Namespaces
Variants
Actions

Difference between revisions of "Riemann-Roch theorem"

From Encyclopedia of Mathematics
Jump to: navigation, search
m (link)
m (fixing spaces)
 
(2 intermediate revisions by 2 users not shown)
Line 1: Line 1:
A theorem expressing the [[Euler characteristic|Euler characteristic]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819801.png" /> of a [[Locally free sheaf|locally free sheaf]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819802.png" /> on an algebraic or analytic variety <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819803.png" /> in terms of the characteristic Chern classes of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819804.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819805.png" /> (cf. [[Chern class|Chern class]]). It can be used to calculate the dimension of the space of sections of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819806.png" /> (the Riemann–Roch problem).
+
<!--
 +
r0819801.png
 +
$#A+1 = 96 n = 2
 +
$#C+1 = 96 : ~/encyclopedia/old_files/data/R081/R.0801980 Riemann\ANDRoch theorem
 +
Automatically converted into TeX, above some diagnostics.
 +
Please remove this comment and the {{TEX|auto}} line below,
 +
if TeX found to be correct.
 +
-->
  
The classical Riemann–Roch theorem relates to the case of non-singular algebraic curves <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819807.png" /> and states that for any [[Divisor|divisor]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819808.png" /> on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r0819809.png" />,
+
{{TEX|auto}}
 +
{{TEX|done}}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198010.png" /></td> <td valign="top" style="width:5%;text-align:right;">(1)</td></tr></table>
+
A theorem expressing the [[Euler characteristic|Euler characteristic]]  $  \chi ( {\mathcal E} ) $
 +
of a [[Locally free sheaf|locally free sheaf]]  $  {\mathcal E} $
 +
on an algebraic or analytic variety  $  X $
 +
in terms of the characteristic Chern classes of  $  {\mathcal E} $
 +
and  $  X $ (cf. [[Chern class|Chern class]]). It can be used to calculate the dimension of the space of sections of  $  {\mathcal E} $ (the Riemann–Roch problem).
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198011.png" /> is the dimension of the space of functions <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198012.png" /> for which <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198013.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198014.png" /> is the [[canonical divisor]] and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198015.png" /> is the genus of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198016.png" />. In the middle of the 19th century B. Riemann used analytic methods to obtain the inequality
+
The classical Riemann–Roch theorem relates to the case of non-singular algebraic curves  $  X $
 +
and states that for any [[Divisor|divisor]] $  D $
 +
on  $  X $,
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198017.png" /></td> </tr></table>
+
$$ \tag{1 }
 +
l( D) - l( K _ {X} - D)  =   \mathop{\rm deg}  D - g + 1,
 +
$$
 +
 
 +
where  $  l( D) = \mathop{\rm dim}  H  ^ {0} ( X, {\mathcal O} _ {X} ( D)) $
 +
is the dimension of the space of functions  $  f \in k( x) $
 +
for which  $  ( f  ) + D \geq  0 $,
 +
$  K _ {X} $
 +
is the [[canonical divisor]] and  $  g $
 +
is the genus of  $  X $.  
 +
In the middle of the 19th century B. Riemann used analytic methods to obtain the inequality
 +
 
 +
$$
 +
l( D)  \geq    \mathop{\rm deg}  D - g + 1.
 +
$$
  
 
The equality (1) was proved by E. Roch.
 
The equality (1) was proved by E. Roch.
  
