles β„’{\mathscr{L}} such that β„’β‹…C=0{\mathscr{L}}\cdot C=0 for each closed curve CC in the special fiber 𝒳s{{\mathscr{X}}}_{s}. Note that N1​(𝒳/S)β„šN^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}} is finite dimensional by applying [Kle66, Prop.Β IV.1.4] to 𝒳s{{\mathscr{X}}}_{s}. We define N1​(𝒳/S)≔N1​(𝒳/S)β„šβŠ—β„šβ„N^{1}({{\mathscr{X}}}/S)\coloneqq N^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}}\otimes_{\mathbb{Q}}\mathbb{R}. An element α∈N1​(𝒳/S)β„š\alpha\in N^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}} (resp.Β OPENα∈N1​(𝒳/S))\alpha\in N^{1}({{\mathscr{X}}}/S)) is called nef if Ξ±β‹…Cβ‰₯0\alpha\cdot C\geq 0 for all closed curves CC in 𝒳s{{\mathscr{X}}}_{s}. We call a line bundle β„’{\mathscr{L}} on 𝒳{{\mathscr{X}}} nef if the class of β„’{\mathscr{L}} in N1​(𝒳/S)N^{1}({{\mathscr{X}}}/S) is nef.

2.4.

We define 𝒡1,1​(X)β„š\mathcal{Z}^{1,1}(X)_{\mathbb{Q}} as the direct limit

(2.1) 𝒡1,1​(X)β„šβ‰”lim→⁑N1​(𝒳/S)β„š,\displaystyle\mathcal{Z}^{1,1}(X)_{\mathbb{Q}}\coloneqq\varinjlim N^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}},

where 𝒳{{\mathscr{X}}} runs over the isomorphism classes of models of XX. The space of closed (1,1)(1,1)-forms on XX is defined as 𝒡1,1​(X)≔𝒡1,1​(X)β„šβŠ—β„šβ„\mathcal{Z}^{1,1}(X)\coloneqq\mathcal{Z}^{1,1}(X)_{\mathbb{Q}}\otimes_{\mathbb{Q}}\mathbb{R}. Let LL be a line bundle on XX. Let βˆ₯⁣βˆ₯{\|\ \|} be a model metric on LanL^{\mathrm{an}} which is determined on 𝒳{{\mathscr{X}}} by a model β„’{\mathscr{L}} of LβŠ—mL^{\otimes m}. We multiply the class of β„’{\mathscr{L}} in N1​(𝒳/S)β„šN^{1}({{\mathscr{X}}}/S)_{\mathbb{Q}} by mβˆ’1m^{-1} which determines a well defined class c1(L,βˆ₯βˆ₯)βˆˆπ’΅1,1(X)β„šβŠ†π’΅1,1(X)c_{1}(L,{\|\ \|})\in\mathcal{Z}^{1,1}(X)_{\mathbb{Q}}\subseteq\mathcal{Z}^{1,1}(X) called the curvature form c1(L,βˆ₯βˆ₯)c_{1}(L,{\|\ \|}) of (L,βˆ₯βˆ₯)(L,{\|\ \|}).

A closed (1,1)(1,1)-form ΞΈ\theta is called semipositive if it is represented by a nef element ΞΈπ’³βˆˆN1​(𝒳/S)\theta_{{\mathscr{X}}}\in N^{1}({{\mathscr{X}}}/S) for some model 𝒳{{\mathscr{X}}} of XX. We say that a model metric βˆ₯⁣βˆ₯{\|\ \|} on LanL^{\mathrm{an}} for a line bundle LL on XX is semipositive if the same holds for the curvature form c1(L,βˆ₯βˆ₯)c_{1}(L,{\|\ \|}).

2.5.

For ΞΈβˆˆπ’΅1,1​(X)\theta\in\mathcal{Z}^{1,1}(X) we denote by

PSHπ’Ÿ(X,ΞΈ)={fβˆˆπ’Ÿ(X)|ΞΈ+c1(π’ͺX,βˆ₯βˆ₯trivβ‹…eβˆ’f)βˆˆπ’΅1,1(X)Β is semipositive}{\rm PSH}_{\mathscr{D}}(X,\theta)=\{f\in{\mathscr{D}}(X)\,|\,\theta+c_{1}(\mathcal{O}_{X},{\|\ \|}_{\rm triv}\cdot e^{-f})\in\mathcal{Z}^{1,1}(X)\text{ is semipositive}\}

the set of ΞΈ\theta-plurisubharmonic (ΞΈ\theta-psh for short) model functions. Recall from [GM16, Prop.Β 3.12] that the set PSHπ’Ÿβ€‹(X,ΞΈ){\rm PSH}_{\mathscr{D}}(X,\theta) is stable under the formation of max.

2.6.

