lectures_on_knot_theory/lec_27_05.tex

141 lines
4.2 KiB
TeX

???????
\begin{theorem}
Such a pairing is isometric to a pairing:
\[
\begin{bmatrix}
1
\end{bmatrix}
\times
\begin{bmatrix}
1
\end{bmatrix}
\rightarrow
\frac{\epsilon}{p^k_{\xi}},
\: \epsilon \in {\pm 1}
\]
\end{theorem}
?????????????
\[
\begin{bmatrix}
1
\end{bmatrix} = 1 \in \quot{\Lambda}{p^k_{\xi} \Lambda }
\]
????????
\begin{theorem}
The jump of the signature function at $\xi$ is equal to
$2 \sum\limits_{k_i \odd} \epsilon_i$. \\
The peak of the signature function is equal to $\sum\limits_{k_i \even} \epsilon_i$.
\[
(\quot{\Lambda}{p^{k_1} \Lambda}, \epsilon_1) \oplus \dots \oplus (\quot{\Lambda}{p^{k_n} \Lambda}, \epsilon_n)
\]
%$(\eta_{k, \xi_l^{+}} -\eta_{k, \xi_l^{-}}$
\end{theorem}
\begin{definition}
A matrix $A$ is called Hermitian if
$\overline{A(t)} = {A(t)}^T$
\end{definition}
\begin{theorem}[Borodzik-Friedl 2015, Borodzik-Conway-Politarczyk 2018]
A square Hermitian matrix $A(t)$ of size $n$ with coefficients in $\mathbb{Z}[t, t^{-1}]$
(or $\mathbb{R}[t, t^{-1}]$ ) represents
the Blanchfield pairing if:
\begin{eqnarray*}
H_1(\bar{X}, \Lambda) = \quot{\Lambda^n }{A\Lambda^n },\\
(x, y) \mapsto {\overline{x}}^T A^{-1} y \in \quot{\Omega}{\Lambda}\\
H_1(\widetilde{X}, \Lambda) \times
H_1(\widetilde{X}, \Lambda) \longrightarrow
\quot{\Omega}{\Lambda},
\end{eqnarray*}
where $\Lambda = \mathbb{Z}[t, t^{-1}]$ or $\mathbb{R}[t, t^{-1}]$, $\Omega = \mathbb{Q}(t)$ or $\mathbb{R}(t)$
\end{theorem}
????????\\field of fractions ??????
\begin{eqnarray*}
H_1(\Sigma(K), \mathbb{Z}) = \quot{\mathbb{Z}^n}{(V + V^T) \mathbb{Z}^n}\\
H_1(\Sigma(K), \mathbb{Z})
\times
H_1(\Sigma(K), \mathbb{Z})
\longrightarrow
= \quot{\mathbb{Q}}{\mathbb{Z}}\\
(a, b) \mapsto a{(V + V^T)}^{-1} b
\end{eqnarray*}
???????????????????\\
???????????????????\\
\begin{eqnarray*}
y \mapsto y + Az \\
\overline{x^T} A^{-1}(y + Az) = \overline{x^T} A^{-1} y + \overline{x^T} \mathbb{1} z = \overline{x^T} A^{-1}y \in \quot{\Omega}{\Lambda} \\
\overline{x^T} \mathbb{1} z \in \Lambda \\
H_1(\widetilde{X}, \Lambda) =
\quot{ \Lambda^n }{(Vt - V) \Lambda^n}
\\
(a, b) \mapsto \overline{a^T}(Vt - V^T)^{-1} (t -1)b
\end{eqnarray*}
(Blanchfield '59)
\begin{theorem}[Kearton '75, Friedl, Powell '15]
There exits a matrix $A$ representing the Blanchfield paring over $\mathbb{Z}[t, t^{-1}]$. The size of $A$ is a size of Seifert form.
\end{theorem}
Remark:
\begin{enumerate}
\item
Over $\mathbb{R}$ we can take $A$ to be diagonal.
\item
The jump of signature function at $\xi$ is
equal to
\[
\lim_{t \rightarrow 0^+} \sign A (e^{it} \xi) - \sign A(e^{-it} \xi).
\]
\item
The minimal size of a matrix $A$ that presents a Blanchfield paring (over $\mathbb{Z}[t, t^{-1}]$) for a knot $K$ is a knot invariant.
\end{enumerate}
\subsection{The unknotting number}
Let $K$ be a knot and $D$ a knot diagram. A crossing change is a modification of a knot diagram by one of following changes
\begin{align*}
\PICorientpluscross \mapsto \PICorientminuscross ,\\
\PICorientminuscross \mapsto\PICorientminuscross.
\end{align*}
The unknotting number $u(K)$ is a number of crossing changes needed to turn a knot into an unknot, where the minimum is taken over all diagrams of a given knot.
\begin{definition}
A Gordian distance $G(K, K^\prime)$ between knots $K$ and $K^\prime$ is the minimal number of crossing changes required to turn $K$ into $K^\prime$.
\end{definition}
\begin{problem}
Prove that:
\[
G(K, K^{\prime\prime})
\leq
G(K, K^{\prime})
+
G(K^\prime, K^{\prime\prime}).
\]
Open problem:
\[
u(K\# K^\prime) = u(K) + u(K^\prime).
\]
\end{problem}
\begin{lemma}[Scharlemann '84]
Unknotting number one knots are prime.
\end{lemma}
\subsection*{Tools to bound unknotting number}
\begin{theorem}
For any symmetric polynomial $\Delta \in \mathbb{Z}[t, t^{-1}]$ such that $\Delta(1) = 1$, there exists a knot $K$ such that:
\begin{enumerate}
\item
$K$ has unknotting number $1$,
\item
$\Delta_K = \Delta$.
\end{enumerate}
\end{theorem}
Let us consider a knot $K$ and its Seifert surface $\Sigma$.
the Seifert form for $K_-$
\\
the Seifert form for $K_+$
\\
$S_- + S_+$ differs from
by a term in the bottom right corner
Let $A$ be a symmetric $n \times n$ matrix over $\mathbb{R}$. Let $A_1, \dots, A_n$ be minors of $A$. \\
Let $\epsilon_0 = 1$
If