The Riemann–Roch theorem for curves is the one-dimensional case of the more general Riemann–Roch–Hirzebruch–Grothendieck theorem. Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198018.png" /> be a non-singular projective variety of dimension <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198019.png" />, and let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198020.png" /> be an appropriate [[Cohomology|cohomology]] theory: either <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198021.png" /> are singular cohomology spaces when the basic field <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198022.png" />, or <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198023.png" /> where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198024.png" /> is a [[Chow ring|Chow ring]], or <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198025.png" /> is the ring associated to the Grothendieck ring <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198026.png" /> (see [[#References|[2]]], [[#References|[7]]]). Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198027.png" /> be a locally free sheaf of rank <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198028.png" /> on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198029.png" />. Universal polynomials for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198030.png" /> with rational coefficients, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198031.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198032.png" />, in the Chern classes <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198033.png" /> of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198034.png" /> are defined in the following way. The factorization
+
The Riemann–Roch theorem for curves is the one-dimensional case of the more general Riemann–Roch–Hirzebruch–Grothendieck theorem. Let $  X $
 +
be a non-singular projective variety of dimension $  n $,  
 +
and let $  H ^ {. } X $
 +
be an appropriate [[Cohomology|cohomology]] theory: either $  H ^ {. } X = H ^ {. } ( X, \mathbf Q ) $
 +
are singular cohomology spaces when the basic field $  k = \mathbf C $,  
 +
or $  H ^ {. } X = A( X) \otimes \mathbf Q $
 +
where $  A( x) $
 +
is a [[Chow ring|Chow ring]], or $  H ^ {. } X $
 +
is the ring associated to the Grothendieck ring $  K  ^ {0} ( X) $ (see [[#References|[2]]], [[#References|[7]]]). Let $  {\mathcal E} $
 +
be a locally free sheaf of rank r $
 +
on $  X $.  
 +
Universal polynomials for $  {\mathcal E} $
 +
with rational coefficients, $  \mathop{\rm ch} (-) $
 +
and $  \mathop{\rm td} (-) $,  
 +
in the Chern classes $  c _ {i} ( {\mathcal E} ) \in H ^ {. } X $
 +
of $  {\mathcal E} $
 +
are defined in the following way. The factorization
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198035.png" /></td> </tr></table>
+
$$
 +
c _ {t} ( E)  = c _ {0} ( {\mathcal E} ) + \dots + c _ {r} ( {\mathcal E} ) t  ^ {r}  = \
 +
\prod _ { i= 1} ^ { r }  ( 1 + a _ {i} t)
 +
$$
  
is examined for the Chern polynomial, where the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198036.png" /> are formal symbols. The exponential Chern character is defined by the formula
+
is examined for the Chern polynomial, where the $  a _ {i} $
 +
are formal symbols. The exponential Chern character is defined by the formula
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198037.png" /></td> </tr></table>
+
$$
 +
\mathop{\rm ch} ( {\mathcal E} )  = \sum _ { i= 1} ^ { r }  e ^ {a _ {i} } \  \left ( e  ^ {x} = 1 + x +
 +
\frac{1}{2!}
 +
x  ^ {2} + \dots \right ) ,
 +
$$
  
 
and, accordingly, the [[Todd class|Todd class]] is defined as
 
and, accordingly, the [[Todd class|Todd class]] is defined as
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198038.png" /></td> </tr></table>
+
$$
 +
\mathop{\rm td} ( {\mathcal E} )  = \prod _ { i= 1} ^ { r } 
 +
\frac{a _ {i} }{1 - e ^ {- a _ {i} } }
 +
;
 +
$$
  
<img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198039.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198040.png" /> are symmetric functions in the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198041.png" /> and they can be written as polynomials in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198042.png" />.
+
$  \mathop{\rm ch} ( {\mathcal E} ) $
 +
and $  \mathop{\rm td} ( {\mathcal E} ) $
 +
are symmetric functions in the $  a _ {i} $
 +
and they can be written as polynomials in $  c _ {i} ( {\mathcal E} ) $.
  
