ScalingStacks

3.2. Embedding the dual complex in the Berkovich space [01EF]

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

3.2. Embedding the dual complex in the Berkovich space

Theorem 3.1.

Let 𝒳\mathcal{X} be any SNC model of XX.

  • (i)

    The image of the evaluation map ev𝒳:Xβ†’Div0⁑(𝒳)π‘βˆ—\ev_{\mathcal{X}}:X\to\Div_{0}(\mathcal{X})^{*}_{\mathbf{R}} coincides with Δ𝒳\Delta_{\mathcal{X}}.

  • (ii)

    There exists a unique continuous (injective) map emb𝒳:Δ𝒳→X\emb_{\mathcal{X}}:\Delta_{\mathcal{X}}\to X such that:

    • (a)

      evπ’³βˆ˜emb𝒳\ev_{\mathcal{X}}\circ\emb_{\mathcal{X}} is the identity on Δ𝒳\Delta_{\mathcal{X}};

    • (b)

      for sβˆˆΞ”π’³s\in\Delta_{\mathcal{X}}, the center of emb𝒳⁑(s)\emb_{\mathcal{X}}(s) on 𝒳\mathcal{X} is the generic point ΞΎJ\xi_{J} of EJE_{J} for the unique subset JβŠ‚IJ\subset I such that ss is contained in the relative interior of ΟƒJ\sigma_{J}.

The proof is given inΒ Β§3.3. Let us derive some consequences.

For any two models Ο€:𝒳→𝒴\pi:\mathcal{X}\to\mathcal{Y}, observe that the natural map Ο€βˆ—t:Div0⁑(𝒳)βˆ—β†’Div0⁑(𝒴)βˆ—{}^{t}\pi^{*}:\Div_{0}(\mathcal{X})^{*}\to\Div_{0}(\mathcal{Y})^{*} maps Δ𝒳\Delta_{\mathcal{X}} onto Δ𝒴\Delta_{\mathcal{Y}} since Ο€βˆ—t∘ev𝒳=ev𝒴{}^{t}\pi^{*}\circ\ev_{\mathcal{X}}=\ev_{\mathcal{Y}} by definition. We may thus form the projective limit lim←𝒳​SNC⁑Δ𝒳\varprojlim_{\mathcal{X}\ \text{SNC}}\Delta_{\mathcal{X}}, and we have

Corollary 3.2.

The maps ev𝒳:Xβ†’Ξ”π’³βŠ‚Div0⁑(𝒳)π‘βˆ—\ev_{\mathcal{X}}:X\to\Delta_{\mathcal{X}}\subset\Div_{0}(\mathcal{X})_{\mathbf{R}}^{*} induce a homeomorphism

(3.1) ev:Xβ†’lim←𝒳​SNC⁑Δ𝒳.\ev:X\to\varprojlim_{\mathcal{X}\ \text{SNC}}\Delta_{\mathcal{X}}.
Proof.

The map ev\ev is well-defined by TheoremΒ 3.1Β (i). It is a homeomorphism onto its image by CorollaryΒ 2.5 and the fact that any model is dominated by an SNC model. As XX is compact, we only need to show that ev⁑(X)\ev(X) is dense in lim←⁑Δ𝒳\varprojlim\Delta_{\mathcal{X}}. Pick s=(s𝒳)π’³βˆˆlim←⁑Δ𝒳s=(s_{\mathcal{X}})_{\mathcal{X}}\in\varprojlim\Delta_{\mathcal{X}} and fix an SNC model 𝒳\mathcal{X}. If 𝒴\mathcal{Y} is an SNC model dominated by 𝒳\mathcal{X}, then evπ’³βˆ˜emb𝒳=id\ev_{\mathcal{X}}\circ\emb_{\mathcal{X}}=\id yields ev𝒴⁑(emb𝒳⁑(s𝒳))=s𝒴\ev_{\mathcal{Y}}(\emb_{\mathcal{X}}(s_{\mathcal{X}}))=s_{\mathcal{Y}}. Hence s=lim𝒳ev⁑(emb𝒳⁑(s𝒳))∈ev⁑(X)Β―s=\lim_{\mathcal{X}}\ev(\emb_{\mathcal{X}}(s_{\mathcal{X}}))\in\overline{\ev(X)}. ∎

Definition 3.3.

