Chapter 19 Exercises
Section 19.1 Set 1
Exercises Exercises
1.
Let \(k\) be a finite field of characteristic \(p\text{.}\) Define the map2.
Let \(k\) be a finite field of order \(q\) and fix a nontrivial additive character (homomorphism) \(\psi: k \to \CC^\times\text{.}\) For \(\chi: k^\times \to \CC^\times\) a nontrivial multiplicative character, define the Gauss sum3.
With notation as in Exercise 19.1.2, verify that \(G_{\psi^{-1}}(\chi) = \chi(-1) G_\psi(\chi)\text{.}\)4.
Fix a choice of \(\chi\) as in Exercise 19.1.2. For \(P(T) = T^n + P_{n-1} T^{n-1} + \cdots + P_0 \in k[T]\) a monic polynomial, define5.
Show that for \(n\) a nonnegative integer,6.
With notation as in Exercise 19.1.5, let \(k'\) be an extension of \(k\) of degree \(v\text{.}\) Let \(\psi': k' \to \CC^\times\) be the additive character given by \(\psi \circ \Trace_{k'/k}\text{.}\) Given \(\chi\text{,}\) let \(\chi'\) be the multiplicative character given by \(\chi \circ \Norm_{k'/k}\text{.}\) For \(P' \in k'[T]\) monic, define \(\lambda'\) by analogy with \(\lambda\text{.}\) For \(P \in k[T]\) monic irreducible, let \(P'\) run over the irreducible factors of \(P\) in \(k'[T]\text{.}\) Prove that7.
Using all of the above, deduce the Davenport–Hasse relationSection 19.2 Set 2
Throughout, let \(\FF_q\) denote a finite field of characteristic \(p\text{.}\)
Exercises Exercises
1.
For \(X\) an algebraic variety over \(\FF_q\text{,}\) we write the zeta function of \(X\) as \(Z(X, q^{-s})\) for2.
For \(X\) equal to the \(n\)-dimensional projective space over \(\FF_q\text{,}\) compute that3.
Prove that the following statements are equivalent.The power series \(Z(X,T)\) represents a rational function in \(T\text{.}\)
There exist \(\alpha_1,\dots,\alpha_r, \beta_1,\dots,\beta_s \in \CC\) such that
\begin{equation*} \#X(\FF_{q^n}) = \alpha_1^n + \cdots + \alpha_r^n - \beta_1^n - \cdots - \beta_s^n \qquad (n=1,2,\dots). \end{equation*}
4.
Let \(X\) be the Grassmannian of \(k\)-dimensional subspaces of \(m\)-space over \(\FF_q\text{.}\)Compute \(\#X(\FF_{q^n})\text{;}\) your answer should be a polynomial in \(q^n\) depending on \(k\) and \(m\text{.}\)
Compute \(Z(X,T)\text{.}\)
5.
Choose \(a_0,\dots,a_r \in \FF_q^\times\text{.}\) For \(d\) a positive integer dividing \(q-1\text{,}\) let \(X_d\) be the projective hypersurface \(a_0 x_0^d + \cdots + a_r x_r^d = 0\text{.}\)Let \(G_d\) be the group of homomorphisms \(\chi: \FF_q^\times \to \CC^\times\) of order \(d\text{.}\) For \(\chi \in G_d\text{,}\) extend the definition of \(\chi\) to \(\FF_q\) by setting \(\chi(0) = 1\) if \(\chi=1\) and \(\chi(0) = 0\) otherwise. Show that
\begin{equation*} 1 + (q-1)\#X_d(\FF_q) = \sum_{(u_0,\dots,u_r) \in X_1} \sum_{\chi_0,\dots,\chi_r \in G_d} \prod_{i=0}^r \chi_i(u_i). \end{equation*}Show that if \(\chi_0,\dots,\chi_r \in G_d\) are neither all equal to 1 nor all distinct from 1, then
\begin{equation*} \sum_{(u_0,\dots,u_d) \in X_1} \prod_{i=0}^r \chi_i(u_i) = 0. \end{equation*}Let \(T\) be the set of tuples \((\chi_0,\dots,\chi_r) \in G_d \setminus \{1\}\) with \(\chi_0 \cdots \chi_r = 1\text{.}\) For \((\chi_0,\dots,\chi_r) \in T\text{,}\) define the Jacobi sum
\begin{equation*} j(\chi_0,\dots\chi_r) = \frac{1}{q-1} \sum_{u_0,\dots,u_r \in \FF_q: u_0 + \cdots + u_r = 0} \chi_0(u_0) \cdots \chi_r(u_r). \end{equation*}Show that\begin{equation*} \#X_d(\FF_q) = 1 + q + \cdots + q^{r-1} + \sum_{(\chi_0,\cdots,\chi_r) \in T} \chi_0(a_0^{-1}) \cdots \chi_r(a_r^{-1}) j(\chi_0,\dots,\chi_r). \end{equation*}Fix an additive character \(\psi: \FF_q \to \CC^\times\text{.}\) Show that
\begin{equation*} j(\chi_0,\dots,\chi_r) = \frac{1}{q} G(\chi_0,\psi) \cdots G(\chi_r, \psi) \end{equation*}where \(G(\chi, \psi)\) denotes the Gauss sum Exercise 19.1.2.