The Riemann–Roch–Hirzebruch theorem: If <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198043.png" /> is a non-singular projective variety or a compact complex variety of dimension <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198044.png" /> and if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198045.png" /> is a vector bundle of rank <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198046.png" /> on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198047.png" />, then
+
The Riemann–Roch–Hirzebruch theorem: If $  X $
 +
is a non-singular projective variety or a compact complex variety of dimension $  n $
 +
and if $  {\mathcal E} $
 +
is a vector bundle of rank r $
 +
on $  X $,  
 +
then
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198048.png" /></td> </tr></table>
+
$$
 +
\chi ( {\mathcal E} )  =   \mathop{\rm deg} (  \mathop{\rm ch} ( {\mathcal E} )  \mathop{\rm td} ( {\mathcal T} _ {X} )) _ {n} ,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198049.png" /> is the tangent sheaf on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198050.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198051.png" /> denotes the component of degree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198052.png" /> in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198053.png" />. This theorem was proved by F. Hirzebruch in the case of the ground field <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198054.png" />. When <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198055.png" /> and the invertible sheaf <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198056.png" />, it leads to the equation
+
where $  {\mathcal T} _ {X} $
 +
is the tangent sheaf on $  X $
 +
and $  \mathop{\rm deg} ( \cdot ) _ {n} $
 +
denotes the component of degree $  n $
 +
in $  H ^ {. } X $.  
 +
This theorem was proved by F. Hirzebruch in the case of the ground field $  \mathbf C $.  
 +
When $  n = 2 $
 +
and the invertible sheaf $  {\mathcal E} = {\mathcal O} _ {X} ( D) $,  
 +
it leads to the equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198057.png" /></td> </tr></table>
+
$$
 +
\chi ( {\mathcal O} _ {X} ( D))  =
 +
\frac{1}{2}
 +
D( D - K _ {X} ) +
 +
\frac{1}{12}
 +
( K _ {X}  ^ {2}
 +
+ c _ {2} ),
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198058.png" /> is the second Chern class of the surface <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198059.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198060.png" /> is its canonical class. In particular, when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198061.png" /> Noether's formula is obtained:
+
where $  c _ {2} = c _ {2} ( X) $
 +
is the second Chern class of the surface $  X $
 +
and $  K _ {X} $
 +
is its canonical class. In particular, when $  D = 0 $
 +
Noether's formula is obtained:
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198062.png" /></td> </tr></table>
+
$$
 +
\chi ( {\mathcal O} _ {X} )  = 1 + p _ {a}  =
 +
\frac{1}{12}
 +
( K _ {X}  ^ {2} + c _ {2} ).
 +
$$
  
For three-dimensional varieties <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198063.png" /> the theorem leads to
+
For three-dimensional varieties $  ( n= 3) $
 +
the theorem leads to
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198064.png" /></td> </tr></table>
+
$$
 +
\chi ( {\mathcal O} _ {X} ( D))  =
 +
\frac{1}{6}
 +
D  ^ {3} -
 +
\frac{1}{4}
 +
D  ^ {2} K _ {X} +
 +
\frac{1}{12}
 +
D( K _ {X}  ^ {2} + c _ {2} ) -  
 +
\frac{1}{24}
 +
K _ {X} c _ {2} .
 +
$$
  
In particular, when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198065.png" />,
+
In particular, when $  D = 0 $,
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198066.png" /></td> </tr></table>
+
$$
 +
\chi ( {\mathcal O} _ {X} )  = -  
 +
\frac{1}{24}
 +
K _ {X} c _ {2} .
 +
$$
  