If YY is a proper variety over an arbitrary field kk, we denote by N1​(Y)β„šN^{1}(Y)_{\mathbb{Q}} the rational vector space Pic⁑(Y)βŠ—β„š{{\rm Pic}\,}(Y)\otimes\mathbb{Q} modulo numerical equivalence. Similarly, we denote by N1​(Y)=N1​(Y)β„šβŠ—β„šβ„N^{1}(Y)=N^{1}(Y)_{\mathbb{Q}}\otimes_{\mathbb{Q}}\mathbb{R} the real vector space Pic⁑(Y)βŠ—β„{{\rm Pic}\,}(Y)\otimes\mathbb{R} modulo numerical equivalence. A class in N1​(Y)N^{1}(Y) is called ample if it is an ℝ>0\mathbb{R}_{>0}-linear combination of classes induced by ample line bundles on YY. An element α∈N1​(Y)β„š\alpha\in N^{1}(Y)_{\mathbb{Q}} (resp.Β OPENα∈N1​(Y))\alpha\in N^{1}(Y)) is called nef if Ξ±β‹…Cβ‰₯0\alpha\cdot C\geq 0 for all closed curves CC in YY.

2.7.

The restriction maps N1​(𝒳/S)β†’N1​(X),[β„’]↦[β„’|X]N^{1}({{\mathscr{X}}}/S)\rightarrow N^{1}(X),\,[{\mathscr{L}}]\mapsto[{\mathscr{L}}|_{X}] induce a linear map {}:𝒡1,1​(X)⟢N1​(X),θ↦{ΞΈ}\{\phantom{a}\}:\mathcal{Z}^{1,1}(X)\longrightarrow N^{1}(X),\,\theta\mapsto\{\theta\}. We call {ΞΈ}\{\theta\} the de Rham class of ΞΈ\theta.

Definition 2.8.

Let XX be a smooth projective variety over KK and ΞΈβˆˆπ’΅1,1​(X)\theta\in\mathcal{Z}^{1,1}(X) with de Rham class {ΞΈ}∈N1​(X)\{\theta\}\in N^{1}(X). The ΞΈ\theta-psh envelope of u∈C0​(X)u\in C^{0}(X) is the function

(2.2) Pθ​(u):X→ℝ,Pθ​(u)​(x)=sup{φ⁑(x)|Ο†βˆˆPSHπ’Ÿβ€‹(X,ΞΈ)βˆ§Ο†β‰€u}.{P}_{\theta}(u)\colon X\to\mathbb{R},\,\,{P}_{\theta}(u)(x)=\sup\{\varphi(x)\,|\,\varphi\in{\rm PSH}_{\mathscr{D}}(X,\theta)\wedge\varphi\leq u\}.

Note that Pθ​(u){P}_{\theta}(u) is a real valued function if and only if there exists a ΞΈ\theta-psh model function. For the existence of a ΞΈ\theta-psh model function, it is necessary that the de Rham class {ΞΈ}\{\theta\} is nef (see [GM16, 4.8] and [BFJ16a, Rem.Β 5.4]). If {ΞΈ}\{\theta\} is ample, then there exists always a ΞΈ\theta-psh model function and hence Pθ​(u){P}_{\theta}(u) is a real valued function. If there is no ΞΈ\theta-psh function, then Pθ​(u)β‰‘βˆ’βˆž{P}_{\theta}(u)\equiv-\infty by definition.

If the residue characteristic is zero and if the de Rham class {ΞΈ}\{\theta\} is ample, our definition of Pθ​(u){P}_{\theta}(u) is by [BFJ16a, Thm.Β 8.3 and Lemma 8.9] equivalent to the definition of Boucksom, Favre, and Jonsson in [BFJ16a, Def.Β 8.1] .

The next Proposition collects elementary properties of envelopes.

Proposition 2.9.

Let u,uβ€²βˆˆC0​(Xan)u,u^{\prime}\in C^{0}(X^{\mathrm{an}}) and ΞΈ,ΞΈβ€²βˆˆπ’΅1,1​(X)\theta,\theta^{\prime}\in\mathcal{Z}^{1,1}(X).

  1. (i)

    If u≀uβ€²u\leq u^{\prime} then Pθ​(u)≀Pθ​(uβ€²){P}_{\theta}(u)\leq{P}_{\theta}(u^{\prime}).

  2. (ii)

    We have Pt​θ+(1βˆ’t)​θ′​(t​u+(1βˆ’t)​uβ€²)β‰₯t​Pθ​(u)+(1βˆ’t)​Pθ′​(uβ€²){P}_{t\theta+(1-t)\theta^{\prime}}(tu+(1-t)u^{\prime})\geq t{P}_{\theta}(u)+(1-t){P}_{\theta^{\prime}}(u^{\prime}) for all t∈[0,1]t\in[0,1].