For any SNC model 𝒳\mathcal{X} we define a continuous map p𝒳:Xβ†’Xp_{\mathcal{X}}:X\to X by

p𝒳:=embπ’³βˆ˜ev𝒳.p_{\mathcal{X}}:=\emb_{\mathcal{X}}\circ\ev_{\mathcal{X}}.

It follows from TheoremΒ 3.1 that p𝒳p_{\mathcal{X}} satisfies pπ’³βˆ˜p𝒳=p𝒳p_{\mathcal{X}}\circ p_{\mathcal{X}}=p_{\mathcal{X}} and p𝒳​(x)=xp_{\mathcal{X}}(x)=x iff x∈emb𝒳⁑(Δ𝒳)x\in\emb_{\mathcal{X}}(\Delta_{\mathcal{X}}). Hence we view p𝒳p_{\mathcal{X}} as a retraction of XX onto the image of the embedding emb𝒳:Δ𝒳→X\emb_{\mathcal{X}}:\Delta_{\mathcal{X}}\to X.

Lemma 3.4.

The retraction map p𝒳p_{\mathcal{X}} satisfies the following properties:

  • (i)

    c𝒳​(x)∈{c𝒳​(p𝒳​(x))}Β―c_{\mathcal{X}}(x)\in\overline{\{c_{\mathcal{X}}(p_{\mathcal{X}}(x))\}} for all x∈Xx\in X.

  • (ii)

    Ο†D∘p𝒳=Ο†D\varphi_{D}\circ p_{\mathcal{X}}=\varphi_{D} for all D∈Div0⁑(𝒳)𝐑D\in\Div_{0}(\mathcal{X})_{\mathbf{R}}.

Proof.

By definition of c𝒳c_{\mathcal{X}} we have c𝒳​(x)∈Eic_{\mathcal{X}}(x)\in E_{i} for a given i∈Ii\in I iff ⟨ev𝒳⁑(x),Ei⟩>0\langle\ev_{\mathcal{X}}(x),E_{i}\rangle>0, and it follows that ev𝒳⁑(x)\ev_{\mathcal{X}}(x) lies in the relative interior of the simplex ΟƒJ\sigma_{J} for the maximal JβŠ‚IJ\subset I such that c𝒳​(x)∈EJc_{\mathcal{X}}(x)\in E_{J}. PropertyΒ (b) in TheoremΒ 3.1 then shows that c𝒳​(p𝒳​(x))c_{\mathcal{X}}(p_{\mathcal{X}}(x)) is the generic point of EJE_{J}, which proves (i).

Let us prove (ii). For each x∈Xx\in X we have

Ο†D​(p𝒳​(x))=⟨D,ev𝒳⁑(p𝒳​(x))⟩=⟨D,evπ’³βˆ˜embπ’³βˆ˜ev𝒳⁑(x)⟩=⟨D,ev𝒳⁑(x)⟩=Ο†D​(x),\varphi_{D}(p_{\mathcal{X}}(x))=\langle D,\ev_{\mathcal{X}}(p_{\mathcal{X}}(x))\rangle=\langle D,\ev_{\mathcal{X}}\circ\emb_{\mathcal{X}}\circ\ev_{\mathcal{X}}(x)\rangle=\langle D,\ev_{\mathcal{X}}(x)\rangle=\varphi_{D}(x),

using the identity evπ’³βˆ˜emb𝒳=id\ev_{\mathcal{X}}\circ\emb_{\mathcal{X}}=\id.

∎

Proposition 3.5.

If 𝒳β‰₯𝒴\mathcal{X}\geq\mathcal{Y} are two SNC models, then

  • (i)

    evπ’΄βˆ˜p𝒳=ev𝒴\ev_{\mathcal{Y}}\circ p_{\mathcal{X}}=\ev_{\mathcal{Y}} and pπ’΄βˆ˜p𝒳=p𝒴p_{\mathcal{Y}}\circ p_{\mathcal{X}}=p_{\mathcal{Y}}.

  • (ii)

    pπ’³βˆ˜emb𝒴=emb𝒴p_{\mathcal{X}}\circ\emb_{\mathcal{Y}}=\emb_{\mathcal{Y}}.

Note that (ii) says that the image in XX of Δ𝒴\Delta_{\mathcal{Y}} is contained in the image of Δ𝒳\Delta_{\mathcal{X}}.

Proof.