In 1957, A. Grothendieck generalized the Riemann–Roch–Hirzebruch theorem to the case of a morphism of non-singular varieties over an arbitrary algebraically closed field (see [[#References|[1]]]). Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198067.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198068.png" /> be the Grothendieck groups of the coherent and locally free sheafs on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198069.png" />, respectively (cf. [[Grothendieck group|Grothendieck group]]). The functor <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198070.png" /> is a covariant functor from the category of schemes and proper morphisms into the category of Abelian groups. In this case, for a [[Proper morphism|proper morphism]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198071.png" /> the morphism <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198072.png" /> is defined by the formula
+
In 1957, A. Grothendieck generalized the Riemann–Roch–Hirzebruch theorem to the case of a morphism of non-singular varieties over an arbitrary algebraically closed field (see [[#References|[1]]]). Let $  K _ {0} X $
 +
and $  K  ^ {0} X $
 +
be the Grothendieck groups of the coherent and locally free sheafs on $  X $,  
 +
respectively (cf. [[Grothendieck group|Grothendieck group]]). The functor $  K _ {0} X $
 +
is a covariant functor from the category of schemes and proper morphisms into the category of Abelian groups. In this case, for a [[Proper morphism|proper morphism]] $  f: X \rightarrow Y $
 +
the morphism $  f:  K _ {0} X \rightarrow K _ {0} Y $
 +
is defined by the formula
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198073.png" /></td> </tr></table>
+
$$
 +
f _ {!} ( {\mathcal F})  = \sum (- 1)  ^ {i} R  ^ {i} f _ {*} ( {\mathcal F} ),
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198074.png" /> is an arbitrary [[Coherent sheaf|coherent sheaf]] on <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198075.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198076.png" /> is a covariant functor into the category of rings. For regular schemes with an ample sheaf, the groups <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198077.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198078.png" /> coincide and are denoted by <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198079.png" />. The Chern character <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198080.png" /> is a homomorphism of rings; <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198081.png" /> is also a covariant functor: The Gizin homomorphism <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198082.png" /> is defined. When <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198083.png" />, the homomorphism <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198084.png" /> is obtained from <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198085.png" /> for homology spaces using [[Poincaré duality|Poincaré duality]]. The theorem as generalized by Grothendieck expresses the measure of deviation from commutativity of the homomorphisms <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198086.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198087.png" />.
+
where $  {\mathcal F} $
 +
is an arbitrary [[Coherent sheaf|coherent sheaf]] on $  X $
 +
and $  K  ^ {0} X $
 +
is a covariant functor into the category of rings. For regular schemes with an ample sheaf, the groups $  K _ {0} X $
 +
and $  K  ^ {0} X $
 +
coincide and are denoted by $  K( X) $.  
 +
The Chern character $  \mathop{\rm ch} : K( X) \rightarrow H ^ {. } X $
 +
is a homomorphism of rings; $  H ^ {. } X $
 +
is also a covariant functor: The Gizin homomorphism $  f _ {*} : H ^ {. } X \rightarrow H ^ {. } Y $
 +
is defined. When $  H ^ {. } X = H ^ {. } ( X, \mathbf Q ) $,  
 +
the homomorphism $  f _ {*} $
 +
is obtained from $  f _ {*} $
 +
for homology spaces using [[Poincaré duality|Poincaré duality]]. The theorem as generalized by Grothendieck expresses the measure of deviation from commutativity of the homomorphisms $  f _ {!} $
 +
and $  \mathop{\rm ch} $.
  
The Riemann–Roch–Hirzebruch–Grothendieck theorem: Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198088.png" /> be a smooth projective morphism of non-singular projective varieties; then for any <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198089.png" /> the equation
+
The Riemann–Roch–Hirzebruch–Grothendieck theorem: Let $  f : X \rightarrow Y $
 +
be a smooth projective morphism of non-singular projective varieties; then for any $  x \in K( X) $
 +
the equation
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198090.png" /></td> </tr></table>
+
$$
 +
\mathop{\rm ch} ( f _ {!} ( x))  = f _ {*} (  \mathop{\rm ch} ( x)  \mathop{\rm td} ( {\mathcal T} _ {f} ))
 +
$$
  
in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198091.png" /> is true, where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198092.png" /> (the relative tangent sheaf of the morphism <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198093.png" />).
+
in $  H ^ {. } X $
 +
is true, where $  {\mathcal T} _ {f} = {\mathcal T} _ {X} - f ^ { * } ( {\mathcal T} _ {Y} ) \in K _ {X} $ (the relative tangent sheaf of the morphism $  f  $).
  
