In the first part of this thesis, we prove two density theorems for quadrature domains in Cn ,n≥2. It is shown that quadrature domains are dense in the class of all product domains of the form D×Ωwhere D⊂Cn−1 is a smoothly bounded pseudoconvex domain satisfying Bell’s Condition R and Ω⊂Cis a smoothly bounded domain. It is also shown that quadrature domains are dense in the class of all smoothly bounded complete Hartogs domains in C2. In the second part of this thesis, we study the behaviour of the critical points of the Green’s function when a sequence of domains Dk⊂Rn con-verges to a limiting domain Din the C∞-topology. It is shown that the limit-ing set of the critical points of the Green’s functions Gkfor domains Dk⊂Care the zeroes of the Bergman kernel of D. This generalizes a result of Solynin and Gustafsson, Sebbar.

On sait que le nombre de racines réelles d’un polynôme à une variable de degré d et à coefficients réels est compris entre 0 et d. Au début des années 90, E. Kostlan prouve que le nombre moyen de racines vaut racine carrée de d, lorsque ces polynômes sont équipées d’une mesure de probabilité adéquate. Ce résultat possède une interprétation géométrique, où les polynômes apparaissent comme sections au-dessus de la sphère de Riemann, et ils peuvent s’étendre au cadre plus général de sections de fibrés en droites amples sur une surface de Riemann. Il s’agit ici du calcul de l’espérance mathématique du nombre de racines réelles de ces polynômes ou sections. Dans cette thèse, on calcule tous les moments centrés de ces variable aléatoires. Comme application de ce calcul, on prouve que la mesure de l’ensemble des polynômes ou sections dont le nombre de racines s’ écartent de la moyenne est majoré de façon effective en fonction de cet écart, un résultat de type concentration de la mesure en probabilité. Dans une deuxième partie, on présente des résultats analogue dans la théorie de Hurwitz réelle, où plutôt que du nombre de racines réelles d’un polynôme aléatoire, on considère le nombre de points critiques réels d’un revêtement ramifié aléatoire de la sphère de Riemann. On calcul la moyenne et tous les moments centrés du nombre de points critiques réels d'un revêtement aléatoire.Les techniques employées dans la preuve de ces résultats sont de nature analytique (noyau de Bergman, estimées L^2) et géométriques (multi-espaces d'Olver, formule de la coaire) / It is well known that the number of real roots of a real degree d polynomial is at most d. In the 90s, E. Kostlan proved that the average number of real roots equals the square root of d, once we equip the space of polynomials with some natural Gaussian measure. This result has a geometric interpretation, in which the real polynomials are sections of a line bundle over the Riemann sphere. We can extend this study in a more general case of a real Riemann surface equipped with ample line bundle and study the expected value of the number of real zeros of a random section. In this thesis, we compute all the central moments of these random variables. As an application, we prove that the measure of the space of real sections whose number of real zeros deviates from the expected one goes to zeros, as the degree of the line bundle goes to infinity.In a second part, we present analogues results in real Hurwitz theory, in which we study the real critical points of a random branched covering of the Riemann sphere. We compute the expected value of this number and also all the central moments.The techniques we use are of analytique nature (Bergman kernel, L^2 estimates) and gometric one (Olver multispaces, coarea formula)

The aim of this dissertation is to give explicit descriptions of the set of proper holomorphic mappings between two complex manifolds with reasonable restrictions on the domain and target spaces. Without any restrictions, this problem is intractable even when posed for do-mains in . We give partial results for special classes of manifolds. We study, broadly, two types of structure results: Descriptive. The first result of this thesis is a structure theorem for finite proper holomorphic mappings between products of connected, hyperbolic open subsets of compact Riemann surfaces. A special case of our result follows from the techniques used in a classical result due to Remmert and Stein, adapted to the above setting. However, the presence of factors that have no boundary or boundaries that consist of a discrete set of points necessitates the use of techniques that are quite divergent from those used by Remmert and Stein. We make use of a finiteness theorem of Imayoshi to deal with these factors. Rigidity. A famous theorem of H. Alexander proves the non-existence of non-injective proper holomorphic self-maps of the unit ball in . ,n >1. Several extensions of this result for various classes of domains have been established since the appearance of Alexander’s result, and it is conjectured that the result is true for all bounded domains in . , n > 1, whose boundary is C2-smooth. This conjecture is still very far from being settled. Our first rigidity result establishes the non-existence of non-injective proper holomorphic self-maps of bounded, balanced pseudo convex domains of finite type (in the sense of D’Angelo) in ,n >1. This generalizes a result in 2, by Coupet, Pan and Sukhov, to higher dimensions. As in Coupet–Pan–Sukhov, the aforementioned domains need not have real-analytic boundaries. However, in higher dimensions, several aspects of their argument do not work. Instead, we exploit the circular symmetry and a recent result in complex dynamics by Opshtein. Our next rigidity result is for bounded symmetric domains. We prove that a proper holomorphic map between two non-planar bounded symmetric domains of the same dimension, one of them being irreducible, is a biholomorphism. Our methods allow us to give a single, all-encompassing argument that unifies the various special cases in which this result is known. Furthermore, our proof of this result does not rely on the fine structure (in the sense of Wolf et al.) of bounded symmetric domains. Thus, we are able to apply our techniques to more general classes of domains. We illustrate this by proving a rigidity result for certain convex balanced domains whose automorphism groups are assumed to only be non-compact. For bounded symmetric domains, our key tool is that of Jordan triple systems, which is used to describe the boundary geometry.

