NAG-002287 · Accepted whole

The degree of a subtorus-invariant residual cycle

5 Tropicalization, multiplicities and balancing

Open in the English reader · PDF page 947

Statement · exact TeX
\begin{theorem}[The degree of a subtorus-invariant residual cycle]
\label{v036:projection:invariantweight}
In the preceding theorem let $L\subset N$ be a saturated
sublattice of rank $d$. Suppose every reduced integral support
$Z$ of $\mathcal C_d(D,h)$ is stable, as a closed subscheme,
under translation by the split subtorus $\mathbb T_L$.
Let $\rho:\mathbb T_N\to\mathbb T_{N/L}$ be the quotient map.
There are distinct closed points $s_1,\ldots,s_t$ of
$\mathbb T_{N/L}$ and positive integers $c_1,\ldots,c_t$ with
\[
                  \mathcal C_d(D,h)
                     =\sum_{j=1}^t c_j[\rho^{-1}(s_j)].
\]
The empty sum is allowed. For every finite projection $Q$ for
which $Q|_L$ is injective, put
\[
                         e_Q=[\mathbf Z^d:Q(L)].
\]
Then
\begin{equation}
\label{v036:projection:eqinvariantweight}
        \delta_Q=e_Q\mu,
        \qquad \mu=\sum_{j=1}^t c_j[\kappa(s_j):k]\in\mathbf Z_{\geq0}.
\end{equation}
In particular, $\delta_Q/e_Q$ is independent of all these
projections and is positive exactly when the residual cycle
is nonzero.
\end{theorem}
Full selected proof · exact TeX
\begin{proof}
Lemma~\ref{cand:elementarylatticesplitting} gives a splitting
$N=L\oplus N_1$. Put $M_0=(N/L)^\vee$ and $M_L=L^\vee$.
Then $k[M]=k[M_0][M_L]$, and translation by $\mathbb T_L$
is the character grading by $M_L$.

Let $I\subset k[M]$ be the prime ideal of an integral support
$Z$. The scheme-theoretic invariance hypothesis says that
the coaction sends $I$ to $I\otimes_k k[M_L]$. If
$f=\sum_\alpha f_\alpha\chi^\alpha\in I$, with
$f_\alpha\in k[M_0]$, independence of the characters in the
second tensor factor shows that each
$f_\alpha\chi^\alpha$ belongs to $I$. Multiplication by the
inverse character then puts $f_\alpha$ in $I\cap k[M_0]$.
Thus
\[
 I=(I\cap k[M_0])k[M],\qquad
 k[Z]=(k[M_0]/(I\cap k[M_0]))[M_L].
\]
The coefficient algebra is a finite-type domain. Since
$\dim Z=d$, the dimension--transcendence-degree equality
\cite[Tag 00P0]{StacksAlg} makes its fraction field algebraic
over $k$. Its finitely many algebra generators are therefore
algebraic over $k$. They span a finite-dimensional domain by
their monic equations, and this domain is a field. It is
$\kappa(s)$ for a closed point $s$ of $\mathbb T_{N/L}$.
This proves that $Z=\rho^{-1}(s)$. Distinct integral supports
give distinct points, establishing the first formula.

Fix one of these points, and write $k'=\kappa(s)$.
Under the splitting its fibre has coordinate ring $k'[M_L]$.
The target characters for $Q$ have the form
\[
                         U_i\longmapsto d_i\chi^{a_i},
                  \qquad d_i\in(k')^\times.
\]
The $a_i$ generate a sublattice of $M_L$ of index $e_Q$;
equivalently, this is the dual index of the injective
map $Q|_L$. Choose one representative for each coset.
Every character has a unique expression as a representative
times a product of the $\chi^{a_i}$, and the units $d_i$
can be absorbed into its coefficient. These representatives
are therefore a basis of $k'[M_L]$ over $k'[U^{\pm1}]$.
A $k$-basis of $k'$ gives a basis of size $e_Q[k':k]$
over $k[U^{\pm1}]$. This is the same character-coset
calculation as in Proposition~\ref{v035:degreeinitialweight}.
The generic field degree of $\rho^{-1}(s)\to
\mathbb T_{\mathbf Z^d}$ is consequently $e_Q[k':k]$.
Substitution in \eqref{v036:projection:eqdegreecycle} of Theorem~\ref{v036:projection:intrinsiccycle} gives
\eqref{v036:projection:eqinvariantweight}. Full residue-field
degrees, including inseparable degrees, enter this computation.
The positive coefficients $c_j$ already contain the unreduced
generic lengths from the actual models.
\end{proof}
Standing context 1 · exact TeX
Let $N$ be a lattice, $M=N^\vee$, and
$\mathbb T_N=\operatorname{Spec}k[M]$. A normalized moment tuple
on $D$ is a homomorphism
\[
                     h:M\longrightarrow D^\times
\]
such that every character value and its inverse are power
bounded. A basis and its inverses suffice for this condition.
For $Q:N\to\mathbf Z^d$, put
\[
 A_Q=K\langle U_1^{\pm1},\ldots,U_d^{\pm1}\rangle,
 \qquad U_j\longmapsto h(Q^*e_j).
\]
We consider those $Q$ for which this map is finite and the given
norm on $D$ is equivalent to the canonical finite-$A_Q$-module
norm. Write $H_Q$ for the completed residue field of the Gauss
point of $\mathcal M(A_Q)$, and put
\[
                 \delta_Q=\dim_{H_Q}(D\widehat\otimes_{A_Q}H_Q).
\]
An actual model is \emph{compatible with $h$} if it contains
the values of a basis of $M$ and their inverses.
Typeset reader excerpt
Typeset mathematical passage from PDF page 948
PDF page 948

Typed proof inputs

Named external registrations
Reader cross-references

Proof map

Verified map inputs · NAG-002287

AND · joint registered proof inputs

Unlinked source pointers

These pointers have no operative linked edge in this map.

  • APP-STACKS-00P0@NAG-002287:15670
  • STACKS-00P0