When <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198094.png" /> is a point, this theorem reduces to the Riemann–Roch–Hirzebruch theorem. There are generalizations (see [[#References|[5]]], [[#References|[6]]], [[#References|[7]]]) when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198095.png" /> is a Noetherian scheme with an ample invertible sheaf, when <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198096.png" /> is a projective morphism whose fibres are locally complete intersections, and also to the case of singular quasi-projective varieties.
+
When $  Y $
 +
is a point, this theorem reduces to the Riemann–Roch–Hirzebruch theorem. There are generalizations (see [[#References|[5]]], [[#References|[6]]], [[#References|[7]]]) when $  Y $
 +
is a Noetherian scheme with an ample invertible sheaf, when $  f $
 +
is a projective morphism whose fibres are locally complete intersections, and also to the case of singular quasi-projective varieties.
  
 
Several versions of the Riemann–Roch theorem are closely connected with the index problem for elliptic operators (see [[Index formulas|Index formulas]]). For example, the Riemann–Roch–Hirzebruch theorem for compact complex varieties is a particular case of the Atiyah–Singer index theorem.
 
Several versions of the Riemann–Roch theorem are closely connected with the index problem for elliptic operators (see [[Index formulas|Index formulas]]). For example, the Riemann–Roch–Hirzebruch theorem for compact complex varieties is a particular case of the Atiyah–Singer index theorem.
Line 63: Line 196:
 
====References====
 
====References====
 
<table><TR><TD valign="top">[1]</TD> <TD valign="top"> A. Borel, J.-P. Serre, "La théorème de Riemann–Roch" ''Bull. Soc. Math. France'' , '''86''' (1958) pp. 97–136</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top"> Yu.I. Manin, "Lectures on the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198097.png" />-functor in algebraic geometry" ''Russian Math. Surveys'' , '''24''' : 5 (1969) pp. 1–89 ''Uspekhi Mat. Nauk'' , '''24''' : 5 (1969) pp. 3–86 {{MR|265355}} {{ZBL|}} </TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top"> R. Hartshorne, "Algebraic geometry" , Springer (1977) {{MR|0463157}} {{ZBL|0367.14001}} </TD></TR><TR><TD valign="top">[4]</TD> <TD valign="top"> F. Hirzebruch, "Topological methods in algebraic geometry" , Springer (1978) (Translated from German) {{MR|1335917}} {{MR|0202713}} {{ZBL|0376.14001}} </TD></TR><TR><TD valign="top">[5]</TD> <TD valign="top"> P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for singular varieties" ''Publ. Math. IHES'' , '''45''' (1975) pp. 101–145</TD></TR><TR><TD valign="top">[6]</TD> <TD valign="top"> P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for topological <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198098.png" />-theory and singular varieties" ''Acta Math.'' , '''143''' : 3–4 (1979) pp. 155–192</TD></TR><TR><TD valign="top">[7]</TD> <TD valign="top"> "Théorie des intersections et théorème de Riemann–Roch" P. Berthelot (ed.) et al. (ed.) , ''Sem. Geom. Alg. 6'' , ''Lect. notes in math.'' , '''225''' , Springer (1971)</TD></TR></table>
 