<p>We write down explicit algebraic formulas for the Szeg\H{o}, Garabedian and Bergman kernels for specific two-connected planar domains. We use these results to derive integral representations for a biholomorphic invariant relating the Bergman and Szeg\H{o} kernels. We use the formulas to study the asymptotic behavior of these kernels as a family of two-connected domains approaches the unit disc. We derive an explicit formula for the Green's function for the Laplacian for special values on two-connected domains. Every two-connected domain is biholomorphic to a unique two-connected domain of the type we consider. This allows one to write down formulas for the kernel functions on a general two-connected domain.</p>

<p>In the theory of holomorphic functions of one complex variable it is often useful to study subharmonic functions. The subharmonic can be described using the Laplace operator. When one studies holomorphic functions of several complex variables one should study the plurisubharmonic functions instead. Here the complex Monge--Ampère operator has a role similar to that of the Laplace operator in the theory of subharmonic functions. The complex Monge--Ampère operator is nonlinear and therefore it is not as well understood as the Laplace operator. We consider two types of boundary value problems for the complex Monge--Ampere equation in certain pseudoconvex domains. In this thesis the right-hand side in the Monge--Ampère equation will always be smooth, strictly positive and meet a monotonicity condition. The first type of boundary value problem we consider is a Dirichlet problem where we look for plurisubharmonic solutions which are zero on the boundary of the domain. We show that this problem has a unique smooth solution if the domain has a smooth bounded plurisubharmonic exhaustion function which is globally Lipschitz and has Monge--Ampère mass larger than one everywhere. We obtain some results on which domains have such a bounded exhaustion function. The second type of boundary value problem we consider is a boundary blow-up problem where we look for plurisubharmonic solutions which tend to infinity at the boundary of the domain. Here we also assume that the right-hand side in the Monge--Ampère equation satisfies a growth condition. We study this problem in strongly pseudoconvex domains with smooth boundary and show that it has solutions which are Hölder continuous with arbitrary Hölder exponent α, 0 ≤ α < 1. We also show a uniqueness result. A result on the growth of the solutions is also proved. This result is used to describe the boundary behavior of the Bergman kernel.</p>

In the theory of holomorphic functions of one complex variable it is often useful to study subharmonic functions. The subharmonic can be described using the Laplace operator. When one studies holomorphic functions of several complex variables one should study the plurisubharmonic functions instead. Here the complex Monge--Ampère operator has a role similar to that of the Laplace operator in the theory of subharmonic functions. The complex Monge--Ampère operator is nonlinear and therefore it is not as well understood as the Laplace operator. We consider two types of boundary value problems for the complex Monge--Ampere equation in certain pseudoconvex domains. In this thesis the right-hand side in the Monge--Ampère equation will always be smooth, strictly positive and meet a monotonicity condition. The first type of boundary value problem we consider is a Dirichlet problem where we look for plurisubharmonic solutions which are zero on the boundary of the domain. We show that this problem has a unique smooth solution if the domain has a smooth bounded plurisubharmonic exhaustion function which is globally Lipschitz and has Monge--Ampère mass larger than one everywhere. We obtain some results on which domains have such a bounded exhaustion function. The second type of boundary value problem we consider is a boundary blow-up problem where we look for plurisubharmonic solutions which tend to infinity at the boundary of the domain. Here we also assume that the right-hand side in the Monge--Ampère equation satisfies a growth condition. We study this problem in strongly pseudoconvex domains with smooth boundary and show that it has solutions which are Hölder continuous with arbitrary Hölder exponent α, 0 ≤ α &lt; 1. We also show a uniqueness result. A result on the growth of the solutions is also proved. This result is used to describe the boundary behavior of the Bergman kernel.