  3. (iii)

    We have Pθ​(u)+c=Pθ​(u+c){P}_{\theta}(u)+c={P}_{\theta}(u+c) for each cβˆˆβ„c\in\mathbb{R}.

  4. (iv)

    We have Pθ​(u)βˆ’v=PΞΈ+d​dc​v​(uβˆ’v){P}_{\theta}(u)-v={P}_{\theta+dd^{c}v}(u-v) for each vβˆˆπ’Ÿβ‘(X)v\in{\mathscr{D}}(X).

  5. (v)

    If Pθ​(u)β‰’βˆ’βˆž{P}_{\theta}(u)\not\equiv-\infty, then we have supXan|Pθ​(u)βˆ’Pθ​(uβ€²)|≀supXan|uβˆ’uβ€²|\sup_{X^{\mathrm{an}}}|{P}_{\theta}(u)-{P}_{\theta}(u^{\prime})|\leq\sup_{X^{\mathrm{an}}}|u-u^{\prime}|.

  6. (vi)

    If ΞΈ\theta is determined on a model 𝒳{{\mathscr{X}}}, if the de Rham class {ΞΈ}\{\theta\} is ample and if ΞΈmβ†’ΞΈ\theta_{m}\to\theta in N1​(𝒳/S)N^{1}({{\mathscr{X}}}/S), then PΞΈm​(u)β†’Pθ​(u)P_{\theta_{m}}(u)\to{P}_{\theta}(u) uniformly on XanX^{\mathrm{an}}.

  7. (vii)

    We have Pt​θ​(t​u)=t​Pθ​(u){P}_{t\theta}(tu)=t{P}_{\theta}(u) for all tβˆˆβ„>0t\in\mathbb{R}_{>0}.

  8. (viii)

    Assume Pθ​(u)β‰’βˆ’βˆž{P}_{\theta}(u)\not\equiv-\infty. Then the envelope Pθ​(u)P_{\theta}(u) is continuous if and only if it is a uniform limit of ΞΈ\theta-psh model functions.

Proof.

The proof of Properties (i)–(vi) in [BFJ16a, Prop.Β 8.2] works in our setup as well. Property (vii) is obvious for tβˆˆβ„š>0t\in\mathbb{Q}_{>0} and an easy approximation argument then shows (vii) in general. We have seen that ΞΈ\theta-psh model functions are closed under max\max and hence the ΞΈ\theta-psh model functions φ≀u\varphi\leq u form a directed family. We conclude that (viii) follows from Dini’s Theorem for nets [Kel75, p.Β 239] and the definition of Pθ​(u)P_{\theta}(u). ∎

Proposition 2.10.

Let LL be an ample line bundle on XX, β„’{\mathscr{L}} an extension to a model 𝒳{{\mathscr{X}}} and ΞΈ=c1(L,βˆ₯βˆ₯β„’)βˆˆπ’΅1,1(X)\theta={c_{1}(L,{\|\ \|}_{\mathscr{L}})}\in\mathcal{Z}^{1,1}(X). For m>0m>0 let

(2.3) π”žm=Im​(H0​(𝒳,β„’βŠ—m)βŠ—Kβˆ˜β„’βŠ—βˆ’mβ†’π’ͺ𝒳){\mathfrak{a}}_{m}=\mbox{\rm Im}\,\bigl(H^{0}({{\mathscr{X}}},{\mathscr{L}}^{\otimes m})\otimes_{K^{\circ}}{\mathscr{L}}^{\otimes-m}\to{\mathcal{O}}_{{\mathscr{X}}}\bigr)

be the mm-th base ideal of β„’{\mathscr{L}} and Ο†m:=mβˆ’1​log⁑|π”žm|\varphi_{m}:=m^{-1}\log|{\mathfrak{a}}_{m}|. Then Ο†m∈PSHπ’Ÿβ€‹(X,ΞΈ)\varphi_{m}\in{\rm PSH}_{\mathscr{D}}(X,\theta) and

(2.4) limmβ†’βˆžΟ†m=supmβˆˆβ„•Ο†m=Pθ​(0)\lim_{m\to\infty}\varphi_{m}=\sup_{m\in\mathbb{N}}\varphi_{m}={P}_{\theta}(0)

pointwise on XanX^{\mathrm{an}}.

Proof.

This is shown as in Step 1 of the proof of [BFJ16a, Thm. 8.5]. ∎

Proposition 2.11.

Let Kβ€²/KK^{\prime}/K be a finite normal extension and let q:Xβ€²:=XβŠ—KKβ€²β†’Xq\colon X^{\prime}:=X\otimes_{K}{K^{\prime}}\to X be the natural projection. For ΞΈβˆˆπ’΅1,1​(X)\theta\in\mathcal{Z}^{1,1}(X) and u∈C0​(Xan)u\in C^{0}({X^{{\mathrm{an}}}}), we have

(2.5) qβˆ—β€‹(Pθ​(u))=Pqβˆ—β€‹ΞΈβ€‹(qβˆ—β€‹(u)).q^{*}(P_{\theta}(u))=P_{q^{*}\theta}(q^{*}(u)).
Proof.

