Typically, a water surface appears black in an ERS-1 SAR image if the windspeed does not exceed 3 m/s. In regions of higher windspeed where the image still appears dark, the presence of some type of surface film is probable. Although at windspeeds exceeding 10 m/s no effect from natural or artificial oil films on wind-induced surface roughness has been observed to date. Hence oil spill monitoring is limited to the 3-10 m/s windspeed range.
For many years radar surveillence from aircraft has taken place along the coasts of Europe and the US. The operation of such aircraft is however expensive and has a limited coverage. With ERS-1 spaceborne radar monitoring became possible and indeed the first proof of this concept came in the early commissioning phase of the satellite.
An experiment in the Haltenbanken area (Norway) released on three separate occasions between July and August 1991, 20 tons of crude oil (Bern et al. 1993). It was demonstrated from this experiment that under certain wind conditions an oil film could be detected both by aircraft and ERS-1 SAR. On 3 December 1992 the Greek oil tanker Aegean Sea ran aground at Torre de Hercules near the entrance to the Ria de Coruna and La Coruna harbour, Spain. The 291 m tanker of 114 036 dwt - carrying 79 096 tons of Brent-type crude oil - broke up and exploded. Practically all of the oil was released into the sea.
Over 200 km of shore was affected, while some 4000 m³ of oil was cleaned from the sea and beaches. The spill was surveyed daily by aircraft. Distribution maps (of which Figure 4 is a synthesis) show that the oil spread to the North-East until 9 December, reaching Pta. Candelaria. Then it spread along the southern coast, reaching the Islas Sisargas by 10 December.
The prevailing winds initially ranged from southerlies to westerlies, changing on 10 December to range from northerlies to easterlies for four days. In the second half of December and first half of January, wind ranged from South-East to North-East, pushing the oil out to sea. Wind velocities were favourable for oil slick observation at all ERS-1 passes (Figs. 1, 2 and 3).
This case study demonstrates that data from the ERS-1 SAR could be a useful tool for oil spillage monitoring, not only because of independence from cloud cover, but also because of its rapid availability. In fact, a network of Earth stations within Europe can receive 'ready-to-use' image data only a few hours after acquisition at the ESA stations of Fucino, Italy and Kiruna, Sweden .
Of the ESA members it is Norway which is currently pushing towards the utilisation of ERS-1 SAR for a monitoring of its extremely long coast. Having set up a national receiving facility at Tromso and having received fundings from the Norwegian Space Centre, the Norwegian State Pollution Control Authority SFT, Statoil, Esso, the Marine Spill Response Corporation (MSRC) and ESA (in the framework ofthe Prodex programme), an efficient scenario has been realised. This includes a fast processing chain at the acquisition station and delivery to the data evaluation centre (Norwegian Defence Research Establishment) via satellite link or landline. Furthermore, an immediate reporting chain to the end-user, the SFT was established.
As the SFT aircraft is operating daily along the coast, verifications of possible oil spills identified on ERS-1 SAR images were carried out by rerouting the aircraft. During 1993 a collaboration with the Directoraat-Generaal Rijkswaterstaat in the Netherlands was also established and a series of low-resolution SAR images from the Dutch coastal waters was sent for realtime verification. The pilot operation phase of the project ended in November 1993 (Wahl et al. 1994) and was very successful for two reasons:
In the Mediterranean where pollution is apparently much higher and the control widely nonexistent, weather conditions are even better for using SAR data on a regular basis for such applications.
[1] Bern T I, Wahl T, Anderssen T & Olsen R, Oil spill detection using satellite-based SAR: Experiences from a field
experiment, Proc. First ERS-1 Symp, Cannes, France 4-6 Nov. 1992.
[2] Wahl T, Anderssen T & Skoelv A, Oil spill detection using
satellite-based SAR: Pilot operation phase, Final Report, Norwegian Defence Research Establishment, 31 January 1994.
Figure 1: ERS-1 SAR image of 13 December 1992, ten days after the
disaster. From aircraft observations it was concluded that the very dark area is heavily polluted sea, while the dark grey area and
dark streaks are older and more dispersed oil. The oil spread northwards at first, and since 10 December westwards.
Figure 2: ERS-1 SAR image of 1 January 1993. Winds have pushed the oil
westwards as far as the Islas Sisargas spoiling many more kilometres of beaches. It seems that the dark areas in the upper part of
the image are not caused by oil but by low winds, but there is no confirmation of this.
Figure 3: ERS-1 SAR image of 17 January 1993. The scene is dominated by
high swell, perhaps hiding remaining oil, or the slick has largely disappeared. There are some dark areas in the Ria de El Ferrol
and near the harbour of La Coruna. This may simply be caused by sheltering effects of the coast.
Figure 4: Situation plan of oil polluted zones as observed from
aircraft between 4 and 28 December 1992, with dates indicated.
Figure 5: The Aegean Sea tanker in flame.
Figure 6: ERS-1 SAR image of La Coruna region on 8 February 1993,
showing evidence of the oil spill lasting at least till this date.
Figure 7: An oil slick of 10x1.5 Km sighted simultaneously by ERS-1 SAR
and the SLAR of the Dutch Directoraat-Generaal Rijkswaterstaat along the coast on a november day - another proof of the
complementarity of ERS-1 for oil spill surveillance. The wind was very light and therefore natural slicks are well visible (image
size 51x51 Km.)
Photo 1: Protection barrier at the beach in Berbes, La Coruna, 4
December.
Photo 2: Oil spill impact on the ecosystems, Canaval, 4 December.
Photo 3: Oil on the surf at Canaval.
ESA EOQ Nr. 44.