<table><TR><TD valign="top">[1]</TD> <TD valign="top"> A. Borel, J.-P. Serre, "La théorème de Riemann–Roch" ''Bull. Soc. Math. France'' , '''86''' (1958) pp. 97–136</TD></TR><TR><TD valign="top">[2]</TD> <TD valign="top"> Yu.I. Manin, "Lectures on the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198097.png" />-functor in algebraic geometry" ''Russian Math. Surveys'' , '''24''' : 5 (1969) pp. 1–89 ''Uspekhi Mat. Nauk'' , '''24''' : 5 (1969) pp. 3–86 {{MR|265355}} {{ZBL|}} </TD></TR><TR><TD valign="top">[3]</TD> <TD valign="top"> R. Hartshorne, "Algebraic geometry" , Springer (1977) {{MR|0463157}} {{ZBL|0367.14001}} </TD></TR><TR><TD valign="top">[4]</TD> <TD valign="top"> F. Hirzebruch, "Topological methods in algebraic geometry" , Springer (1978) (Translated from German) {{MR|1335917}} {{MR|0202713}} {{ZBL|0376.14001}} </TD></TR><TR><TD valign="top">[5]</TD> <TD valign="top"> P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for singular varieties" ''Publ. Math. IHES'' , '''45''' (1975) pp. 101–145</TD></TR><TR><TD valign="top">[6]</TD> <TD valign="top"> P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for topological <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/r/r081/r081980/r08198098.png" />-theory and singular varieties" ''Acta Math.'' , '''143''' : 3–4 (1979) pp. 155–192</TD></TR><TR><TD valign="top">[7]</TD> <TD valign="top"> "Théorie des intersections et théorème de Riemann–Roch" P. Berthelot (ed.) et al. (ed.) , ''Sem. Geom. Alg. 6'' , ''Lect. notes in math.'' , '''225''' , Springer (1971)</TD></TR></table>
 
 
  
 
====Comments====
 
====Comments====

Latest revision as of 03:10, 2 March 2022


A theorem expressing the Euler characteristic $ \chi ( {\mathcal E} ) $ of a locally free sheaf $ {\mathcal E} $ on an algebraic or analytic variety $ X $ in terms of the characteristic Chern classes of $ {\mathcal E} $ and $ X $ (cf. Chern class). It can be used to calculate the dimension of the space of sections of $ {\mathcal E} $ (the Riemann–Roch problem).

The classical Riemann–Roch theorem relates to the case of non-singular algebraic curves $ X $ and states that for any divisor $ D $ on $ X $,

$$ \tag{1 } l( D) - l( K _ {X} - D) = \mathop{\rm deg} D - g + 1, $$

where $ l( D) = \mathop{\rm dim} H ^ {0} ( X, {\mathcal O} _ {X} ( D)) $ is the dimension of the space of functions $ f \in k( x) $ for which $ ( f ) + D \geq 0 $, $ K _ {X} $ is the canonical divisor and $ g $ is the genus of $ X $. In the middle of the 19th century B. Riemann used analytic methods to obtain the inequality

$$ l( D) \geq \mathop{\rm deg} D - g + 1. $$

The equality (1) was proved by E. Roch.

The Riemann–Roch theorem for curves is the one-dimensional case of the more general Riemann–Roch–Hirzebruch–Grothendieck theorem. Let $ X $ be a non-singular projective variety of dimension $ n $, and let $ H ^ {. } X $ be an appropriate cohomology theory: either $ H ^ {. } X = H ^ {. } ( X, \mathbf Q ) $ are singular cohomology spaces when the basic field $ k = \mathbf C $, or $ H ^ {. } X = A( X) \otimes \mathbf Q $ where $ A( x) $ is a Chow ring, or $ H ^ {. } X $ is the ring associated to the Grothendieck ring $ K ^ {0} ( X) $ (see [2], [7]). Let $ {\mathcal E} $ be a locally free sheaf of rank $ r $ on $ X $. Universal polynomials for $ {\mathcal E} $ with rational coefficients, $ \mathop{\rm ch} (-) $ and $ \mathop{\rm td} (-) $, in the Chern classes $ c _ {i} ( {\mathcal E} ) \in H ^ {. } X $ of $ {\mathcal E} $ are defined in the following way. The factorization

$$ c _ {t} ( E) = c _ {0} ( {\mathcal E} ) + \dots + c _ {r} ( {\mathcal E} ) t ^ {r} = \ \prod _ { i= 1} ^ { r } ( 1 + a _ {i} t) $$

