Exercises of Algebraic Geometry I

Michele Bolognesi
M ICHELE B OLOGNESI , IMAG, P LACE E UGENE B ATAILLON , 34095 M ONTPELLIER ,
F RANCE
E-mail address: [Link]@[Link]
 Contents

Introduction 4
Chapter 1. Affine varieties 5
Bibliography 9

3
 Introduction

These notes collect a series of solved exercises for the course of Algebraic [Link]
of them from the book Algebraic Geometry by R. Hartshorne [Har]. Many others from the
notes by Ph. Ellia [PhE]. I am very thankful to Alex Massarenti for providing most of the
resolutions presented in here.

4
 CHAPTER 1

Affine varieties

Exercise 1. [Har, Exercise 1.1]

(a) The coordinate ring of the curve C = {y − x2 = 0} ⊂ A2 is given by
A(C ) = k [ x, y]/(y − x2 ) ∼
= k[ x, x2 ] ∼
= k [ x ].
(b) A( Z ) = k [ x, y]/( xy − 1) is isomorphic to the localization of k [ x ] at x. Let f :
A( Z ) → k [ x ] be a morphism of k-algebras. Since x ∈ A( Z ) is invertible f ( x ) ∈ k.
Therefore, f can not be an isomorphism.

Exercise 2. [Har, Exercise 1.3] Consider Y = { x2 − yz = xz − x = 0} ⊂ A3 . Then

Y = { x 2 − y = z − 1 = 0} ∪ { x = y = 0} ∪ { x = z = 0},
and Y is the union of two lines and a plane irreducible curve of degree two. In order to
show that each component is irreducible, one shows that k [ x, y, z]/equations is an integral
domain, i.e. has no zero divisor.
In particular, the coordinate ring of each irreducible component is isomorphic to k [t].

Exercise 3. [Har, Exercise 1.5] Let B be a finitely generated k-algebra. Then we may
write B = k [ x1 , ..., xn ]/I for some ideal I = ( f 1 , ..., f r ) in k [ x1 , ..., xn ]. Let X = { f 1 = ... =
f r = 0} ⊆ An . Let f ∈ I ( X ) then, by the Nullstellensatz we have f k ∈ I for some k > 0.
Now, B does not have nilpotents, so f ∈ I. Clearly I ⊆ I ( X ). This yields I = I ( X ) and
B∼ = A ( X ).
Conversely, assume to have B = A( X ) for some algebraic set X ⊂ An . Let I ( X ) be the
ideal of X. Then B ∼ = k[ x1 , ..., xn ]/I ( X ) is a finitely generated k-algebra. Let f ∈ B be a
nilpotent element. Then f k = 0 for some k, that is f k ∈ I. Since I is radical we get f ∈ I,
that is f = 0 in B.

Exercise 4. [Har, Exercise 1.8] Let Y ⊂ An be an affine variety of dimension r. Let

H ⊂ An be an hypersurface such that Y is not contained in H and Y ∩ H 6= ∅. Since Y
is not contained in H we have I ( H ) * I (Y ). Let f be the polynomial defining H. Then,
the irreducible components of Y ∩ H corresponds to the minimal prime ideals of A(Y )
containing f . Note that Y * H implies that f is not a zero-divisor in A(Y ). By the Haup-
tidealsatz any minimal prime ideal containing f has height one. Finally, by [Har, Theorem
1.8A] we get that the any irreducible component of Y ∩ H has dimension dim(Y ) − 1.

Exercise 5. [Har, Exercise 1.9] Let a ⊆ k[ x1 , ..., xn ] be an ideal that can be generated by
r elements f 1 , ..., f r . Note that { f i = 0} defines an hypersurface for any i = 1, ..., r. We
apply r times Exercise 1.8 and we distinguish two cases:
5
6 1. AFFINE VARIETIES

- at any step the variety Hk = { f 1 = ... f k = 0} is not contained in the hypersurface

{ f k+1 = 0}. Then at each step the dimension of the intersection drops by one. We
get that the dimension of each irreducible component of Y is n − r,
- if Hk is contained in { f k+1 = 0} for some k, then the intersection with { f k+1 = 0}
will not drop the dimension. Then each irreducible component of Y has dimen-
sion greater than n − r.
In any case we have that the dimension of each irreducible component of Y is greater or
equal than n − r.

