The ERS contribution to Oil Spill Monitoring - From R&D towards an operational service - J.P. Pedersen, T.Bauna, L.G. Seljelv, L. Steinbakk, R.T.Enoksen Tromsø Satellite Station, N-9291 Tromsø, Norway Phone : +47 776 00250 Fax: +47 776 00299 e-mail: janp@tss.no web site: www.tss.no Abstract A project for utilisation of ERS-1 SAR data for detection of oil spill at sea was started in 1991 as part of the national ERS-1 programme. The service development has undergone a lot of phases, and since 1998 the service has been provided on an operational basis from TSS. The main objective of the service is near real-time detection and early warning of possible oil spills at sea for national pollution control authorities and for offshore oil companies. Information on possible oil spills is provided to the customers within two hours after data acquisition. TSS is operational 24 hours a day, and both day and night passes are analysed. ERS SAR data was the first data source applied for the development work. The reduced data availability from ERS limits the use of these satellites for operational purposes. RADARSAT-1 is now the primary satellite for this service. The improved data availability from ENVISAT is expected to contribute to the utilisation of ENVISAT data for the operational service from TSS. 1. INTRODUCTION A project for utilisation of ERS-1 SAR data for detection of oil spill at sea was started in 1991 as part of the Norwegian Space Centre s (NSC) national ERS-1 programme. The service development has undergone a lot of phases. During the first years the focus was on basic research related to documentation of the capabilities and the limitations. Tromsø Satellite Station (TSS) took over the responsibility in 1994 when the project entered the demonstration and pre-operational phase. Since 1998 the service has been provided on an operational basis from TSS. The main objective of the service is near real-time detection and early warning of possible oil spills at sea for national pollution control authorities and for offshore oil companies. SAR image data from the ERS satellites were used during the first phases to document the SAR detection capabilities and limitations and to develop and demonstrate the operational concepts. Two important factors that have contributed to the success are the user contribution even from the beginning of the development as well as the involvement of R&D institutions. The Norwegian Pollution Control Authority (SFT) has been involved since 1991, first through presentation of requirements and now as a paying customer. During the first years the development project was administrated by the Norwegian Defence Research Establishment (FFI), and their knowledge on SAR imaging has been the basis for the developing the service quality. 2. CUSTOMERS AND THEIR REQUIREMENTS The background for this project was the assumption that the SAR instrument could be capable to detect oil pollution at the sea surface, and that this capability could be of benefit for the pollution control authorities to for pollution monitoring. Special focus was from the beginning given to the ocean areas in Northern Europe with important ship lanes and offshore oil industry activities. The pollution control authorities in some North European countries were identified as the most likely customers for using radar satellites for pollution monitoring, and some of the authorities were in turn approached by the project team. The objective was to engage the pollution monitoring authorities into the project, both for demonstrating the pollution detection capabilities and to assess their requirements for operational use of the satellite technology. The national pollution control authorities can roughly be divided into two main categories. The first category requires information about a possible oil spill in 1-2 hours after data acquisition. Typical information includes spill position, time, extent, source identification, and a probability estimate for being a real spill. The customer is then responsible for the verification, especially through application of additional information and observations. The second category receives EO
data and performs the analysis and interpretation by themselves. For the pollution control customers, the satellite based information is of importance also for pre-flight planning obtaining optimal utilisation of surveillance aircraft. During the last few years, the offshore oil industry has become a customer group for this service. The are several reasons for this development: focus on environmental protection and greenness a need for surveillance of automatic offshore oil production installations requirement from national pollution control authorities to oil companies to contribute actively to pollution monitoring. 3. APPLIED DATA The SAR data from the ERS-1 satellite was applied during the initial R&D phases from 1991 and onwards. The initial detection capabilities and limitations were established by using the ERS-1 data. During the initial phases the SAR Lowresolution (200 meter) images were produced at TSS and sent via Internet to FFI for analysis. The first request for explanation of a suspicious feature close to an offshore oil rig was sent from the Norwegian Pollution Control Authority (SFT) to an offshore oil company was sent very early in the project phase. The project established contacts with other pollution control authorities e.g. in the Netherlands, and ERS-1 SAR data was distributed to the Dutch Pollution Control Authorities under an ESA AO-project. ERS-2 SAR images were introduced from TSS into the service development after the launch of the satellite. During the first periods after launch ERS-1 and ERS-2 were cross-compared in terms of oil spill detection capabilities. These comparisons showed that the performances of the two satellites were identical for these purposes. During the intermediate periods from 1994 until 1998 a large number of ERS-1 and ERS-2 SAR data has been acquired at TSS and applied for the service and market development. The