%0 Conference Proceedings
%T Neural network as a new approach for data assimilation
%8 16-20 out. 2000
%A Nowosad, A. G.,
%A Campos Velho, H. F. de,
%A Rios Neto, A.,
%@affiliation INPE-Sao Jose dos Campos-12227-010-SP-Brasil
%B Congresso Brasileiro de Meteorologia, 11.
%C Rio de Janeiro
%K data assimilation, neural networks, Kalman filter, nonlinear dynamics.
%X Multilayer Perceptron Neural Networks are tested as a new method for data assimilation in DYNAMO meteorological model. The approach "emulates" the Kalman Filter data assimilation method avoiding recalculation of the gain matrix at each instant of assimilation. A new prodedure for training the networks is also presented, based on a modification in the backpropagation algorithm. An Adaptive Extended Kalman Filter was used to provide examples for network training.