Abstract: In this research, Jordan Neural Network (J-NN) is applied to emulate an Extended Kalman Filter (EKF) in a data assimilation experiment. The J-NN is a recurrent Artificial Neural Network (ANN), which one there is a reverse-feed from output neurons to input layer of the ANN. The dynamical model studied here is based on the one-dimensional shallow water equation DYNAMO-1D. This code is simple, but the DYNAMO-1D is rich for representing some atmospheric motions, such as Rossby and gravity waves. It has been shown in the literature that the ability of the EKF to track non-linear models depends on the frequency and the non-linear nature from the dynamics. In some cases, only EKF of fourth-order moment works well. ANN is an alternative solution for this computational complexity problem. After training, the ANN has lower computational cost than EKF, considering second and, of course, fourth order KF. The J-NN shows good performance to carry out the assimilation for the DYNAMO model.