is examined for the Chern polynomial, where the $ a _ {i} $ are formal symbols. The exponential Chern character is defined by the formula

$$ \mathop{\rm ch} ( {\mathcal E} ) = \sum _ { i= 1} ^ { r } e ^ {a _ {i} } \ \left ( e ^ {x} = 1 + x + \frac{1}{2!} x ^ {2} + \dots \right ) , $$

and, accordingly, the Todd class is defined as

$$ \mathop{\rm td} ( {\mathcal E} ) = \prod _ { i= 1} ^ { r } \frac{a _ {i} }{1 - e ^ {- a _ {i} } } ; $$

$ \mathop{\rm ch} ( {\mathcal E} ) $ and $ \mathop{\rm td} ( {\mathcal E} ) $ are symmetric functions in the $ a _ {i} $ and they can be written as polynomials in $ c _ {i} ( {\mathcal E} ) $.

The Riemann–Roch–Hirzebruch theorem: If $ X $ is a non-singular projective variety or a compact complex variety of dimension $ n $ and if $ {\mathcal E} $ is a vector bundle of rank $ r $ on $ X $, then

$$ \chi ( {\mathcal E} ) = \mathop{\rm deg} ( \mathop{\rm ch} ( {\mathcal E} ) \mathop{\rm td} ( {\mathcal T} _ {X} )) _ {n} , $$

where $ {\mathcal T} _ {X} $ is the tangent sheaf on $ X $ and $ \mathop{\rm deg} ( \cdot ) _ {n} $ denotes the component of degree $ n $ in $ H ^ {. } X $. This theorem was proved by F. Hirzebruch in the case of the ground field $ \mathbf C $. When $ n = 2 $ and the invertible sheaf $ {\mathcal E} = {\mathcal O} _ {X} ( D) $, it leads to the equation

$$ \chi ( {\mathcal O} _ {X} ( D)) = \frac{1}{2} D( D - K _ {X} ) + \frac{1}{12} ( K _ {X} ^ {2} + c _ {2} ), $$

where $ c _ {2} = c _ {2} ( X) $ is the second Chern class of the surface $ X $ and $ K _ {X} $ is its canonical class. In particular, when $ D = 0 $ Noether's formula is obtained:

$$ \chi ( {\mathcal O} _ {X} ) = 1 + p _ {a} = \frac{1}{12} ( K _ {X} ^ {2} + c _ {2} ). $$

For three-dimensional varieties $ ( n= 3) $ the theorem leads to

$$ \chi ( {\mathcal O} _ {X} ( D)) = \frac{1}{6} D ^ {3} - \frac{1}{4} D ^ {2} K _ {X} + \frac{1}{12} D( K _ {X} ^ {2} + c _ {2} ) - \frac{1}{24} K _ {X} c _ {2} . $$

In particular, when $ D = 0 $,

$$ \chi ( {\mathcal O} _ {X} ) = - \frac{1}{24} K _ {X} c _ {2} . $$

In 1957, A. Grothendieck generalized the Riemann–Roch–Hirzebruch theorem to the case of a morphism of non-singular varieties over an arbitrary algebraically closed field (see [1]). Let $ K _ {0} X $ and $ K ^ {0} X $ be the Grothendieck groups of the coherent and locally free sheafs on $ X $, respectively (cf. Grothendieck group). The functor $ K _ {0} X $ is a covariant functor from the category of schemes and proper morphisms into the category of Abelian groups. In this case, for a proper morphism $ f: X \rightarrow Y $ the morphism $ f: K _ {0} X \rightarrow K _ {0} Y $ is defined by the formula

$$ f _ {!} ( {\mathcal F}) = \sum (- 1) ^ {i} R ^ {i} f _ {*} ( {\mathcal F} ), $$

