Une étude de sensibilité aux échelles kilométriques et hectométriques de simulations idéalisées de convection profonde montre qu’une résolution horizontale minimale de 1 km est nécessaire pour commencer à bien représenter les structures convectives et qu'il faut améliorer la turbulence dans les nuages convectifs. Une simulation LES (50 m de résolution) d'un nuage convectif profond permet d’obtenir les flux turbulents de référence, dégradés ensuite à différentes résolutions (2, 1 et 0.5 km), et d'évaluer ainsi la paramétrisation actuelle de la turbulence au sein des nuages convectifs. Les défauts mis en évidence sont une énergie cinétique turbulente insuffisante, liée à une sous-estimation de la production thermique notamment dans des zones à contre-gradient, et des vitesses verticales trop fortes. Une paramétrisation alternative de certains flux turbulents, basée sur des gradients horizontaux, montre une meilleure partition entre mouvements résolus et sous-maille à 1 km de résolution. / The purpose of adaptive observation (AO) strategies is to design optimal observation networks in a prognostic way to provide guidance on how to deploy future observations. The overarching objective is to improve forecast skill. Most techniques focus on adding observations. Some AO techniques account for the dynamical aspects of the atmosphere using the adjoint model and for the data assimilation system (DAS), which is usually either a 3D or 4D-Var (ie. solved by the minimization of a cost function). But these techniques rely on a single (linearisation) trajectory. One issue is to estimate how the uncertainty related to the trajectory aects the eciency of one technique in particular : the KFS. An ensemble-based approach is used to assess the sensitivity to the trajectory within this deterministic approach (ie. with the adjoint model). Experiments in a toy model show that the trajectory uncertainties can lead to signicantly diering deployments of observations when using a deterministic AO method (with adjoint model and VDAS). This is especially true when we lack knowledge on the VDAS component. During this work a new tool for observation targeting called Variance Reduction Field (VRF) has been developed. This technique computes the expected variance reduction of a forecast Score function that quanties forecast quality. The increase of forecast quality that is a reduction of variance of that function is linked to the location of an assimilated test probe. Each model grid point is tested as a potential location. The VRF has been implemented in a Lorenz 96 model using two approaches. The rst one is based on a deterministic simulation. The second approach consists of using an ensemble data assimilation and prediction system. The ensemble approach can be easily implemented when we already have an assimilation ensemble and a forecast ensemble. It does not need the use of the adjoint model. The implementation in real NWP system of the VRF has not been conducted during this work. However a preliminary study has been done to implement the VRF within OOPS (2013 version). After a description of the different components of OOPS, the elements required for the implementation of the VRF are described.