The possibility of increasing flight duration, particularly for small and micro fixed-wing Unmanned Aerial Systems (UAS), is typically hindered by the payload and size of the platform. In maritime missions, Atmospheric Energy Harvesting (AEH) near the Marine Atmospheric Boundary Layer (MABL), is a low-cost solution, with very promising results, that may allow drastic increases in flight duration by exploiting wind energy. The utilization of atmospheric energy is not a straightforward process. Understanding the mechanics behind the dynamic soaring process, which is the most promising form of AEH in marine applications, is key. A proper identification and characterization of the wind field is required. In addition, a smart trajectory generation and tracking module shall maximize the energy gain or to increase the flying time. The atmospheric energy represents a series of complex processes, specially when it comes to the interaction between the ocean and the atmosphere in the MABL. However, the MABL has drastic gradient changes in which soaring maneuvers can be performed. This paper focuses in the study of the soaring energy transfer mechanisms, based on the the interaction between the MABL and the surface of the ocean and on observations of birds that perform swooping maneuvers. This allows the generation of energy efficient trajectories that will provide a UAS with extended time duration in an oceanic mission. The trajectory generation method considers interpolating splines that contain wind information embedded into the control points, i.e. tension, velocity and direction unitary vector, aiming to replicate the biologically-inspired Rayleigh cycle. Preliminary results in simulations are presented, showing that there are potential energy gains after executing splines that follow basic Rayleigh cycles.