Sea ice information for navigation and offshore operations

1. Frontier regions for exploration
2. Current sources of information
3. Ice monitoring with satellite SAR
4. The Northern Sea Route - ICEWATCH
5. The Sea Ice Information Service (EOS Ltd.)
6. The Ice Prediction and Analysis Platform (GEC Marconi Research Centre)
7. Current developments
8. Contacts

Frontier regions for exploration

Ice information is required by a wide spectrum of users operating at high latitudes. These include fishing activities in areas such as the Barents Sea and the region around Svalbard, and merchant vessels on-route through ice-infested regions in the Baltic or the Canadian, Alaskan and European Arctic. Greater exploitation of the Arctic for its offshore oil and gas reserves has lead to a requirement not only for accurate and timely monitoring but also reliable design statistics for offshore construction. Additional users, specifically in coastal regions, include coastguard and harbour authorities.

[Figure 1. An oil rig in the Cook Inlet, US. Courtesy: US Minerals Management Service]

Icebreakers fulfill an essential role in ice-infested waters, keeping ports in the Baltic Sea open in the winter months for example, and chaperoning vessels not adapted to ice conditions. These services are often provided free or at a fixed price, and so maximisation of efficiency and optimisation of the service brings direct benefits to the operator as well as the customer.

Requirements for each activity overlap considerably but differ in detail. The extent of the ice provides safety limits for those concerned with avoiding ice altogether. Ice-strengthened vessels which can risk navigation within the icepack require information on the distribution of ice age and on pack motion, which is used to avoid high pressure ridges, and identify navigable leads. The end product required is a service which incorporates all the available data combined with expert analysis.

[Figure 2. The ice-strengthened floating drilling unit BeauDril Kulluk with icebreaker. Courtesy: US Minerals Management Service]

Many countries maintain ice monitoring and forecasting services including China, Russia, the USA, the Scandinavian countries, Canada and the Baltic States. These services are essentially tactical in nature, providing information on conditions for planning within current operations and for optimal routing of icebreakers. The forecasts consist of daily regional (500km2 coverage) ice surveys in the form of ice maps which include information on ice type and concentration.

However, specific requirements such as strategic information prior to operations, multi-temporal information on ice pack drift, and spatially detailed nowcasts and forecasts for critical areas such as straits are not fulfilled by these services.

Current sources of information

Currently, useful albeit sparse information is provided in the form of reports from vessels in the ice, and from icebreakers and weather stations. Other sources include airborne survey by radar, but these aircraft are expensive and provide only limited coverage. Although the radar sensor itself may be insensitive to weather conditions, the aircraft upon which it is flown can be grounded for a week or more at a time by adverse weather conditions.

Satellite image data have proved to be of major benefit for regional ice surveillance - large scale monitoring of sea ice is routinely performed using Special Sensor Microwave Imager (SSM/I) data which have a coarse spatial resolution of around 50km, but these data are independent of cloud cover and observe the whole Arctic twice daily. The infrared/optical sensor NOAA AVHRR has a higher ground resolution (1 km) but is seriously limited in operability in cloud cover which is common at the ice edge (as illustrated in Figure 3) and by darkness, when only thermal infrared imagery can be used.

[Figure 3. Ice chart derived from visible band AVHRR imagery, illustrating the impact of cloud cover (white areas). Courtesy: Danish Meteorological Institute.]

ERS Synthetic Aperture Radar (SAR) imagery, available up to 84 °N, have a spatial resolution of approximately 25m (though the lower resolution 100m Fast Delivery products are usually used for ice surveillance) and a coverage of 100km by 100km, thus providing a vital bridge between coarse resolution satellite data and airborne radar at a resolution which allows leads and other features important in successful navigation to be resolved. The SAR has an all-weather capability and is not affected by the long dark polar winter. To obtain daily ice maps of regions which are typically 500km by 500 km it is thus useful to combine detailed high resolution information from the SAR (delivered 2-3 hours after acquisition) with other data sets SSM/I and AVHRR (when available).

Ice monitoring with satellite SAR

The SAR is an active microwave instrument, which measures the degree of backscatter from radar pulses which are reflected from the Earth’s surface back to the satellite. In SAR images, with a resolution of 100m, one can distinguish different ice types and map leads, polynyas, shear zones, landfast ice, drifting ice and the location of the ice edge.

The radar backscattering properties of ice vary with age, as illustrated in figure 4, because the response from the ice is related primarily to properties such as its surface scattering behaviour (roughness and presence of melt ponds or frost flowers), and its volume scattering properties (salinity, different snow/ice types).

[Figure 4. How SAR backscatter measured from ice varies with ice age. Courtesy: Nansen Environmental and Remote Sensing Center]

Using the combination of the backscatter from ice, and the context of surrounding ice, its texture and the shape of the floes, the state and condition of sea ice can be accurately characterised by SAR.

Figure 5 illustrates the ability of SAR to provide high resolution information. The SAR stripe cuts across the Odden ice tongue apparent in the coincident SSM/I ice chart. The tongue consists of small first-year and multi-year floes and locally formed grease and pancake ice. These ice types were classified from the SAR imagery and documented with in situ ship observations (see photographs) from around area B. Dark signatures (above both B and C) are due to grease ice, while brighter signatures (below and around C, around B) are pancake ice.