availability of these data also represented the basis for developing and demonstrating pre-operational service infrastructure. These developments have generally been done in close cooperation with the anchor tenants involved in the project. Various man-made and natural oil spill like features have been observed in the ERS SAR images acquired, processed and analysed by TSS. Especially data from the North Sea and the Baltic Sea areas have been applied for these purposes. Important results form this work can be summarised as: At low wind conditions, e.g. at wind speeds below app 2 m/s, areas of no wind occurs as dark features in radar images and can hardly be discriminated from an oil spill. At high wind conditions, e.g. at wind speeds above app 12-13 m/s, an oil spill will hardly be visible in a radar image due to the wave conditions causing mixing of oil and water. The ERS radar is capable to detect oil at the sea surface at wind speeds beyond what initially was anticipated. Important strategies of the project were to assess the market development in parallel with the technical development, and to develop the pre-operational service infrastructure based on user requirements. The near routine availability of the ERS SAR data from TSS was an important contribution to the success of this strategy. A large volume of data was available at very favourable conditions, and the pre-operational concepts could be developed and demonstrated to the anchor tenants. The near routine information about possible oil spills derived from the ERS SAR images was in turn applied in the customers operational surveillance systems. During this phase, the knowledge among the customers concerning the capabilities and limitations were derived, and in turn this resulted in updated customer requirements. Special focus was made from the customers on frequent coverage and access to the data and derived information in near real-time. Near real-time means in this context within 1-2 hours after the data has been acquired. One of the important experiences now was that the experimental ERS satellites were not optimal for entering into the full operational phase. The infrastructure at TSS fulfilled the near real-time requirement on data and information provision, but the problem was the limited data access. The narrow ERS swath width of 100 km in combination with the limited operational capability per orbit result in a coverage frequency even at the high latitudes that can not meet the customer requirements in terms of revisit frequency. The Canadian RADARSAT-1 satellite was launched in Autumn 1996, and through its position as a network station and data distributor TSS can apply the SAR data from this satellite for the oil spill service. Compared to ERS, RADARSAT-1 operates SAR imaging modes having a swath width of 300 km, which in combination with the improved operational capability gives an increased repetitive coverage. The satellite is also
operated based on customer requests. In total, the availability of the SAR data from RADARSAT-1 contributed to the operationalisation of the oil spill detection service provided by TSS. Data from ERS is, however, still applied as a supplement to the RADARSAT-1 data to improve the total coverage. The more favourable data costs of ERS also makes these data more suited for service demonstrations for new, potential customers. Service costs have also been a concern among the customers. TSS has therefore invested in cost-benefit studies where the unit costs of the satellite service has been compared with the costs of the traditional monitoring systems. These studies have shown that the unit costs e.g. per square kilometre for a satellite service is even below the corresponding costs of the traditional service. Typical average square kilometre costs of an aircraft based service is app. 0.08 Euro and typical 0,08 Euro for a satellite based service. The key to success is, however, not to compare the two different service in a competitive approach, but rather to focus on the benefits from an integrated service utilising both elements. 4. SERVICE CONCEPT The core elements of the service rely upon the RADARSAT and the ERS SAR facilities at TSS. The radar processors are capable to generate a radar image in less than 10 minutes after data acquisition. It has previously been documented that both RADARSAT and ERS SAR images having a reduced pixel spacing of 100 meters is very suitable for detecting oil at the sea. Figure 1: ERS-2 (left) and RADARSAT (right) SAR images of a confirmed oil spill at an offshore oil field in the North Sea. Image size approx. 70 x 70 km. Copyright ESA, TSS, 1996, CSA/RADARSAT, 1996. A dedicated service workstation is applied for information extraction, presentation and distribution. This workstation utilises integrated use of SAR data, wind model data and geographical information during the analysis and interpretation. In addition, the workstation includes functionality for message generation and distribution, and for service database handling. The radar capabilities to detect oil relates to the dampening effect from oil to the surface capillary and short ocean waves. The radar backscatter is reduced in areas with oil present, and an oil spill will hence occur as a darker feature on a brighter background in the radar image. The capabilities and the limitations of satellite radar data to detect oil spills at sea with a certain probability of being a real spill have been evaluated since the availability of ERS-1 SAR data in 1991. Figure 1 presents two image examples on oil spill at sea. The left image is an ERS SAR image of an offshore oil field in the North Sea. This image was acquired on August 22, and shows platform and a dark feature that was classified as a medium/high probability for being a real oil spill. On the following day a RADARSAT ScanSAR Narrow_Far image was acquired over the same area. The large dark feature is the possible oil spill observed on the preceding day. Notice, however, the different forms of the features at the two times.
