Data portal expands its services for sea-ice drift

15 March 2018

Sea ice represents only a comparatively thin skin on the ocean, where it is cold enough to freeze. As such, it can easily be moved about by wind or currents. The majority of sea ice is young, relatively thin drift ice, and the remainder is dense, massive pack ice. In addition to data from buoys anchored in the ice, which provide information on its movement by transmitting their position on a daily basis, sea-ice drift can be mapped with the help of satellite data. This comprehensive information on ice movement provides valuable clues to understanding the dynamic processes at work in sea ice and how they shape the Arctic climate system. In addition, the insights into sea-ice drift gleaned from satellite data are used to validate computer models that simulate sea ice, and to refine e.g. sea-ice forecasts.


That being said, sea-ice movement estimates based on satellite data also include uncertainties (technically referred to as uncertainties and systematic errors), which can be due e.g. to measurement errors, or errors in the analytical methods used to ‘translate’ the raw data from the satellites into movement patterns. As such, these uncertainties and systematic errors have to be kept in mind when using drift data. The sea-ice drift data available at and its uncertainties have now been substantially expanded in terms of both the timeframe covered and the base data used. Portal users can now choose to download the monthly sea-ice drift for each product and the corresponding uncertainties in map form, or to download the actual data in NetCDF format. Calculations of the uncertainties for various sea-ice drift products, including those released by major data centres, were developed by the Helmholtz research initiative ‘Regional Climate Change’ (REKLIM). ‘The sea-ice drift data offers researchers the chance to more accurately investigate the loss of sea ice in the Arctic by allowing them to view not only the ice concentration and thickness, but also horizontal ice transport,’ explains Dr Hiroshi Sumata, a sea-ice physicist at the AWI and head of the project for calculating sea-ice drift uncertainties. ‘We’re very happy to be able to promote this type of work in the Helmholtz research initiative REKLIM,’ adds Dr Klaus Grosfeld, Managing Director of REKLIM.


In the context of analysing the satellite-based datasets, two pattern recognition methods – the Eulerian and Lagrangian methods – are used to track the movement of individual ice structures (packets). For the Eulerian products, the initial position of the sea-ice ‘parcels’ to be tracked lies on a fixed coordinate system; for the Lagrangian products, the initial position is not fixed, and changes from day to day, which makes it easier to identify suitable structures. In both types of product, the movement of the individual sea-ice ‘parcels’ in a specific time interval is portrayed as a vector (ice displacement / movement). In this regard, the advantages of the Eulerian products are their homogenous spatial and temporal data coverage and their ease of use, thanks to the fixed coordinate system. All of the calculations offered at are based on Eulerian products.


With the new datasets, scientists now have access to monthly sea-ice drift vectors and estimates of the uncertainties and systematic errors (bias) for the components of ice drift in the Arctic Ocean.


‘Over the past few years, numerous products for sea-ice drift have been made available. Yet in some cases they differ considerably, which is largely due to the use of different satellites or different algorithms. The advantage of our work is that we calculate the uncertainties and show them. Our data can be directly used for sea-ice analyses, model studies and data assimilation,” adds sea-ice physicist Hiroshi Sumata.


The four available sea-ice drift products hail from the following data centres:

1. Ocean and Sea Ice Satellite Application Facility (OSI SAF) (OSI-405)

2. Ifremer (CERSAT)

3. National Snow and Ice Data Center (Polar Pathfinder Daily 25 km EASE-Grid Sea Ice Motion Vectors, Version 2) and

4. National Institute of Polar Research (from Noriaki KIMURA), Japan.

The new datasets for November to April in the years 2010 to 2015 can now be downloaded as either graphics or raw data from the portal (here).



Dr. Hiroshi Sumata (AWI)

You have any further questions?