Show that \(\#X_d(\FF_q) = \#X_e(\FF_q)\) for \(e = \gcd(d, q-1)\text{.}\)
Using the Davenport–Hasse relation, show that the rationality, functional equation, and Riemann hypothesis hold for \(Z(X_d, T)\text{.}\)
Section 19.3 Set 3
Throughout, let \(\FF_q\) denote a finite field of characteristic \(p\text{.}\) Assume the Weil conjectures for curves and abelian varieties unless otherwise specified.
Exercises Exercises
1.
Let \(X\) be a nonzero abelian variety over \(\FF_q\text{.}\) Prove that if \(q \geq 5\text{,}\) then the group \(X(\FF_q)\) is nontrivial.2.
Let \(X\) be a curve over \(\FF_q\) such that \(\#X(\FF_q) = 1\text{.}\)If \(q=3\) or \(q=4\text{,}\) prove that
\begin{equation*} Z(X,T) = \frac{1 - qT + qT^2}{(1-T)(1-qT)}. \end{equation*}If \(=2\text{,}\) prove that the genus of \(X\) is at most 4, and that there are at most 6 possibilities for \(Z(X,T)\text{.}\)
Show that each of the 8 possibilities occurs for a unique \(X\) up to isomorphism.
3.
Let \(X\) be an abelian variety of dimension \(g\) over \(\FF_q\text{.}\) Assuming only the existence of complex numbers \(\alpha_1,\dots,\alpha_{2g}\) such that4.
Using the Honda–Tate theorem, prove that if \(A_1, A_2\) are abelian varieties over \(\FF_q\) and \(P_1(A_1, T)\) divides \(P_1(A_2, T)\text{,}\) then \(A_1\) is isogenous to the product of \(A_2\) with some other abelian variety.5.
Let \(X\) be a curve of genus \(g\) over \(\FF_q\text{.}\) Prove the following refinement of the Weil bound due to Serre:6.
Let \(P(T) = \sum_{i=0}^{2g} a_i T^i\) be a polynomial over \(\ZZ\) such that \(a_0 = 1\text{,}\) \(a_{g+i} = q^i a_{g-i}\) for all \(i\text{,}\) all roots of \(P(T)\) in \(\CC\) lie on the circle \(|T| = q^{-1/2}\text{,}\) and in addition \(a_g\) is not divisible by \(p\) (that is, \(P\) is an ordinary Weil polynomial). Use the Honda–Tate theorem to show that \(P(T)\) occurs as \(P_1(A,T)\) for some abelian variety \(A\) over \(\FF_q\) (without raising \(P\) to a power).Section 19.4 Set 4
Exercises Exercises
1.
Let \(K\) be a number field. Using the Chebotarëv density theorem, prove that the Frobenius elements corresponding to maximal ideals of \(\mathfrak{o}_K\) are dense in the absolute Galois group \(G_K\text{.}\)2.
In this exercise, we prove Theorem 11.1.4.Let \(f(T) = \sum_{n=0}^\infty a_n T^n\) be a power series over an arbitrary field \(K\text{.}\) Prove that \(f(T)\) represents a rational function over \(K\) if and only if for some positive integer \(m\text{,}\) the determinants of the \((m+1) \times (m+1)\) matrices \(A_{n,m} = (a_{n+i+j})_{i,j=0}^m\) vanish for all sufficiently large \(n\text{.}\)
Let \(f(T) = \sum_{n=0}^\infty a_n T^n\) be a power series over \(\ZZ\text{.}\) Let \(r>0\) be a real number such that over \(\QQ_p\text{,}\) there exists a polynomial \(P(T)\) of degree \(d\lt m\) such that \(P(T)f(T)\) converges for \(|T| \lt r+\epsilon\) for some \(\epsilon > 0\text{.}\) (We do not assume that \(P\) has coefficients in \(\ZZ\text{.}\)) Prove that for some \(C > 0\text{,}\) \(\left| \det(A_{n,m}) \right|_p \leq C r^{-n(m-d)}\) for all \(n\text{.}\)
Let \(f(T) = \sum_{n=0}^\infty a_n T^n\) be a power series over \(\ZZ\text{.}\) Let \(R\) and \(r\) be real numbers with \(Rr > 1\) such that over \(\CC\text{,}\) \(f(T)\) converges for \(|T| \lt R\text{;}\) and over \(\QQ_p\text{,}\) \(f(T)\) is the ratio of two series that converge for \(|T| \lt r\text{.}\) Prove that \(f\) represents a rational function.
