Skip to main navigation Skip to main content
The University of Southampton
Ocean and Earth Science, National Oceanography Centre Southampton

Research project: Iceberg forecasting - from days to decades (ICECAST)

Currently Active:

Icebergs present a hazard to shipping and marine operations, and play an active role in climate change as the great ice sheets of Greenland and Antarctica decline in size. Initially funded by a NERC Innovation Grant, we are working in partnership with the Met Office to provide the capability for forecasting and predicting the drift of icebergs on timescale from days to decades.

Since 2010, we have been developing a model to represent the drift and melting of icebergs in the polar oceans as an interactive option in the NEMO ocean model. The NEMO framework is widely used throughout Europe, and the UK Met Office adopted NEMO around 2010. The Met Office has developed seamless prediction systems that use NEMO for daily forecasts in FOAM (Forecasting Ocean Assimilation Model), for seasonal forecasts in GloSea5 (Global Seasonal Forecasting) and for climate prediction in HadGEM3 (the latest in the Hadley Centre family of climate models).

We initially developed NEMO-ICB, in which icebergs interact with the 3D ocean. We subsequently developed SAS-ICB, in which icebergs respond to surface conditions simulated in NEMO (SAS stands for “Stand-Alone Surface forced”). These two models are used for quite different purposes. NEMO-ICB is designed for the climate and earth system models that are used in multi-decadal climate projections, while SAS-ICB is designed for forecasting iceberg drift on timescales from days to seasons.

Iceberg distributions
Iceberg distributions in each hemisphere, Jan-June (L), July-Dec (R)

Our 30-year global simulations with NEMO-ICB reveal how icebergs across a range of sizes drift away from Antarctica and Greenland. We have examined how and where these icebergs melt, and how this affects ocean salinity, temperature and currents. The Met Office is currently adopting NEMO-ICB in HadGEM3 as they undertake next-generation climate model projections.

SAS-ICB is many times more computationally efficient than NEMO-ICB, making it ideal for multiple short-term forecasts. We have tested SAS-ICB in the Southern Ocean and the northwest Atlantic, using selected observations. In the Southern Ocean, we use satellite observations of giant iceberg tracks over 2014, while in the northwest Atlantic we use digitized charts of generally smaller icebergs that are largely based on daily aircraft surveillance. The SAS-ICB predictions are forced with daily data from FOAM. It takes around 6 hours of computing to simulate iceberg drifts for up to 1 year. We process many individual iceberg trajectories to obtain composite maps of mean iceberg density, size and age, per 1° grid squares.

Validation S Atlantic
Validating iceberg forecasts with giant iceberg tracks - S Atlantlc

In the case of Southern Ocean iceberg drift, simulated icebergs (colours) are simulated to move in broadly the same direction as observed icebergs (symbols), but the simulated drift is too fast, and all the icebergs melt too soon. This is likely due to currently specifying relatively small iceberg sizes, and the use of climatological winds at this testing stage.

Validation NW Atlantic
Validating iceberg forecasts with digitized iceberg charts-NW Atlantic

In the case of northwest Atlantic iceberg drift, the simulation bears more encouraging resemblance to the observations, here for the first half of 2015. After further tests, we hope to use seasonal (4-month) forecasts of surface ocean currents, temperatures and winds, from GloSea5, to develop corresponding iceberg forecasts that may be used in long-range planning of offshore operations, in polar regions of both hemispheres.

This research was initially funded by a NERC Innovation Grant (ref. NE/M007820/1) over December 2014 to August 2015.

PhDs and Other opportunities


Associated research themes

High Resolution Global Modelling

Share Share this on Facebook Share this on Twitter Share this on Weibo
Privacy Settings