Evaluating nonlinearity and chaos in maritime traffic

Abstract

This paper addresses the existence of nonlinear dynamics and chaotic behavior in the Spanish maritime transport services using monthly temporarily data for the period 1992-2013. For this purpose in first instance the Box-Jenkins approach for time series analysis as a preparatory step in order to remove linear dependence. Afterwards nonlinearity and chaotic behavior are evaluated by means of the BDS and the Lyapunov test respectively. The results are especially relevant for modeling and prediction of maritime container traffic and for the design and evaluation of public policies related to the planning and management of port investments. Our findings suggest that there is evidence for the existence of nonlinearity in the underlying dynamics of the number of traffic containers that is not compatible with a chaotic behavior.