The Ice Service has access to the data from a number of short-range, 10 days or less, sea ice forecasts and has prepared visualizations of these to highlight the key parameters of sea ice concentration, thickness, and drift. We present here the static graphics out to 6 days for the present day, and 24, 48, 72, 96, 120, 144, and 168 hours. Depending on the model production time there is an additional offset of 14 to 27 hours to these time points. These forecast models have been selected as their spatial resolutions are better than 5 km, and therefore approaching a scale to which they can show the finer details of the sea ice edge and conditions within Svalbard fjords.
As these forecasts are made externally to the Ice Service, we are unable to guarantee their availability or accuracy. These should therefore be considered demonstration or experimental, and used at own risk. Please contact us if you require any help interpreting these. Below the graphics, you will find a section providing a brief description, and links to further information, on the forecast models. The white paper 'Knowledge needs in sea ice forecasting for navigation in Svalbard and the High Arctic', gives a good overview of the current and future work that needs to be carried out to facilitate sea ice forecasting into the future.
Due to the forecasts being supplied from different sources, and having varying processing times, this page will receive updates at around the following UTC times throughout the day: GOFS3.1 at 04:15, neXtSIM at 11:00, Barents-2.5km at 11:40, and the distance and direction plot at 16:30. The times of the latest updates are shown below:
|Day 1 (24 hours)|
|Day 2 (48 hours)|
|Day 3 (72 hours)||Not Applicable|
|Day 4 (96 hours)||Not Applicable|
|Day 5 (120 hours)||Not Applicable|
|Day 6 (144 hours)||Not Applicable|
To help understand the dynamics of the sea ice, we have also prepared an animation in MPEG-4 (.avi) format. This provides a visualization of every time step of each model.
The Forecast Models
The forecast models are initialized using different datasets to those used to draw the ice charts, and are then forced using different weather and ocean forecasts. A description of the different forecast models is provided below. This provides information on the basic set up, with links to further detailed, technical, information.
The Global Ocean Forecasting System (GOFS) 3.1 forecast is produced by the Naval Research Laboratory: Ocean Dynamics and Prediction Branch and provides a 180-hour forecast on a global grid that is 0.08° longitude by 0.04° latitude. This equates to a 2 by 4 km grid at 80° N. The ocean forecast is from the HYbrid Coordinate Ocean Model (HYCOM), with 41 vertical layers, coupled to the Los Alamos National Laboratory Community Ice CodE (CICE), and with data assimilation being handled by the Navy Coupled Ocean Data Assimilation (NCODA) system. Atmospheric, weather, forcing comes from the NAVy Global Environmental Model (NAVGEM).
The initial sea ice concentration is a combination of Advanced Microwave Scanning Radiometer 2 (AMSR2) passive microwave and the Multisensor Analyzed Sea Ice Extent (MASIE). MASIE incorporates the Interactive Multisensor Snow and Ice Mapping System (IMS) that runs at the U.S. National Ice Center (USNIC). More Information on the sea ice data assimilation and prediction capabilities can be found in the presentation by Allard and others, 2014, and papers by Metzger and others, 2014, and Posey and others, 2015.
The neXtSIM (ARCTIC_ANALYSISFORECAST_PHY_ICE_002_011) forecast is produced by the Copernicus Marine Service (CMEMS) and provides a 168-hour forecast over a 3 by 3 km grid covering the Arctic, but excluding the Canadian Archipelago, Baffin Bay and Hudson Bay. neXtSIM is forced with surface atmosphere forecasting from the European Centre for Medium-Range Weather Forecasts (ECMWF) and ocean forecasts from the CMEMS TOPAZ4 (ARCTIC_ANALYSIS_FORECAST_PHYS_002_001_a) forecast.
The forecast assimilates the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT) OSI SAF sea ice concentrations, derived from both Special Sensor Microwave - Imager/Sounder (SSMIS) (OSI-401) and AMSR2 (OSI-408) passive microwave.
The sea ice model uses a Brittle-Bingham-Maxwell (BBM) sea ice rheology on an adaptive Lagrangian triangular mesh of 10 km average cell length. This is then interpolated to the 3 km resolution regular grid. Outputs provided at hourly frequency are sea ice concentrations, thickness, drift velocity and snow depths. More information on neXtSIM and its capabilities can be found in the papers by Rampal and others, 2019, and Williams and others, 2021.
The Barents-2.5km forecast is produced by the Ocean and Ice section of the Research and Development department of MET Norway and provides a 66-hour forecast for the Barents, Norwegian and Greenland Seas. The ocean model is the Regional Ocean Modeling System (ROMS), coupled to the CICE sea ice model using OASIS-MCT, and with 4DVAR data assimilation. Atmospheric, weather forcing is provided by the AROME-Arctic convective scale forecasting system at MET Norway.
The forecast assimilates AMSR2 sea ice concentration information from the University of Bremen.
A key use of forecast models is to provide a warning of potential hazard, e.g. for a weather forecast things like approaching bad weather such as gales or thunderstorms. With sea ice a typical use is to assess whether the ice edge is advancing or retreating from a fixed position, such as a vessel undertaking a survey that requires maintaining a constant location.
We present here a comparison of the three forecast models for this type of scenario, using a location east of Jan Mayen. The model graphics coverage and location point are shown in the map below.
The forecast of distance and direction is shown below, with the blue line representing GOFS3.1, the red line neXtSIM, and the green line Barents-2.5km. On days when the ice chart is produced, the distance and direction to this is shown by a filled green star for very open drift ice (10%), and an outline yellow star for open drift ice (40%).