Cosmology with weak-lensing peak counts

Abstract: Weak gravitational lensing (WL) causes distortions of galaxy images and probes massive structures on large scales, allowing us to understand the late-time evolution of the Universe. One way to extract the cosmological information from WL is to use peak statistics. Peaks are tracers of massive halos and therefore probe the mass function. They retain non-Gaussian information and have already been shown as a promising tool to constrain cosmology. In this work, we develop a new model to predict WL peak counts. The model generates fast simulations based on halo sampling and selects peaks from the derived lensing maps. This approach has three main advantages. First, the model is very fast: only several seconds are required to perform a realization. Second, including realistic conditions is straightforward. Third, the model provides the full distribution information because of its stochasticity. We show that our model agrees well with N-body simulations. Then, we study the impacts of the cosmology-dependent covariance on constraints and explore different parameter inference methods. A special focus is put on approximate Bayesian computation (ABC), an accept-reject sampler without the need to estimate the likelihood. We show that ABC is able to yield robust constraints with much reduced time costs. Several filtering techniques are studied to improve the extraction of multiscale information. Finally, the new model is applied to the CFHTLenS, KiDS DR1/2, and DES SV data sets. Our preliminary results agree with the Planck constraints assuming the Lambda-CDM model. Overall, this thesis forges an innovative tool for future WL surveys. The manuscript provides a brief review on WL peak counts.