Exercise 6. [Har, Exercise 1.11] The curve Y is the image of the morphism

φ : A1 −→ A3
t 7−→ (t , t4 , t5 )
3

Note that since A1 is irreducible Y is irreducible as well. Therefore I = I (Y ) is prime. Fur-

thermore dim(Y ) = dim( A(Y )) = 1 and by [Har, Theorem 1.8A] we get height( I (Y )) = 2.
Note that the three polynomials z2 − x2 y, xz − y2 and yz − x3 are in I (Y ) and they are in-
dependent.
Let J = (z2 − x2 y, xz − y2 , yz − x3 ) ⊆ I (Y ). By [Ku, Page 138] we have that I (Y ) = J and
that we need three elements to generate I (Y ).

Exercise 7. [Har, Exercise 1.12] Consider the polynomial

f = ( x2 − 1 + iy)( x2 − 1 − iy) = x4 − 2x2 + y2 + 1.

Since R[ x, y] ⊂ C[ x, y] are unique factorization domains and f splits in C[ x, y] as a product

of two irreducible polynomials of degree two, we conclude that f is irreducible in R[ x, y].
On the other hand, Z ( f ) = {(1, 0), (−1, 0)} is the union of two points. Therefore f ∈
R[ x, y] is irreducible but Z ( f ) ⊂ A2 is reducible.

Exercise 8. [Har, Exercise 2.1]

a is homogeneous and hence defines a cone in An+1 . The polynomial f vanishes on all
the elements of this cone (including 0 since f has positive degree) so f t ∈ a for some t > 0
by the usual Nullstellensatz.

Exercise 9. [Har, Exercise 2.2]

(iii) implies (i) is trivial because all monomials xid belong to Sd . (i) implies (ii): If Z ( a)
√
is empty, then in An+1 ,√Z ( a) is either empty or (0, . . . , 0), so a must be S or the irrelevant
ideal. (ii) implies (iii): a contains xi , so there is some m s.t. xim ∈ a for all i, so a contains
Sm(n+1) as any monomial of degree m(n + 1) must have xim as a factor for some i.

√ Exercise 10. [Har, Exercise 2.3] (a),(b),(c),(e) are clear. For (d), clearly I ( Z ( a)) contains
a. Since Z ( a) is nonempty, any nonzero homogeneous polynomial vanishing on it must √
have positive degree. By 2.1, this implies that f t ∈ a. Therefore I ( Z ( a)) is contained in a
as it is a homogeneous ideal.
 1. AFFINE VARIETIES 7

Exercise 11. [Har, Exercise 2.4] (a) Follows from 2.3d,e, and 2.2.
(b) If Y = Y1 ∪ Y2 , then I (Y ) = I (Y1 ) ∩ I (Y2 ) ⊃ I (Y1 ) I (Y2 ). Therefore if I (Y ) is prime,
I (Y ) must be either I (Y1 ) or I (Y2 ), so Y is Y1 or Y2 . On the other hand if Y is not prime,
then ab ∈ I (Y ), with a 6∈ I (Y ), b 6∈ I (Y ). Therefore Y is the union of the proper subsets
Y ∩ Z ( a), Y ∩ Z (b) and is therefore not irreducible.
(c) I (Pn ) = 0 which is a prime ideal.

Exercise 12. [Har, Exercise 2.9] Let Y ⊆ An be an affine variety. Consider the homeo-
morphism
φ0 : U0 = Pn \ { x0 = 0} −→ An
x1
[ x0 : ... : xn ] 7−→ ( x0 , ..., xxn0 )
Finally, let Y be the projective closure of Y.
Let F ∈ I (Y ), then f (y1 , ..., yn ) = F (1, x0 , ..., xn ) where yi = xx0i vanishes on Y = Y ∩ U0 .
We get that f ∈ I (Y ) and x0s β( f ) = F for some s. Therefore, F ∈ ( β( I (Y ))), where β is the
homogeneization with respect to x0 .
Now let F ∈ β( I (Y )), then F = g1 β( f 1 ) + ... + gr β( f r ) for some f 1 , ..., f r ∈ I (Y ), that is
F = g1 x0s1 f 1 ( xx10 , ..., xxn0 ) + ... + gr x0sr f r ( xx01 , ..., xxn0 ). Hence F ∈ I (Y ).
Let Y ⊂ A3 be the affine twisted cubic. Then I (Y ) = ( x3 − z, x2 − y) while I (Y ) = ( xz −
y2 , yw − z2 , xw − yz). Note that I (Y ) can not be generated by two elements because Y ⊂ P3
is not a scheme-theoretic complete intersection.

