The Brauer-Manin obstruction for sections of the fundamental group (Q531343): Difference between revisions

The main object of the paper under review is the following exact sequence of profinite groups: \[ 1\to \pi_1(X\times_kk^{\text{{sep}}},\bar x) \to \pi_1(X,\bar x) \to G_k \to 1 \] where \(k\) is an arbitrary field, \(k^{\text{{sep}}}\) is its separable closure, \(X\) is a geometrically connected variety over \(k\), \(\bar x\in X\) is a geometric point, \(\pi_1\) stands for the fundamental group, and \(G_k\) is the absolute Galois group of \(k\). This sequence is split by a section \(s_a: G_k\to\pi_1(X)\) as soon as \(X\) has a \(k\)-point \(a\). In the case where \(X\) is a smooth projective curve of genus at least 2 and \(k/\mathbb Q\) is finitely generated, Grothendieck's conjecture asserts that the correspondence \(a\mapsto s_a\) is a bijection of \(X(k)\) with the set of \(\pi_1(X\times_kk^{\text{{sep}}})\)-conjugacy classes of sections of \(\pi_1(X,k)\). The author proposes, in the case where \(k\) is a number field, an analogue of the well-known Brauer--Manin obstruction for an adelic point to be global. Namely, he introduces the notion of an adelic section and shows the existence of a Brauer--Manin obstruction for it to be global. He applies this obstruction to show that the open subset \(U\) of the Reichardt--Lind curve defined by \[ 2y^2=x^4-17, y\neq 0, \] does not admit a section of its fundamental group \(\pi_1(U/\mathbb Q)\), thus confirming the section conjecture in this case. He then extends this approach to arithmetic twists of the Reichardt--Lind curve and isotrivial families of such twists. He notes that the method fails for the Selmer curve \(3X^3+4Y^3+5Z^3=0\). Some conditional results are provided, stating that the local version of the section conjecture, together with the conjecture that the Brauer--Manin obstruction is the only obstruction to the existence of rational points on curves, imply the section conjecture for curves over number fields.