SmartICE – Sea-ice Monitoring And Real-Time Information for Coastal Environments
SmartICE seeks to address these information needs using networks of in situ sensors that generate daily observations of changing sea ice conditions at hazardous travel locations, particularly during freeze-up and break-up, combined with user-based satellite image classifications of sea ice state. Novel information technology will adjust data interpretation and visualization according to user needs and preferences.
SmartICE is a community-university-industry collaboration that integrates adapted technology, remote sensing and Inuit Knowledge to promote safe travel for all stakeholders in northern coastal environments. While community participation in SmartICE is key to addressing local needs and conditions, SmartICE is intended to augment and integrate Inuit sea ice knowledge, not replace it.
The SmartICE system is a climate change adaptation tool and technological innovation of ArcticNet’s Nunatsiavut Nuluak project, which is addressing Inuit concerns about the impacts of climate change and modernization on communities and environments of Northern Labrador. It is also linked to the Nunatsiavut Government’s initiative: SakKijânginnatuk Nunalik – Building sustainable communities in the coastal subarctic – understanding the risks and developing tools and best practices for climate change adaptation in Nunatsiavut. University partners include Memorial University and Trent University. SmartICE is partnering with C-CORE, for remote sensing expertise and the Canadian Ice Service.
The variables of interest for navigation over sea ice are ice thickness, concentration and roughness. For example, safe travel for snowmobiles over sea ice requires continuous, smooth ice that is at least 15 cm thick. The main elements of a SmartICE information system are:
A) A network of in-situ sensors that measure sea ice thickness, and other community/industry defined variables, at designated locations and transmits daily data to a central server.
B) Repeat satellite imagery from which sea-ice surface conditions (e.g., concentration, roughness, water content) are mapped following user-defined classification systems.
C) Information technology that integrates in situ and remotely sensed sea ice data to generate raw and processed digital products that match the needs of user groups, from ice navigation managers to Inuit ice experts to recreational ice users.
To obtain sea ice thickness (and possible other) measurements there will be a suite of simple, low-cost sensors that are deployed from a boat or aircraft, prior to freeze-up and allowed to freeze into the ice. This unique feature means they remain in the ice all season and operate through the most dangerous periods for over-ice travel. In addition to ice thickness, sensors can measure other ice and environmental parameters identified by user groups. Data are recorded regularly but averaged and transmitted once a day via satellite to a central server. Measurement uncertainty will also be communicated as part of the dataset. Ideally the device will be maintenance-free during the monitoring season (e.g. sufficient battery power). A key requirement of these sensors is that they be sufficiently affordable to permit multiple unit deployment along traditional sea ice routes. They are not, however, designed to be expendable and ideally will be tracked and recovered after sea ice break-up.
User-based sea ice classification of satellite imagery
Each sea ice season there will be a series of Synthetic Aperture Radar (SAR) satellite images acquired for monitoring the evolution of sea ice at selected locations of interest. SAR is a microwave remote sensing technology that produces images based on the amount of backscatter reflected from a target surface. It can be exploited for sea ice mapping and classification and, because SAR uses an active microwave system, imagery can be collected in all weather conditions, day and night, at a range of spatial resolutions and scales. In close collaboration with partners, SmartICE will define user needs in terms of ice classification for safe travel on the ice (e.g. smooth vs. rough ice, dry vs. slushy ice, mobile vs. pressure ridged ice), record field observations of hazardous ice types with the support of community ice monitors, establish image characteristics and processing techniques that best discriminate different ice types on satellite imagery, and develop a sea ice catalogue that combines user ice types and image classes to address ice travel information needs.
Data management and visualization system
Another key collaborative component of SmartICE is the development of an information technology and user interface that integrates the fine-scale sensor network data and the broad scale classified satellite data to generate a range of user specified products. It is anticipated that basic features such as floe edge location and sea-ice types will be made available for each study location; hazardous ice maps, however, would be generated for self-identified users, depending on travel type (snowmobile vs. ship), user experience (professional ice harvester vs. recreational snowmobiler), trip length (long-distance trips may evaluate recent sea ice changes and trends) and other factors identified by user groups. Conceptually the data will be presented in layers allowing different data fields to be viewed over a base map of the study location.
Fig 1. Figure showing SmartICE inputs (upper plot) and simulated output (lower plot). The upper plot shows a SAR image (Cosmo-SkyMed, HH polarisation, StripMap acquisition mode: 3-5 m resolution, approximately 40 km x 40 km area of interest) acquired over Nain, Labrador and associated sea-ice thickness measurements (magenta triangles annotated with station identifiers and the thickness measurements). The image and thickness measurements were acquired on March 27th, 2013 as part of preliminary SmartICE fieldwork. Note the track of the Umiaq bulk carrier vessel that travels to Voisey’s Bay can be seen at bottom of the SAR image close to station ST124. In addition, snowmachine tracks can be seen criss-crossing the bay in front of Nain. The lower plot shows a simulated SmartICE thickness map generated from the input data.
Thank you to our research partners and funders for their contribution to this project: