With this work, we intend to achieve reduced dynamical models for an incompressible turbulent external flow. The model is constructed from a set of particle image velocimetry (PIV) measurements, using the velocity field data obtained in previous experiments. The Proper Orthogonal Decomposition (POD) is as a technique which allows the extraction of the significant modes of a flow. It permits to identify modes without any a priori hypothesis, and these are ordered by their associated level of energy. The POD modes are used to formulate a dynamical system which only contains the main features of the flow, a low order dynamical system(LODS). This is achieved by means of a Galerkin projection of the Navier-Stokes Equations with the goal to obtain a system of ordinary differential equations describing the main aspects of the flow without demanding excessive computational requirements.