Two theoretical approaches for simulating nonlinear focused ultrasound fields generated by a diagnostic convex array are compared. The first model is based on the three-dimensional Westervelt equation and describes the full structure of the array field with high accuracy. However, it requires great computational resources and is technically difficult. The second model is based on an axially symmetric form of the parabolic KZK equation for estimating the strength of nonlinear effects in the focal region of a beam, which reduces the computational time by a factor of several hundreds. To establish the boundary conditions to the KZK model, the radius and the focal length of a circular piston source are defined such that the simulated field on the beam axis in the linear case fits the real structure of the field in the focal region. It is shown that the parabolic model can be used to accurately describe the spatial and temporal structure of the field generated by a diagnostic transducer in the focal region of the beam along its axis and in the plane of the beam’s electronic focusing.