Splitting the extension Kβ€²/KK^{\prime}/K into a purely inseparble part and a Galois part, we can reduce to two cases. In the first case of a purely insparable extension, the result follows from Lemma 2.12 below. In the second case of a Galois extension, we can apply the argument of [BFJ15, Lemma A.4]. ∎

Lemma 2.12.

Let LL be a line bundle on XX. Let Kβ€²/KK^{\prime}/K be a finite purely inseparable extension and let q:X′≔XβŠ—KKβ€²β†’Xq\colon X^{\prime}\coloneqq X\otimes_{K}{K^{\prime}}\to X be the natural projection. Then the map qβˆ—q^{*} induces a bijection between the set of model metrics on LL and the set of model metrics on qβˆ—β€‹(L)q^{*}(L). Moreover this bijection identifies semipositive metrics on LL and on qβˆ—β€‹(L)q^{*}(L).

Proof.

We always consider the GG-topology induced by the strictly KK-affinoid domains. We claim that the map q:(X′)an→Xanq\colon(X^{\prime})^{\rm an}\to{X^{{\mathrm{an}}}} is a homeomorphism and that it also identifies the GG-topologies. In fact, this follows easily from the following claim:

Step 1: Let VV be a strictly affinoid space over KK and V′≔Vβ€‹βŠ—^K​Kβ€²V^{\prime}\coloneqq V\hat{\otimes}_{K}K^{\prime}. Then the natural projection q:Vβ€²β†’Vq\colon V^{\prime}\to V is a homeomorphism which identifies the GG-topologies.

Let pe=[Kβ€²:K]p^{e}=[K^{\prime}:K] be the degree of the purely inseparable field extension. It is clear that for every g∈π’ͺ⁑(Vβ€²)g\in\mathcal{O}(V^{\prime}), there is f∈π’ͺ⁑(V)f\in\mathcal{O}(V) with

(2.6) gpe=f∘q.g^{p^{e}}=f\circ q.

This property easily shows that q:Vβ€²β†’Vq\colon V^{\prime}\to V is a homeomorphism which we read now as an identification. Using that (2.6) holds also for rational functions gg on Vβ€²V^{\prime} and ff on VV, we see that VV and Vβ€²V^{\prime} have the same strictly rational domains. By the Gerritzen–Grauert theorem [BGR84, Cor.Β 7.3.5/3], we deduce the Step 1.

Next we prove the bijective correspondence between the model metrics on LL and on Lβ€²L^{\prime}. For this, it is enough to show that we have a bijective correspondence between model functions on Xan{X^{{\mathrm{an}}}} and model functions on (Xβ€²)an(X^{\prime})^{\rm an}.

We recall from [GM16, Def.Β 2.8, 2.11] that a piecewise β„š\mathbb{Q}-linear function on a strictly KK-analytic space WW is a function f:W→ℝf:W\to\mathbb{R} such that there is a GG-covering {Ui}i∈I\{U_{i}\}_{i\in I} of WW by strictly affinoid domains, analytic functions Ξ³i∈π’ͺ​(Ui)Γ—\gamma_{i}\in\mathcal{O}(U_{i})^{\times} and non-zero miβˆˆβ„•m_{i}\in\mathbb{N} with mi​f=βˆ’log⁑|Ξ³i|m_{i}f=-\log|\gamma_{i}| on UiU_{i} for every i∈Ii\in I.

By [GM16, Rem.Β 2.6, Prop.Β 2.10], model functions and piecewise β„š\mathbb{Q}-linear functions are the same and hence we have to check the bijective correspondence between piecewise β„š\mathbb{Q}-linear functions on Xan{X^{{\mathrm{an}}}} and (Xβ€²)an(X^{\prime})^{\rm an}. This can be checked GG-locally and hence it is enough to prove the following:

Step 2: Using the same assumptions as in Step 1, the map f↦f∘qf\mapsto f\circ q is an isomorphism from the group of piecewise β„š\mathbb{Q}-linear functions on VV onto the group of piecewise β„š\mathbb{Q}-linear functions on Vβ€²V^{\prime}.

Using the above definition of piecewise β„š\mathbb{Q}-linear functions, Step 1 and (2.6) yield easily Step 2.

To deduce the lemma, it remains to check that the identification between the model metrics on LL and Lβ€²L^{\prime} preserves semipositivity. This is an easy consequence of the projection formula applied to finite morphisms between closed curves in the special fibers of models. ∎

Original mathematics by the credited authors. Source-backed reader collection; mathematical self-containment is not assessed.