Diagnostic models for estimating mean sea level change using hybrid model of exponential smoothing and neural network

A diagnostics model was proposed to estimate the mean sea level change by hybridizing exponential smoothing and neural network. The model integrated the linear characteristics of the exponential smoothing model and the nonlinear pattern of the neural network. Mean sea level data were obtained from the measurements of Jason-2 satellite altimeter mission from 2008 – 2014. The results showed that the diagnostic model obtained by hybridization of the exponential smoothing and neural network model provide an alternative prediction model for the mean sea level change in South China Sea.