Stochastic weather generators (SWG) are statistical models designed to provide arbitrarily long realistic simulations of atmospheric variables such as temperature , precipitation and wind at specific locations or areas, at daily or hourly temporal scales. Such simulated series serve as input for environmental, ecological or agricultural impact studies or for long time planning of renewable energy resources. In this talk I'll introduce the main ideas behind this models. Afterwards I will present a spatio-temporal generator based on a latent Gaussian field and its application to simulating daily precipitation in the south of Norway.