(i) amounts to the fact that Ο†D∘p𝒳=Ο†D\varphi_{D}\circ p_{\mathcal{X}}=\varphi_{D} for all D∈Div0⁑(𝒴)D\in\Div_{0}(\mathcal{Y}), which is a special case of Lemma 3.4.

Let us now prove (ii). The map emb𝒴′:=pπ’³βˆ˜emb𝒴:Δ𝒴→X\emb^{\prime}_{\mathcal{Y}}:=p_{\mathcal{X}}\circ\emb_{\mathcal{Y}}:\Delta_{\mathcal{Y}}\to X is continuous, and the previous identity implies that evπ’΄βˆ˜emb𝒴′=evπ’΄βˆ˜emb𝒴=id\ev_{\mathcal{Y}}\circ\emb^{\prime}_{\mathcal{Y}}=\ev_{\mathcal{Y}}\circ\emb_{\mathcal{Y}}=\id. By the uniqueness part of TheoremΒ 3.1 it suffices to prove that cπ’΄βˆ˜emb𝒴′=cπ’΄βˆ˜emb𝒴c_{\mathcal{Y}}\circ\emb^{\prime}_{\mathcal{Y}}=c_{\mathcal{Y}}\circ\emb_{\mathcal{Y}} on Δ𝒴\Delta_{\mathcal{Y}}. Pick sβˆˆΞ”π’΄s\in\Delta_{\mathcal{Y}} and set x:=emb𝒴⁑(s)x:=\emb_{\mathcal{Y}}(s), xβ€²:=emb𝒴′⁑(s)x^{\prime}:=\emb^{\prime}_{\mathcal{Y}}(s). On the one hand (i) shows that

p𝒴​(xβ€²)=pπ’΄βˆ˜pπ’³βˆ˜emb𝒴⁑(s)=pπ’΄βˆ˜emb𝒴⁑(s)=x,p_{\mathcal{Y}}(x^{\prime})=p_{\mathcal{Y}}\circ p_{\mathcal{X}}\circ\emb_{\mathcal{Y}}(s)=p_{\mathcal{Y}}\circ\emb_{\mathcal{Y}}(s)=x,

so c𝒴​(xβ€²)∈{c𝒴​(x)}Β―c_{\mathcal{Y}}(x^{\prime})\in\overline{\{c_{\mathcal{Y}}(x)\}} by (i) of LemmaΒ 3.4. On the other hand p𝒳​(x)=xβ€²p_{\mathcal{X}}(x)=x^{\prime} by definition, so c𝒳​(x)∈{c𝒳​(xβ€²)}Β―c_{\mathcal{X}}(x)\in\overline{\{c_{\mathcal{X}}(x^{\prime})\}} and hence c𝒴​(x)∈{c𝒴​(xβ€²)}Β―c_{\mathcal{Y}}(x)\in\overline{\{c_{\mathcal{Y}}(x^{\prime})\}} by continuity of the map 𝒳→𝒴\mathcal{X}\to\mathcal{Y} for the Zariski topology. ∎

Definition 3.6.

We define the subset XqmβŠ‚XX^{\mathrm{qm}}\subset X of quasimonomial points as

Xqm:=⋃𝒳emb𝒳⁑(Δ𝒳),X^{\mathrm{qm}}:=\bigcup_{\mathcal{X}}\emb_{\mathcal{X}}(\Delta_{\mathcal{X}}),

where 𝒳\mathcal{X} ranges over SNC models of 𝒳\mathcal{X}.

Corollary 3.7.

We have lim𝒳p𝒳=id\lim_{\mathcal{X}}p_{\mathcal{X}}=\id pointwise on XX. Hence XqmX^{\mathrm{qm}} is dense in XX.

Of course, we already knew from CorollaryΒ 2.4 that XdivβŠ‚XqmX^{\mathrm{div}}\subset X^{\mathrm{qm}} is dense in XX.

Proof.

By CorollaryΒ 3.2 it suffices to show that lim𝒳ev∘p𝒳=ev\lim_{\mathcal{X}}\ev\circ p_{\mathcal{X}}=\ev, which amounts to proving lim𝒳evπ’΄βˆ˜p𝒳=ev𝒴\lim_{\mathcal{X}}\ev_{\mathcal{Y}}\circ p_{\mathcal{X}}=\ev_{\mathcal{Y}} for each 𝒴\mathcal{Y}. This follows from (i) of PropositionΒ 3.5. ∎

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