The problem of the so-called dark vessels is a major security issue related to the marine traffic and environment. In this respect, the automatic ship detection from satellite imagery is one of the possible countermeasures in the fight against illegal activities. Technically, such a way of detecting dark vessels mainly relies on the identification of the far field wake released by a hull which might be visible over a significant distance from the hull itself and which might last in the water surface for a reasonable amount of time to be seen by a satellite. However, the shape and the persistence of such a wake field, which includes both the steady wave pattern and the turbulent wake behind the hull, is strongly affected by other phenomena mainly related to the weather conditions, then including the presence of surface wind waves or other patterns on the sea surface. Moreover, there might be unfavorable conditions for the satellite to produce accurate images to be used in this recognition process. This complex situation is faced in a project funded by the Italian Ministry of University and Research called UEIKAP (Unveil and Explore the In-depth Knowledge of earth observation data for maritime Applications) in which an Artificial Intelligence (AI) system for the automatic identification of dark vessels from optical and SAR (Synthetic Aperture Radar) images is under development. The training of the AI framework is based on both publicly available image datasets and ad-hoc synthetic data. The latter are generated by using Computational Fluid Dynamics simulations on a set of hulls in different operating conditions. The proposed analysis focused on the numerical issues related to the prediction of the far-field wake of a hull in the light of the final purpose of the UEIKAP project. A comparison of two different CFD approaches is shown and discussed in terms of pro and cons of both methods and the obtained results.