Rigid cohomology over Laurent series fields (Q265731): Difference between revisions

Let \(k\) be a finite field of characteristic \(p\). Let \(\mathcal{V}\) be a complete discrete valuation ring with residue field \(k\) and fraction field \(K\) of characteristic 0. Several rings of \(p\)-adic power series are classically associated to these data: the Robba ring \(\mathcal{R}_{K}\) (Laurent series with coefficients in \(K\) converging on some open annulus \(\{ \eta < |t| < 1\}\) with \(\eta\) unprescribed), the bounded Robba ring \(\mathcal{E}^\dagger_{K}\) (subring of \(\mathcal{R}_{K}\) where the series have bounded coefficients) and the Amice ring \(\mathcal{E}_{K}\) (completion of \(\mathcal{E}^\dagger_{K}\) with respect to the \(p\)-adic topology). When moding out by the maximal ideal of \(\mathcal{V}\), all those rings give \(k((t))\). Note that neither \(\mathcal{R}_{K}\) nor \(\mathcal{E}_{K}\) contains the other and that the only way to relate them involves passing through \(\mathcal{E}^\dagger_{K}\). Recall also that there exists a theory of \(p\)-adic cohomology for varieties over the field of Laurent series \(k((t))\), namely Berthelot's rigid cohomology (see [Cohomologie rigide et cohomologie rigide supports propres, Première partie, Prépublication IRMAR 96--03, Université de Rennes (1996)] or [\textit{B. Le Stum}, Rigid cohomology. Cambridge: Cambridge University Press (2007; Zbl 1131.14001)]), and that it actually takes values in \(\mathcal{E}_{K}\). The aim of the book under review is to construct a new cohomology theory for varieties over \(k((t))\) that takes values in the smaller ring \(\mathcal{E}^\dagger_{K}\) and gives back the classical rigid cohomology after tensoring with \(\mathcal{E}_{K}\). One immediate advantage of the theory is that one may also tensor with \(\mathcal{R}_{K}\), which is a ring of arithmetic importance. For instance, the \(p\)-adic analogue of the local monodromy theorem takes place over this ring. The book consists of 5 chapters, the first one being the introduction. In Chapter 2, the authors construct cohomology groups \(H^i_{\text{rig}}(-/\mathcal{E}^\dagger_{K})\) for varieties over \(k((t))\) that are embeddable (which roughly speaking means that such a variety \(X\) may be embedded in the special fiber of a formal \(\mathcal{V}[[t]]\)-scheme (for the \(p\)-adic topology) that is smooth around \(X\)). They closely follow Berthelot's construction except for the fact that they use Huber's adic spaces instead of Tate's rigid spaces. They also introduce a category of relative coefficients (isocrystals) over \(\mathcal{E}^\dagger_{K}\). Chapter 3 contains a proof of a \(p\)-adic local monodromy theorem following that of Kedlaya in the classical setting (see [Finiteness of rigid cohomology with coefficients. Duke Math. J. 134, No. 1, 15-97 (2006; Zbl 1133.14019)]). Several results follow, including the finite dimensionality of the groups \(H^i_{\text{rig}}(X/\mathcal{E}^\dagger_{K})\) for \(X\) is smooth and embeddable, as well as the fact that tensoring with \(\mathcal{E}_{K}\) gives back the classical rigid cohomology groups. Chapter 4 is devoted to cohomological descent and the aim is to extend the definition of the groups \(H^i_{\text{rig}}(-/\mathcal{E}^\dagger_{K})\) to any variety over \(k((t))\). In order to do so, the authors define an overconvergent site following \textit{B. Le Stum} [Mém. Soc. Math. Fr., Nouv. Sér. 127, 108 p. (2011; Zbl 1246.14028)] and use \textit{D. Zureick-Brown}'s strategy to prove the cohomological descent (see [Res. Math. Sci. 1, Paper No. 8, 20 p. (2014; Zbl 1349.14073)]). In Chapter 5, the authors show that their cohomology groups may be endowed with Gauss-Manin connections, resulting in a structure of \((\varphi,\nabla)\)-module. Using \textit{A. Marmora}'s construction [Compos. Math. 144, No. 2, 439--483 (2008; Zbl 1162.12003)], they can then associate \(p\)-adic Weil-Deligne representations. The book ends with applications to the monodromy-weight conjecture and independence of \(\ell\) results. The book is thorough and very carefully written, with useful appendices on classical rigid cohomology, adic spaces and cohomological descent. Moreover, instead of deducing results from the known cases in classical rigid cohomology (when possible), the authors have the choice of writing down complete proofs in their setting. This makes the exposition clearer and the book self-contained. I believe that it will soon become a reference on the subject and that the theory exposed here will quickly develop further and find more applications.