Exercise 13. [Har, Exercise 2.10]

(a) Obvious.
(b) They have the same ideal, which is prime if and only if they are irreducible.
(c) By ex. 2.6 of [Har], we have S(Y )) = dim(Y ) + 1, and dim(Y ) = dim(S(Y )), hence the
claim.

Exercise 14. [Har, Exercise 2.13]

We assume that P2 is isomorphic to its image which is easy to check, so that curves in
the image of P2 correspond to curves in P2 .
The map is given by

( x0 : x1 : x2 ) → ( x02 : x12 : x22 : x0 x1 : x1 x2 : x2 x0 ).

Any curve in P2 is defined by some polynomial f ( x0 : x1 : x2 ) = 0, f homogeneous,
and therefore also by the polynomial f ( x0 ; x1 ; x2 )2 = g( x02 : x12 : x22 : x0 x1 : x1 x2 : x2 x0 ) for
some polynomial g. Then some factor of this polynomial g defines a suitable hypersurface
containing the image of the curve Z.

Exercise 15. [Har, Exercise 2.14]

The image of ψ is the set Y defined by the equations of the form x ab xcd = x ac xbd . Proof:
the image is clearly contained in Y . Conversely if ( x00 : x10 : · · · : xrs ) ∈ Y then we may
assume that x00 is nonzero. But then the point is the image of ( x00 : x10 : · · · : xr0 ) × ( x00 :
x01 : · · · : x0s ) ∈ Pr × Ps .
8 1. AFFINE VARIETIES

Exercise 15. [Har, Exercise 2.15]

a) ( a0 : a1 ) × (b0 : b1 ) = ( a0 b0 : a0 b1 : a1 b0 : a1 b1 ) = (w : x : y : z), and the image of
P1 × P1 is then the subvariety xy − zw = 0 as in Ex. 15.
b) Q is isomorphic to P1 × P1 , so we can take the two families of lines to correspond to
point × line and line × point. We check that these are lines inside Q ⊂ P3; for example the
image of ( a0 : a1 ) × P1 is the set of points (w : x : y : z) ∈ P3 with a1 w = a0 y, a1 x = a0 z.
THis is the intersection of two projective planes in P3 , that is a line.
c) For example, the closed subset x = y of Q is not one of these lines, and it is closed
in the Zariski topology of Q, hence Q is not homemorphic to P1 × P1 with the product
topology.
Exercise 16. [Har, Exercise 2.17]
a) By Ex. 1.8 of [Har], the intersection of q hypersurfaces has dimension at least n − q. If
a is generated by q elements then Z (Y ) is the intersection of q hypersurfaces and therefore,
by Ex. 2.8 of [Har], has dimension at least n − q.
b) If I (Y ) can be generated by r elements then Y is the intersection of their hypersur-
faces.
c) Y can be seen as the intersection of H1 = Z ( x2 − wy) and H2 = Z (y3 + wz2 −
2xyz) as one can obtain powers of the three quadrics that generate I (Y ) by combining H1
and H2 . For example ( xy − wz)2 = w(y3 + wz2 − 2xyz) + y2 ( x2 − wy) and (y2 − xz)2 =
y(y3 + wz2 − 2xyz) + z2 ( x2 − wy), and similarly for the third. By the way, I (Y ) has no
homogeneous elements of degree 0 or 1 and we know the space of homogeneous elements
of degree 2 is 3 dimensional, so any set of generators must have at least 3 elements.
d)I think that this is still unknown.
 Bibliography

[PhE] P H . E LLIA, Un’introduzione light alla geometria algebrica, [Link].

[Har] R. H ARTSHORNE, Algebraic geometry, Graduate Texts in Mathematics, no. 52, Springer-Verlag, New
York-Heidelberg, 1977.
[Ku] E. K UNZ, Introduction to Commutative Algebra and Algebraic Geometry, Birkhäuser, first edition, 1984.