In mathematics, the Poincaré residue is a generalization, to several complex variables and complex manifold theory, of the residue at a pole of complex function theory. It is just one of a number of such possible extensions.
Given a hypersurface
X\subsetPn
d
F
n
\omega
Pn
k>0
X
\operatorname{Res}(\omega)\inHn-1(X;C)
n=1
When Poincaré first introduced residues[1] he was studying period integrals of the form
wherefor\underset{\Gamma}\iint\omega
\Gamma\in
2 H 2(P -D)
\omega
D
wherefor\int\gammaRes(\omega)
\gamma\inH1(D)
\Gamma=T(\gamma)
\gamma
\varepsilon
\gamma
D*
on an affine chart where\omega=
q(x,y)dx\wedgedy p(x,y)
p(x,y)
N
\degq(x,y)\leqN-3
which are both cohomologous forms.Res(\omega)=-
qdx \partialp/\partialy =
qdy \partialp/\partialx
Given the setup in the introduction, let
p | |
A | |
k(X) |
p
Pn
k
d
d:
p-1 | |
A | |
k-1 |
(X)\to
p | |
A | |
k(X) |
Define
l{K}k(X)=
| |||||||||
|
as the rational de-Rham cohomology groups. They form a filtration
corresponding to the Hodge filtration.l{K}1(X)\subsetl{K}2(X)\subset … \subsetl{K}n(X)= Hn+1(Pn+1-X)
Consider an
(n-1)
\gamma\inHn-1(X;C)
T(\gamma)
\gamma
\gamma x S1
X
n
n
\omega
\intT(-)\omega:Hn-1(X;C)\toC
then we get a linear transformation on the homology classes. Homology/cohomology duality implies that this is a cohomology class
\operatorname{Res}(\omega)\inHn-1(X;C)
which we call the residue. Notice if we restrict to the case
n=1
1
P1
Res:Hn(Pn\setminusX)\toHn-1(X)
There is a simple recursive method for computing the residues which reduces to the classical case of
n=1
1
\operatorname{Res}\left( | dz |
z |
+a\right)=1
If we consider a chart containing
X
w
n
X
dw | |
wk |
\wedge\rho
Then we can write it out as
1 | |
(k-1) |
\left(
d\rho | |
wk-1 |
+d\left(
\rho | |
wk-1 |
\right)\right)
This shows that the two cohomology classes
\left[
dw | |
wk |
\wedge\rho\right]=\left[
d\rho | |
(k-1)wk-1 |
\right]
are equal. We have thus reduced the order of the pole hence we can use recursion to get a pole of order
1
\omega
\operatorname{Res}\left(\alpha\wedge
dw | |
w |
+\beta\right)=\alpha|X
For example, consider the curve
X\subsetP2
Ft(x,y,z)=t(x3+y3+z3)-3xyz
Then, we can apply the previous algorithm to compute the residue of
\omega=
\Omega | |
Ft |
=
xdy\wedgedz-ydx\wedgedz+zdx\wedgedy | |
t(x3+y3+z3)-3xyz |
Since
\begin{align} -zdy\wedge\left(
\partialFt | |
\partialx |
dx+
\partialFt | |
\partialy |
dy+
\partialFt | |
\partialz |
dz\right)&=z
\partialFt | |
\partialx |
dx\wedgedy-z
\partialFt | |
\partialz |
dy\wedgedz\\ ydz\wedge\left(
\partialFt | |
\partialx |
dx+
\partialFt | |
\partialy |
dy+
\partialFt | |
\partialz |
dz\right)&=-y
\partialFt | |
\partialx |
dx\wedgedz-y
\partialFt | |
\partialy |
dy\wedgedz\end{align}
3Ft-z
\partialFt | |
\partialx |
-y
\partialFt | |
\partialy |
=x
\partialFt | |
\partialx |
we have that
\omega=
ydz-zdy | |
\partialFt/\partialx |
\wedge
dFt | |
Ft |
+
3dy\wedgedz | |
\partialFt/\partialx |
This implies that
\operatorname{Res}(\omega)=
ydz-zdy | |
\partialFt/\partialx |