where $ {\mathcal F} $ is an arbitrary coherent sheaf on $ X $ and $ K ^ {0} X $ is a covariant functor into the category of rings. For regular schemes with an ample sheaf, the groups $ K _ {0} X $ and $ K ^ {0} X $ coincide and are denoted by $ K( X) $. The Chern character $ \mathop{\rm ch} : K( X) \rightarrow H ^ {. } X $ is a homomorphism of rings; $ H ^ {. } X $ is also a covariant functor: The Gizin homomorphism $ f _ {*} : H ^ {. } X \rightarrow H ^ {. } Y $ is defined. When $ H ^ {. } X = H ^ {. } ( X, \mathbf Q ) $, the homomorphism $ f _ {*} $ is obtained from $ f _ {*} $ for homology spaces using Poincaré duality. The theorem as generalized by Grothendieck expresses the measure of deviation from commutativity of the homomorphisms $ f _ {!} $ and $ \mathop{\rm ch} $.

The Riemann–Roch–Hirzebruch–Grothendieck theorem: Let $ f : X \rightarrow Y $ be a smooth projective morphism of non-singular projective varieties; then for any $ x \in K( X) $ the equation

$$ \mathop{\rm ch} ( f _ {!} ( x)) = f _ {*} ( \mathop{\rm ch} ( x) \mathop{\rm td} ( {\mathcal T} _ {f} )) $$

in $ H ^ {. } X $ is true, where $ {\mathcal T} _ {f} = {\mathcal T} _ {X} - f ^ { * } ( {\mathcal T} _ {Y} ) \in K _ {X} $ (the relative tangent sheaf of the morphism $ f $).

When $ Y $ is a point, this theorem reduces to the Riemann–Roch–Hirzebruch theorem. There are generalizations (see [5], [6], [7]) when $ Y $ is a Noetherian scheme with an ample invertible sheaf, when $ f $ is a projective morphism whose fibres are locally complete intersections, and also to the case of singular quasi-projective varieties.

Several versions of the Riemann–Roch theorem are closely connected with the index problem for elliptic operators (see Index formulas). For example, the Riemann–Roch–Hirzebruch theorem for compact complex varieties is a particular case of the Atiyah–Singer index theorem.

References

[1] A. Borel, J.-P. Serre, "La théorème de Riemann–Roch" Bull. Soc. Math. France , 86 (1958) pp. 97–136
[2] Yu.I. Manin, "Lectures on the -functor in algebraic geometry" Russian Math. Surveys , 24 : 5 (1969) pp. 1–89 Uspekhi Mat. Nauk , 24 : 5 (1969) pp. 3–86 MR265355
[3] R. Hartshorne, "Algebraic geometry" , Springer (1977) MR0463157 Zbl 0367.14001
[4] F. Hirzebruch, "Topological methods in algebraic geometry" , Springer (1978) (Translated from German) MR1335917 MR0202713 Zbl 0376.14001
[5] P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for singular varieties" Publ. Math. IHES , 45 (1975) pp. 101–145
[6] P. Baum, W. Fulton, R. MacPherson, "Riemann–Roch for topological -theory and singular varieties" Acta Math. , 143 : 3–4 (1979) pp. 155–192
[7] "Théorie des intersections et théorème de Riemann–Roch" P. Berthelot (ed.) et al. (ed.) , Sem. Geom. Alg. 6 , Lect. notes in math. , 225 , Springer (1971)

Comments

There are analogues of the Riemann–Roch theorem in algebraic number theory and arithmetic algebraic geometry.

References

[a1] S. Lang, "Algebraic number theory" , Addison-Wesley (1970) MR0282947 Zbl 0211.38404
[a2] K. Szpiro, "Sem. sur les pinceaux arithmétiques: La conjecture de Mordell" Astérisque , 127 (1985)
How to Cite This Entry:
Riemann-Roch theorem. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Riemann-Roch_theorem&oldid=42129
This article was adapted from an original article by Val.S. Kulikov (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article