Statistical analysis of ice load occurrence in floe ice fields

Ships navigating in the Arctic and Antarctic regions frequently encounter the threat of ice. Floe ice exhibits various shapes and sizes, with significant variations in thickness and concentration across different ice fields. Estimating the number of ice loads accurately is crucial for strategic navigation and fatigue damage assessment. This paper focuses on the ice load occurrence in floe ice fields which can be determined by the time intervals between contacts. Previous study has shown that the occurrence in level ice fields follows a Poisson distribution and the time intervals follow an exponential distribution. Based on the measurement data on Polar oly and Research Vessel (PSRV) S.A. Agulhas II during 2018/19 Antarctic voyage which encountered extensive floe ice, this paper examines the applicability of previously proposed patterns in level ice fields to floe ice fields. A mixture of exponential distributions is proposed to fit the time intervals in floe ice fields and the statistical test is adopted to evaluate the goodness of fit. The relationships between the parameters of ice conditions and distribution models are discussed. Monte Carlo method and theoretical method are adopted to obtain the ice load occurrence based on the distribution of time intervals. To assess the uncertainty associated with the estimated parameters and the ice load occurrence, Bayesian approach and Bootstrapping are adopted. To further investigate the influence of ice load occurrence on short-term extreme distributions of local ice loads, three methods are employed and the distributions are found to be highly consistent.