Iceberg drift modelling and validation of applied metocean hindcast data

An iceberg drift model covering the Barents and Kara Seas has been developed. The skills of the model relies both on the ability to describe physical actions from the environment on the icebergs and the accuracy of the applied metocean variables (wind, waves and currents). Experiences from the East Coast of Canada show that iceberg modelling may work reasonably well and indicate that iceberg drift models are able to fulfil both of the above mentioned requirements. By applying similar models in other regions, it may be assumed that wind, waves and currents affect the iceberg in a similar way as at the East Coast of Canada. However, the reliability of available metocean data sources will vary significantly from region to region. Due to this, a study with the objective to evaluate the quality of the underlying metocean models has been performed. A significant amount of recorded wind, wave and current data from various regions in the Barents Sea have been applied in comparisons with hindcast data from selected atmospheric and oceanographic models. Results show that the quality of wind and wave data applied by the iceberg drift model is very good. Regarding current velocity, there is a poor match between data from the applied oceanographic model and measurements. A method for improving the current magnitude has been introduced. The relative importance of winds, waves and currents on iceberg drift has also been investigated. In general, currents are most important for iceberg drift. However, in open waters, the wave drift may become the most important forcing. The presented iceberg drift model is considered to provide good results in situations with strong winds (and waves) and low currents while situations with low winds will give less reliable results. It is concluded that the quality of incorporated metocean data in any iceberg drift model need to be documented in order to fully understand possible limitations in iceberg drift simulations. Further work should focus on improvements in oceanographic modelling in order to establish a more reliable oceanographic hindcast for the Barents Sea. (C) 2009 Elsevier B.V. All rights reserved.