No description available.

Considerando germes de hipersuperfícies em \'C POT.n+1\' com conjunto singular de dimensão s, Massey em [14] introduz um conjunto de (s+1) números chamados de números de Lê. Estes números se mostram como a generalização natural do número de Milnor para singularidades isoladas. O principal objetivo deste trabalho é mostrar como estes números são obtidos e que os números de Lê de uma hipersuperfície singular estão relacionados com os números de Lê de uma certa sequência de hipersuperfícies singulares \'X IND.0\',...,\'X IND.s-1\' que se aproxima da singularidade original de tal forma que os conjuntos críticos de seus termos \'X IND.i\' têm dimensão i. Essas relações são dadas pelas fórmulas de Lê-Iomdine. / For any germ of hypersurface in \'C POT. n+1\' with singular set of dimension s, Massey in [14] introduces a set of (s+1) numbers called Lê numbers. These numbers are a natural generalization of the Milnor number for isolated singularity hypersurfaces. The main goal of this work is to show how to obtain these numbers and to show the Lê numbers of a singular hypersurface are related with the the Lê numbers of a sequence of singular hypersurfaces \'X IND.0\',...,\'X IND.s-1\' which approach the original singularity in such a way that the critical set of each \'X IND.i\' has dimension i. These relationship are given by the Lê-Iomdine formulas.

Considerando germes de hipersuperfícies em \'C POT.n+1\' com conjunto singular de dimensão s, Massey em [14] introduz um conjunto de (s+1) números chamados de números de Lê. Estes números se mostram como a generalização natural do número de Milnor para singularidades isoladas. O principal objetivo deste trabalho é mostrar como estes números são obtidos e que os números de Lê de uma hipersuperfície singular estão relacionados com os números de Lê de uma certa sequência de hipersuperfícies singulares \'X IND.0\',...,\'X IND.s-1\' que se aproxima da singularidade original de tal forma que os conjuntos críticos de seus termos \'X IND.i\' têm dimensão i. Essas relações são dadas pelas fórmulas de Lê-Iomdine. / For any germ of hypersurface in \'C POT. n+1\' with singular set of dimension s, Massey in [14] introduces a set of (s+1) numbers called Lê numbers. These numbers are a natural generalization of the Milnor number for isolated singularity hypersurfaces. The main goal of this work is to show how to obtain these numbers and to show the Lê numbers of a singular hypersurface are related with the the Lê numbers of a sequence of singular hypersurfaces \'X IND.0\',...,\'X IND.s-1\' which approach the original singularity in such a way that the critical set of each \'X IND.i\' has dimension i. These relationship are given by the Lê-Iomdine formulas.

Das erste Thema dieser Dissertation ist der Defekt projektiver Hyperflächen. Es scheint, dass Hyperflächen mit Defekt einen verhältnismäßig großen singulären Ort besitzen. Diese Aussage wird im ersten Kapitel der Dissertation präzisiert und für Hyperflächen mit beliebigen isolierten Singularitäten über einem Körper der Charakteristik null, sowie für gewisse Klassen von Hyperflächen in positiver Charakteristik bewiesen. Darüber hinaus lässt sich die Dichte von Hyperflächen ohne Defekt über einem endlichen Körper abschätzen. Schließlich wird gezeigt, dass eine nicht-faktorielle Hyperfläche der Dimension drei mit isolierten Singularitäten stets Defekt besitzt. Das zweite Kapitel der Dissertation behandelt Bertini-Sätze über endlichen Körpern, aufbauend auf Poonens Formel für die Dichte glatter Hyperflächenschnitte in einer glatten Umgebungsvarietät. Diese wird auf quasiglatte Hyperflächen in simpliziellen torischen Varietäten verallgemeinert. Die Hauptanwendung ist zu zeigen, dass Hyperflächen mit einem in Relation zum Grad großen singulären Ort die Dichte null haben. Weiterhin enthält das Kapitel einen Bertini-Irreduzibilitätssatz, der auf einer Arbeit von Charles und Poonen beruht. Im dritten Kapitel werden ebenfalls Dichten über endlichen Körpern untersucht. Zunächst werden gewisse Faserungen über glatten projektiven Basisvarietäten in einem gewichteten projektiven Raum betrachtet. Das erste Resultat ist ein Bertini-Satz für glatte Faserungen, der Poonens Formel über glatte Hyperflächen impliziert. Der letzte Abschnitt behandelt elliptische Kurven über einem Funktionskörper einer Varietät der Dimension mindestens zwei. Die zuvor entwickelten Techniken ermöglichen es, eine untere Schranke für die Dichte solcher Kurven mit Mordell-Weil-Rang null anzugeben. Dies verbessert ein Ergebnis von Kloosterman. / The first topic of this dissertation is the defect of projective hypersurfaces. It is indicated that hypersurfaces with defect have a rather large singular locus. In the first chapter of this thesis, this will be made precise and proven for hypersurfaces with arbitrary isolated singularities over a field of characteristic zero, and for certain classes of hypersurfaces in positive characteristic. Moreover, over a finite field, an estimate on the density of hypersurfaces without defect is given. Finally, it is shown that a non-factorial threefold hypersurface with isolated singularities always has defect. The second chapter of this dissertation deals with Bertini theorems over finite fields building upon Poonen’s formula for the density of smooth hypersurface sections in a smooth ambient variety. This will be extended to quasismooth hypersurfaces in simplicial toric varieties. The main application is to show that hypersurfaces admitting a large singular locus compared to their degree have density zero. Furthermore, the chapter contains a Bertini irreducibility theorem for simplicial toric varieties generalizing work of Charles and Poonen. The third chapter continues with density questions over finite fields. In the beginning, certain fibrations over smooth projective bases living in a weighted projective space are considered. The first result is a Bertini-type theorem for smooth fibrations, giving back Poonen’s formula on smooth hypersurfaces. The final section deals with elliptic curves over a function field of a variety of dimension at least two. The techniques developed in the first two sections allow to produce a lower bound on the density of such curves with Mordell-Weil rank zero, improving an estimate of Kloosterman.