Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal OCE2116

TitleTracking and prediction of the path of the Giant Pine Island iceberg
Investigator Bigg, Grant - University of Sheffield, Department of Geography
Team Member
Dr. Marsh, Robert - University of Southampton, School of Ocean and Earth Sciences
SummaryThe NERC-funded project has 4 main objectives: 1. To monitor the passage of this giant iceberg, derived from Pine Island Glacier, over ~6 months, including the products of any major fracturing events; 2. Tomodel its trajectory, and melting, within the coupled ocean-iceberg model NEMO-ICB; 3. To forecast its future movement and melting; 4. To forecast the giant icebergís likely dynamical impact on the Southern Ocean. Two principal methods will be used to achieve these aims. One is the use of iceberg tracking, using SAR imagery. The other involves use of the coupled ocean-iceberg model NEMO-ICB, which is a newly developed part of the UK Meteorological Office's ocean component of the climate model, to do the forecast elements. The Terrasar-X data will be used in the work towards the first objective - suffice it here to say that the iceberg module within NEMO will be used to both model the existing and progressing trajectory and melting of the PIG iceberg (and the products of any break-ups), and to predict its path and melting until dissolution. With respect to the tracking, the iceberg will be tracked using the ITSARI (IceTracking using SAR Instruments) software (Silva and Bigg, 2005; Hall et al.,2012), which can differentiate between icebergs and sea-ice, a necessary tracking capability during the months when the giant iceberg is in the Antarctic Coastal Current zone. If the iceberg were to fracture during the monitoring period all major fragments would be tracked separately. Both the track and iceberg(s) area would be monitored to inform the updating of the iceberg modelling. The latter gives an estimate of the iceberg melting rate to compare with the coupled ocean-iceberg modelís prediction. It is planned to track the iceberg for about 9 months from its calving, on July 8, 2013. So far, it has only moved a very small amount, presumably due to trapping by winter seaice in this part of the Bellingshausen Sea. We therefore will not need regular coverage until the summer. We hope to use other imagery, but from the Terrasar-X mission we anticipate requiring a Scansar HH polarization image from each of the times of late July, late October, late November, and then bimonthly through December to February or March. The core period of the iceberg modelling will run from January to March 2014. Our deliverables within the context of this application are therefore to i) produce initial track and size data by early January to be used to initialise the model; ii) to continue monitoring iceberg movement, fracture and area during the modelling period. By that stage, the modelling component, having been verified from initial movement over the first 6 months, will give predictions of the Pine Island iceberg movement through the rest of 2014. The funding obtained is for a 6 months project, between Sheffield and Southampton, with Dr. Wilton employed to carry out the tracking at Sheffield and another postdoc to do the modelling in Southampton, under Dr. Marsh's direction. Hall J.A. et al., 2012, Identification and tracking of individual sea ice floes from ENVISAT Wide Swath SAR images: a case study from Fram Strait, Rem. Sens. Lett., 3, 295-304 Silva,T.A.M. and Bigg, G.R., 2005, Computer-based identification and tracking of Antarctic icebergs in SAR images, Rem.Sens. Environ., 94, 287-297.
Final ReportThe impact of varying satellite instruments and modes on parameter estimation of giant Antarctic iceberg B31 A giant iceberg, B31, calved from Pine Island Glacier in West Antarctica on 11th November 2013. This was tracked over the following 18 months, using visible MODIS imagery and two SAR systems, from the TerraSAR-X and Radarsat2 satellites, the former using two contrasting modes. On occasion images were available from more than one system within a few days. Estimates of size-related parameters - area, major axis, minor axis and perimeter - of the iceberg were calculated using the ITSARI image analysis system. It was found that, in general, variation between estimates of axis and area between SAR modes and instruments was small and statistically insignificant, although the scatter between instruments was larger than that between modes of one satellite. Estimation of these parameters using MODIS imagery, however, was far more erratic. Estimates of B31ís perimeter were the most inconsistent between instruments, as resolution is integral to estimation of this parameter. It was noteworthy that for this size estimate the coarser MODIS and Radarsat2 results were similar. This suggests an effective maximum in segment complexity for iceberg edges.

Back to list of proposals

© DLR 2004-2016