Dans cette thèse, nous étudions le volume et la caractéristique d'Euler de sous-variétés aléatoires de codimension r ∈ {1, . . . , n} dans une variété ambiante M de dimension n. Dans un premier modèle, dit des ondes riemanniennes aléatoires, M est une variété riemannienne fermée. Nous considérons alors le lieu Zλ des zéros communs de r combinaisons linéaires aléatoires indépendantes de fonctions propres du laplacien associées à des valeurs propres inférieures à λ 0. Nous obtenons alors les asymptotiques du volume moyen et de la caractéristique d'Euler moyenne de Zλ lorsque λ tend vers l'infini. Dans un second modèle, M est le lieu réel d'une variété projective définie sur les réels. On s'intéresse dans ce cadre au lieu d'annulation réel Zd d'une section holomorphe réelle globale aléatoire de E⊗Ld, où E est un fibré hermitien de rang r, L est un fibré en droites hermitien ample et tous deux sont définis sur les réels. Nous estimons alors les moyennes du volume et de la caractéristique d'Euler de Zd quand d tend vers l'infini. Dans ce modèle algébrique réel, nous calculons aussi l'asymptotique de la variance du volume de Zd pour 1 r < n. Nous en déduisons, dans ce cas, des résultats asymptotiques d'équidistribution de Zd dans M / We study the volume and Euler characteristic of codimension r ∈ {1, . . . , n} random submanifolds in a dimension n manifold M. First, we consider Riemannian random waves. That is M is a closed Riemannian manifold and we study the common zero set Zλ of r independent random linear combinations of eigenfunctions of the Laplacian associated to eigenvalues smaller than λ 0. We compute estimates for the mean volume and Euler characteristic of Zλ as λ goes to infinity. We also consider a model of random real algebraic manifolds. In this setting, M is the real locus of a projective manifold defined over the reals. Then, we consider the real vanishing locus Zd of a random real global holomorphic section of E ⊗ Ld, where E is a rank r Hermitian vector bundle, L is an ample Hermitian line bundle and both these bundles are defined over the reals. We compute the asymptotics of the mean volume and Euler characteristic of Zd as d goes to infinity. In this real algebraic setting, we also compute the asymptotic of the variance of the volume of Zd, when 1 r < n. In this case, we prove asympotic equidistribution results for Zd in M

Cette thèse comporte deux parties: - Dans la première partie, nous traitons d'abord une version kahlérienne du célèbre théorème d'extension d'Ohsawa-Takegoshi, puis, un problème de prolongement des courants positifs fermés. Notre motivation provient de la conjecture de Siu sur l'invariance des plurigenres dans le cas d'une famille kahlérienne. En effet, dans la preuve du célèbre théorème d'invariance des plurigenres de Siu, le théorème d'extension d'Ohsawa-Takegoshi joue un rôle important. Il est donc naturel de penser que la preuve de la conjecture fera également intervenir un théorème d'extension de type Ohsawa-Takegoshi dans le cas kahlérien. Suite aux difficultés techniques qui proviennent de la régularisation des fonctions quasi-psh sur les variétés kahlériennes compactes, nous obtenons seulement deux cas particuliers du résultat espéré. Pour ce qui est du prolongement des courants positifs fermés, notre résultat est un cas particulier de la conjecture qui prédit que tout courant positif fermé défini sur le fibré central d'une classe de cohomologie kahlérienne tordue par la classe de Chern du fibré canonique admet un prolongement. - Dans la deuxième partie, nous nous intéressons à l'unicité des solutions des équations de type Monge-Ampère généralisées. Il s'agit d'une généralisation d'un théorème de Bando-Mabuchi concernant les métriques de Kahler-Einstein sur les variétés de Fano. Nous suivons la méthode introduite par Berndtsson et généralisons son résultat en travaillant avec un courant positif fermé à la place d'une paire klt dans son contexte. Les propriétés de convexité des métriques de Bergman jouent un rôle important dans cette partie / This thesis consists in two parts: -In the first part, we first deal with a Kahler version of the famous Ohsawa-Takegoshi extension theorem; then, a problem of extending the closed positive currents. Our motivation comes from the Siu's conjecture on the invariance of plurigenera over a Kahler family. Indeed, in the proof of his famous theorem, the Ohsawa-Takegoshi theorem plays an important role. It is, therefore, natural to think that the proof for the conjecture involves an extension theorem of Ohsawa-Takegoshi type in the Kahler case. Because of the technical difficulties coming from the regularization process of quasi-psh functions over the compact Kahler manifolds, we only obtain two special cases of the hoped result. As for the extension of closed positive currents, our result is a special case of the conjecture which predicts that every closed positive current defined over the central fiber in a Kahler cohomology class twisted by the first Chern class of the canonical bundle admits an extension. -In the second part, we are interested in the uniqueness of the solutions of the equations of generalized Monge-Ampère type, a generalized Bando-Mabuchi theorem concerning the Kahler-Einstein metrics over Fano manifolds. We follow the method introduced by Berndtsson and generalize his result by working with a closed positive current in place of a klt pair in his context. The properties of the convexity of the Bergman metrics play an important role in this part