This chapter explores the use of Gaussian processes (GPs) for modeling physicochemical variables in aquatic environments and the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) for planning sampling trajectories. Gaussian processes allow obtaining accurate estimates and quantifying uncertainty, crucial for real-time decision-making. The definition of uncertainty in Gaussian processes is not data-dependent, which facilitates the implementation of a Receding Horizon algorithm to plan trajectories based on predicted uncertainty. The CMA-ES algorithm optimizes the sampling trajectory, maximizing the information collected. Compared to other algorithms, the CMA-ES proves to be more effective in reducing uncertainty and adapting to the geometry of the search space.