A generalized ice drift spectrum based on measurements in the Beaufort Sea

The objective of the present paper is to propose a mathematical model for the frequency distribution of ice drift velocity. Measured data of the ice drift speed are employed as a basis for this model. The time series of ice drift were collected in the Beaufort Sea during the time interval 2006-2017. For this purpose, acoustic Doppler current profilers (ADCP) were applied. Wind is the main influencing factor for ice drifting in the oceans. The Nansen number, which typically describes the relationship between the wind speed and the ice drift speed as a percentage, is applied to quantify the degree of association. The frequency distribution of the ice drift velocity can accordingly be expected to be closely related to the downscaled wind turbulence spectrum. In the present study, the Fourier transforms of the time series of wind speed and ice drift speed are applied to investigate the properties of the frequency spectrum. The wind velocity has a strong influence on ice drift behavior, especially during the winter season. The Nansen number is found to be approximately 2.5 % within the study area. Accordingly, the Kaimal and Harris standard turbulence spectral density models can be adopted to represent the ice drift spectral density for the winter season. However, for the summer season, the ice drift spectral density exhibits a different decay in the high-frequency range. The proposed generalized mathematical model of the ice drift spectrum contains more parameters than the Kaimal and Harris turbulence spectral density models and is hence more flexible in fitting the ice drift behavior during both the winter and summer seasons. The proposed ice drift spectrum is accordingly found to provide an improved fitting of the empirical spectra estimated from the ADCP measurements compared to the Kaimal and Harris turbulence spectral density models.