[Figure 5. An SSM/I ice concentration chart (left) with coincident ERS SAR strip (second left) and ship observations of pancake ice (third left) and grease ice (right). Courtesy: Nansen Environmental and Remote Sensing Center]

The Northern Sea Route - ICEWATCH

[Figure 6. The Northern Sea Route. Courtesy: Rand McNally & Company]

The Northern Sea Route is the sailing route along the coast north of Russia from the Barents Sea in the West to the Bering Strait in the East (see map). This route can reduce the transit time between Europe and the Pacific by approximately 10 days, though ice conditions restrict sea transportation which requires ice class vessels as well as icebreaker assistance throughout the year. In summer there is traffic in the whole sailing route, whereas in winter it is mainly the western part which is used serving the ports on the Yenisei River. Additionally, offshore oil exploration and production activities in areas such as the Eastern Barents and Kara Seas require information for structural design, and for monitoring purposes.

An extensive ice monitoring and forecasting service has been built up over the past 50 years in Russia for the Northern Sea Route in order to serve the icebreaker service, in which satellite SAR has not been used. ICEWATCH, the first joint project in earth observation between Russian Space Agency (RSA) and European Space Agency (ESA), aims to implement satellite monitoring of the Northern Sea Route by combined use of ESA ERS SAR, RKA Okean SLR and other remote sensing data to support activities in the area.

The Nansen Environmental and Remote Sensing Center (NERSC) first demonstrated use of ERS-1 SAR data for near real-time ice mapping in the Northern Sea Route in August 1991, only a few weeks after the launch of the ERS-1 satellite. SAR-derived sea ice maps were then sent by telefax to the French polar vessel L´Astrolabe during her voyage through the Northeast Passage from Norway to Japan. Figure 6 shows an ice map derived from an ERS SAR image in comparison with a coincident ice chart from the Russian Ice Service. Note the much higher detail provided by the SAR and the lack of information due to cloud cover in the coincident optical AVHRR image. Other demonstrations have been carried out by the Nansen Centers in Bergen and St. Petersburg on several of the Russian icebreakers operation in ice convoy services along the Siberian coast and rivers, where a scientist from the Nansen Center in St. Petersburg stayed onboard the icebreakers and analyzed the SAR images in co-operation with the captain and ice pilots. With the Nansen Center in St Petersburg serving the Northern Sea Route area, the Nansen Center in Bergen serves the Greenland Sea, Barents Sea and Svalbard.

[Figure 7. Ice map dervied from a SAR image, with coincident ice chart derived from SSM/I and AVHRR. Note areas neighbouring SAR image position affected by cloud in the chart. Courtesy: Nansen Environmental and Remote Sensing Center]

Methods for the near real-time distribution of the high resolution, 100m, ERS SAR images and maps from the Tromsoe Satellite Station to icebreakers operating in the Northern Sea Route have been demonstrated using of the INMARSAT-A communication satellite. For example, on January 25-26, 1996 the icebreaker Taymir was sailing from Dikson to Beliy Island (70°-80°E) in 100% ice. With a PC and modem connected to the INMARSAT station onboard, ERS-1 SAR images were received 5 hours after the satellite overpass. In the image areas of rough ice and hummocks (brighter signature) could be clearly distinguished from smooth undeformed ice (darker signatures). Based on this information the icebreaker changed its course and selected a much quicker and safer sailing route.

Sea Ice Information Service (EOS Ltd.)

Earth Observation Sciences (UK) are currently seeking to resolve salient issues in the field of ice charting. Firstly, to fuse effectively heterogenous datasets to derive unambiguous and reliable ice information, and secondly to determine optimal mechanisms for its dissemination, tailored to the customer’s needs.

A Sea Ice Workstation, supported by the BNSC and DRA, was developed to prove the concept of usefully combining different types of satellite data, with most of the processing targeted at ice classification into open water, first-year and multi-year ice. Contextual information was provided through data from the SSM/I and AVHRR sensors while detailed analysis of local areas was achieved through the SAR on ERS.

The SIWS was tested between 1994 and 1996 in the supply of support products to operators engaged in activities in ice-infested waters. Charts were generated displaying ice motion vectors, ice types, ice concentrations and ice edge location, depending on the user’s requirements, which were supplied to a number of users including an oil production platform operating in the Barents Sea; meteorological organisations for input to forecasting models; shipping operators for route planning, and a Royal Navy submarine. In this case, coded ice charts were received on a regular basis when the vessel was submerged under ice.

The SAR data processing algorithm in a new operational PC-based system uses SSM/I ice classifications to define initial assumptions about the surface in an attempt to resolve discrepancies and contradictions in the results from these sources, see figure 7. The resultant charts are being made available digitally to vessels with Inmarsat or Orbcomm receivers. Digital manipulation of the data on-board offers the chance to use the resolution of the sensors to enhance knowledge of specific areas, and to link it with other bridge-based information systems. This approach is the basis of a new collaboration between EOS, The Weather Service of Northern Norway (VNN) and others to model the movement of vessels through sea ice, as a meant of providing a decision support tool for the ice pilots.

