Parameter identification for a nonlinear non-smooth thermodynamic system of sea ice

This study is intended to provide a method to estimate the sea ice parameters in a nonlinear non-smooth thermodynamic system (NNTS) from the sea ice temperature measurements. A new optimization algorithm is constructed to estimate the coefficients describing the salinity and other two parameters in NNTS. Another simulation for sea ice temperature during different measurement period is operated. Results show that a better simulation of the temperature distribution is possible with the estimated parameters than with the experimental parameters.