3.
Let \(\pi\) be an element of an algebraic closure of \(\QQ_p\) satisfying \(\pi^{p-1} = -p\text{.}\) (You may use without proof the fact that \(\ZZ_p[\pi]\) is a discrete valuation ring with maximal ideal \((\pi)\text{.}\)) Define the power seriesProve that \(E_\pi(T) \in 1 + \pi \ZZ_p[\pi] \llbracket T \rrbracket\text{.}\)
Prove that \(E_\pi(T)\) has radius of convergence strictly greater than 1. In particular, it makes sense to evaluate it at any element of \(\ZZ_p[\pi]\text{.}\)
Prove that if \(t \in \ZZ_p\) satisfies \(t^p = t\text{,}\) then \(E_\pi(t)^p = 1\text{.}\)
4.
With notation as in Exercise 19.4.3, let \(n\) be a positive integer and define5.
Set \(q = p^n\) and letSection 19.5 Set 5
Exercises Exercises
1.
Define the ringsLet \(V_1, V_2\) be two finite-dimensional vector spaces over \(F\) equipped with endomorphisms \(\varphi_1,\varphi_2\) satisfying, for some positive integer \(n\text{,}\)
\begin{align*} \det(1-\varphi_1 T, V_1)^{-1} &\equiv 1 + x_1 T + \cdots + x_n T^n \pmod{T^{n+1} F \llbracket T \rrbracket}, \\ \det(1-\varphi_2 T, V_2)^{-1} &\equiv 1 + y_1 T + \cdots + y_n T^n \pmod{T^{n+1} F \llbracket T \rrbracket}. \end{align*}Prove that\begin{equation*} \det(1 - (\varphi_1 \otimes \varphi_2) T, V_1 \otimes_F V_2)^{-1} \equiv f \pmod{T^{n+1} F \llbracket T \rrbracket}. \end{equation*}Deduce that \(f \in R \llbracket T \rrbracket\text{.}\)
2.
Prove that there is a unique (up to unique natural isomorphism) functor \(\Lambda\) from rings to rings with the following properties.The underlying functor from rings to additive groups takes \(R\) to \(\Lambda(R) = 1 + T R \llbracket T \rrbracket\) with the usual series multiplication.
For any ring \(R\text{,}\) the multiplication map \(\ast\) on \(\Lambda(R)\) satisfies
\begin{equation*} (1-aT)^{-1} \ast (1-bT)^{-1} = (1-abT)^{-1} \qquad (a,b \in R). \end{equation*}
3.
Let \(X_1,X_2\) be two varieties over \(\FF_q\text{.}\) Prove that in \(\Lambda(\ZZ)\text{,}\) we have4.
Let \(K\) be a field of characteristic 0. Let \(P(x) \in K[x]\) be a monic polynomial of degree \(2g+1\) with no repeated roots.Let \(X\) be the affine scheme \(\Spec K[x,y]/(y^2-P(x))\text{.}\) Prove that \(\Omega^1_{X/K}\) is freely generated by \(dx/y\text{.}\)
Prove that \(H^1_{\dR}(X)\) admits the basis
\begin{equation*} x^i \frac{dx}{y} \qquad (i=0,\dots,2g-1). \end{equation*}Let \(Y\) be the affine scheme \(\Spec K[x,y,z]/(y^2-P(x),yz-1)\text{.}\) Prove that \(H^1_{\dR}(Y)\) admits the basis
\begin{equation*} x^i \frac{dx}{y}, \qquad (i=0,\dots,2g-1); \qquad x^i \frac{dx}{y^2} \qquad (i=0,\dots,2g). \end{equation*}
5.
Let \(p>2\) be a prime. Let \(\overline{P} \in \FF_p[x]\) be a monic polynomial of degree \(2g+1\) with no repeated roots.Put \(\overline{X} = \Spec \FF_p[x,y]/(y^2 - \overline{P}(x))\text{.}\) Prove that \(H^1_{\MW}(\overline{X})\) admits the basis
\begin{equation*} x^i \frac{dx}{y} \qquad (i=0,\dots,2g-1). \end{equation*}Put \(\overline{Y} = \Spec \FF_p[x,y,z]/(y^2-\overline{P}(x),yz-1)\text{.}\) Prove that \(H^1_{\MW}(\overline{Y})\) admits the basis
\begin{equation*} x^i \frac{dx}{y}, \qquad (i=0,\dots,2g-1); \qquad x^i \frac{dx}{y^2} \qquad (i=0,\dots,2g). \end{equation*}
Section 19.6 Supplementary exercises
These exercises were not assigned during the course, but were added subsequently.