[Figure 8. An SSM/I and AVHRR derived ice chart with coincident classified SAR imagery. Courtesy: Earth Observation Sciences]

The Ice Prediction and Analysis Platform (GEC-Marconi Research Centre)

In August 1994, GEC-Marconi Research Centre (UK) to provided support to NunaOil, who were conducting a seismic survey in the ice-infested Greenland Sea. Maps of local ice concentration and movement derived from ERS SAR images were delivered in near real time by fax to the MV Thetis, the vessel carrying out the survey. In addition to NunaOil, the Canarctic Shipping Company are also evaluating the products with a view to their full operational use.

ERS-1 SAR images were received at the West Freugh ground station in Scotland and transmitted via conventional land lines to the GEC-Marconi Research Centre. The images were processed automatically on a dedicated system, the Ice Prediction and Analysis Platform (IPAP) to produce ice-concentration and ice-movement maps. Results from the classifications are stored to be used to generate multi-temporal ice edge and pack ice motion products.

A validation study of the supervised classification algorithms available within IPAP yielded an accuracy of over 98% in the area around Bent Horn oil terminal using coincident in-situ information from the Canadian icebreaker ‘MV Arctic’.

[Figure 9. Ice edge product, showing temporal variation between 5th and 8th March 1992. Courtesy: GEC Marconi Research Centre]

Two approaches to the problem of tracking ice motion have been developed: a method for tracking pack movement, and a classification-based algorithm which identifies regions matching from one SAR image to the next, useful for monitoring individual ice floes in the Marginal Ice Zone, for example.

Figure 8 shows an ice edge product. The area to the west of the image consists of first-year ice with a fringing of grease ice. The image is overlaid by two edge lines, one generated from the displayed image, and one from the next overpass of the satellite three days later.

[Figure 10. Ice concentration chart. Courtesy: GEC Marconi Research Centre]

Ice concentration and motion maps derived from this information such as figure 9, were then faxed to the master of the ship via the Inmarsat communications satellite within 3 hours of the satellite overpass, which in a number of cases allowed sufficient time to make operational decisions regarding the survey route and the deployment of the hydrophone arrays.

Current developments

Benefits from accurate and timely ice information are immediate and tangible. The ice workstations generate results which provide additional information to standard ice charting services, particularly useful for certain operations undertaken for extended time periods in ice-infested waters which need information on pack movement and leads, avoiding the need for icebreakers to act as costly chaperones. Where icebreakers are needed, routes through leads or the thinnest ice may be identified reducing fuel consumption and increasing the rate of progress to the ultimate destination. During the ice season in the Baltic Sea where ice breakers are used extensively to keep harbours open, approximately 4000 ships visit the 23 Finnish harbours alone, each staying for an average four days in ice-infested waters. The cost in delays, etc. to this traffic amounts to $200 million per year, in addition to costs to harbour authorities and related industry.

Continuity of data is essential in establishing operational services. The RADARSAT satellite from the Canadian Space Agency, launched in 1995, carries a wide swath SAR instrument which has been tuned to measure ice, and the data are now being tested for incorporation into services as an additional data source. The ERS-2 SAR will be succeeded in 1999 by ASAR on ENVISAT, which will also have a wide swath mode useful for regional ice monitoring.

Contacts

For further information on the relevant service, contact:

Prof. Ola Johannessen
Nansen Environmental and Remote Sensing Center
Edvard Griegsvi 3a
N-5037 Solheimsviken Bergen (Norway)
Tel.: +47 55 29 72 88
Fax.: + 47 55 20 00 50
Email: ola.johannessen@nrsc.no

Leonid Bobylev
Nansen International Environmental and Remote Sensing Center
Korpusnaya str. 18,
197110 St. Petersburg, (Russia)
Tel.: + 7 812 235 7493
Fax.: + 7 812 230 7994
Email: nansen@sovam.com

Dr Neil McIntyre
Earth Observation Sciences Ltd.
Farnham Business Park
Farnham, Surrey (UK)
Tel.: +44 1252 721444
Fax: + 44 1252 712552
Email: Neilm@eos.co.uk

Dr Ralph Cordey
GEC-Marconi Research Centre
West Hanningfield Road
Great Baddow
Chelmsford
Essex CM2 8HN (UK)
Tel.: +44 1245 473331
Fax.: +44 1245 475244
Email: Ralph.Cordey@gecm.com

Acknowledgments

Illustrative material was provided by the following organisations: US Ministry of Interior Minerals Management Service, Danish Meteorological Institute, Nansen Environmental and Remote Sensing Center, Rand McNally & Company, Earth Observation Sciences, Ltd., GEC-Marconi Research Centre
ESA gratefully acknowledges all contributions

Keywords: ESA European Space Agency - Agence spatiale europeenne, observation de la terre, earth observation, satellite remote sensing, teledetection, geophysique, altimetrie, radar, chimique atmospherique, geophysics, altimetry, radar, atmospheric chemistry