The accidental release of crude oil into the ocean environment is of increasing concern because of the damage to marine and coastal wildlife as well as its adverse impact on humans that depend on the sea. The BP spill in the Gulf of Mexico has made it obvious that, as the demand for oil expands and drilling takes place in technically challenging locations, the likelihood of accidents will increase. The Arctic is a particular concern because, unlike spills in the warm waters of the Gulf, natural dissipation and dispersion of the oil is slowed by the cold temperatures. Moreover, human efforts at remediation are severely hampered by the extreme weather, by ice cover and by winter darkness. The National Energy Board and the CBC state that Canada is not set to start offshore drilling for arctic oil until 2014. However in August 2010 an exploration licence was granted to Chevron.

Responsibility for regulation and surveillance of the oil industries and remediation in the event of a spill rests with various Canadian agencies. A brief overview of those agencies is provided in the report Oil Spills and Responsible Canadian Agencies, December 2010.

Arctic oil spills could result in oil floating on the surface of the ocean or it could released into icy waters. Lying under the ice, the oil finds its way into pockets or pools on the ice surface. Where the ice is broken up, it is intimately mixed with the rubble. In warm climates the volatile oil components evaporate rapidly and disappear into the atmosphere but, in the arctic, evaporation is reduced and oil may become trapped under ice. Because the bottom surface of the ice is not flat, oil tends to coagulate in pockets. The heavier components tend to sink and are carried by ocean currents leaving the lighter components behind. Therefore the spill may be spread farther than in the open ocean.

Remediation often involves booms and skimmers. In the arctic, collection is affected by the cold temperature, which renders oil more viscous. Skimmers may become clogged especially when the oil is mixed with ice. Booms are not effective in high winds, which are frequent over arctic waters. Oil burn-off may be difficult in extremely low temperatures.

It is important to detect oil spills early so that remediation can limit environmental damage. Ideally this would involve a systematic program of surveillance designed to provide detection of spills over wide-areas including oil platform locations and shipping lanes. Such programs exist, at least partially, at the national level, for example Canada's Integrated Satellite Tracking Of Pollution (ISTOP) program. This is based on the detection of oily ship discharges by radar satellite, principally RADARSAT-2. In radar images of the open ocean, oil reduces the radar backscatter so that the spill appears dark. The Synthetic Aperture Radar (SAR) provides high resolution images of the ocean with a swath width of up to 500 km.

In the case of discharges from a ship, the ship can often be located in the SAR image and associated unequivocally with the spill. However, this is insufficient for litigation and it is necessary to identify the ship using Maritime Patrol Aircraft (MPA) and collect clear evidence by sampling the spill. Therefore the wide-area satellite surveillance provides a cue for a chain of activities. This reduces the overall costs because satellite SAR surveillance has the by far the lowest cost per square kilometer followed by MPA surveillance and so on.

Canada's National Aerial Surveillance Program (NASP) monitors the Great Lakes and the St. Lawrence Seaway as well as the arctic during the northern shipping season. The program uses Dash 7 and 8 aircraft fitted with side looking radar and various sensors covering the IR and visible spectrum as well as AIS for ship identification. Commercial aircraft augment the fleet for surveillance off Newfoundland.

The SAR detection of oil on the open ocean, which is based on the suppression of the capillary waves responsible for radar scatter, is reasonably well established. However, at the level of an individual pixel it is difficult to draw any conclusions and an operator or automatic detection algorithm must utilize pixels over a significant area to classify a dark streak or patch on the ocean as a spill. In narrow arctic passages, bordered by ice, the detection of oil can become difficult. If oil becomes mixed with ice, the spill cannot be detected by the usual SAR methods.

Arctic ice can be broadly classified into first-year ice and multi-year ice. First-year ice with a thickness of up to 2 m tends to be very saline, while multi-year ice is typically over 2.5 m thick and is less saline. The ice is often covered with snow of thickness 10 to 20 cm. Presently, the only way for oil under or in the ice to be detected reliably is to bore a hole through it and take a sample. Bearing in mind the possibility of extreme sub-zero temperatures, high winds, uncertain ice thickness, shifting and cracking ice, an ongoing operational surveillance program based on this method is not feasible. However, it is possible to detect oil at the ice surface reliably by laser excited fluorescence. The fluorosensor may be airborne and the swath width of current instruments is up to 800 m. The costs of a wide-area surveillance program with this technique are prohibitive.

There is a requirement for new satellite borne techniques for oil detection that can operate reliably day and night, year round, over icy waters. Because of the long hours of winter darkness, passive sensors are not ideal but hyper-spectral sensors may have a role. Active sensors, such as SAR are a possibility but it seems unlikely that a satellite borne fluorosensor is an option because of the laser power needed. Unfortunately the attenuation of radar waves associated with the high conductivity of saline pockets within the body of the ice seems to suggest that detection of oil under the ice may be difficult using space-borne SARs operating at the usual microwave wavelengths of a few centimeters.

Nevertheless there is some potential that SAR could provide useful information about spills in ice. Though previous satellite borne SAR, such as RADARSAT-1, did not result in an oil-in-ice detection capability, the advent of fully polarimetric SAR is more promising. The use of lower radar frequencies, such as L-band or P-band is another option, which would reduce attenuation. A partial analysis is provided in J.K.E. Tunaley, Detection of Oil Spills in or with Ice, LRDC Technical Report, November 2010. Verification of detection performance is contingent upon performing controlled spills. However, this is not popular as described in a CBC report.