The primary service analysis area covers both the Norwegian coastal waters and more central European coastal waters within the acquisition area of TSS. Figure 2.a shows the conceptual view of the service data and information flow from TSS to the customer, from acquisition and processing of SAR data at TSS, via image analysis and interpretation utilising the workstation, followed by early warning alert to the customer. The available data is analysed and the customers informed about identified high and medium oil spill candidates, either by using fax/phone and/or a dedicated PC based tool, SARA. The distributed information includes the time, position, and the probability of the identified feature being a spill. The position identifier includes spill centre position, length, width and orientation in four directions (N-S, E-W, NE- SW, NW-SE). If a vessel (and/or any other potential source) could be identified, the position will be included in the message. Quick-look images and related interpretation information of identified high and medium oil spill candidates are distributed using the SARA tool. SAR Data Pre-processing Service Workstation - Wind/geographical info. - NRT Screening - Early W arning - Data transmission TSS Operations Surveillance Aircraft Operations Early W arning Alert SARA Customer Site - Descision process - Verification - Monitoring - Combatting Figure 2: a (left). Oil spill service operational concept. b (right): Coordination of aircraft operations with satellite coverage. The aircraft flight track is adjusted according to the satellite track, and the sequence of the data processing at TSS is adjusted according to the aircraft position. The customers and the aircraft operators are routinely informed about satellite acquisition schedules. For the operations in Norwegian waters, TSS receives the flight plans and a best co-ordination between aircraft operations and the satellite acquisition and processing scheduling is sought (Figure 2.b). 5. CONCLUSIONS A national R&D project focusing on the capabilities of the ERS-1 SAR instrument for detection of oil spills at sea was initiated in 1991. This project later got a status as an ESA AO/Pilot Project. This project has undergone multiple phases, leading to the operational oil spill service now being provided by Tromsø Satellite Station (TSS). The main strategies of the development activities was to develop pre-operational service infrastructure based on real customer requirements in parallel with market development. ERS-1 followed by ERS-2 SAR data were the first data sources applied in the development work. These data has clearly documented the capabilities and the limitations concerning oil spill detection at sea. The ERS data was also important both for the infrastructure and the initial market development. As the customers gained experience with the satellite technology they were able to update their requirements also based from integration of the satellite branch into their existing surveillance systems. It was then documented that the ERS SAR data was not the best source for the operational service. Especially the reduced data availability and coverage limits the full operational use. The documented detection capabilities combined with the improved temporal and spatial coverage of RADARSAT compared to ERS meant that RADARSAT represented an important technical contribution towards an operational
development. The key to the successful utilisation of RADARSAT for operational oil spill monitoring depends, however, upon the pricing policy from the satellite owner. Even though the prices e.g. per square km are comparable for ERS and RADARSAT, the customers are concerned about the total high costs for a service based solely on RADARSAT. It has therefore become very clear that a pricing policy giving incentives to the operational, large volume customers need to be realised in order to fully utilise the potential of RADARSAT and future radar satellites for operational monitoring. The availability of the ERS SAR data in 1991 initiated the development work which now has led to a fully operational service at TSS, though using RADARSAT SAR data supported with ERS-2 SAR data. The expected availability of the ENVISAT ASAR Wide mode data is expected to increase the ESA data contribution to the operational service. ACKNOWLEDGEMENTS The results presented in this paper were originally obtained under the Norwegian Oil Spill Detection project, which now has resulted in the operational service at TSS. The first project phases were funded by the Norwegian Space Centre, ESA, TSS and the customers. Also a special thank to the scientists at FFI for their significant contribution to the service development. REFERENCES J.P. Pedersen, F. Landmark, T. Bauna, R.T. Enoksen, "Oil Spill Detection Activities in Norwegian Waters", TSS Report TSS-OLJ-1627/JPP-96, 11.11.96, Paper presented at RADARSAT Workshop 19-21.11.96. T. Wahl, Å. Skøelv, T. Anderssen, J.P. Pedersen, J.H. Andersen, O.A. Follum, G.D. Strøm, T-I. Bern, H. Hamnes, R. Solberg, "Radar satellites: A new tool for pollution monitoring in coastal waters", Coastal Management 24:61-71, 1996. J.P. Pedersen (Editor), "Utilisation of ERS SAR Data for Near real-time Detection of Oil Spills at Sea", TSS Report TSS-CEO-008/JPP-96, February 1996, Executive Summary from Final Report to the EU/CEO Applications Proof-of-Concept Study. J.P. Pedersen, L.G. Seljelv, T. Bauna, G.D. Strøm, O.A. Follum, J.H. Andersen, T. Wahl, Å. Skøelv, «Towards an Operational Oil Spill Detection Service in the Mediterranean? The Norwegian Experience: A Pre-operational Early Warning Detection Service Using ERS SAR Data», Spill Science & Technology Bulletin, Vol. 3, No 1/2, pp. 41-46, 1996. J.P. Pedersen, et al, «Norway Enhances Oil Spill Detection with RADARSAT», Earth Observation Magazine, Vol. 6, No. 5, May 1997. S. Navrud, «Oil Spill Monitoring in the Baltic Sea Using Earth observation Data», ENCO Environmental Consultants a.s. Final report September 1997, Revised February 1998. J.P. Pedersen, L.G. Seljelv, T. Bauna, et.al, «EUROIL - Environmental monitoring by Using Radar Satellites for Oil Spill Detection», Proposal prepared for the Commission of the European Communities DG XII/D, TSS Report October 1997 G. Engen, K.A. Høgda, H. Johnsen, " A new method for wind field retrieval from SAR data", NORUT IT project publication (In preparation). J.P. Pedersen, "Near real-time detection of oil spills: Annual Report 1997", TSS Project Report TSS-OLJ-0248/JPP-97. J.P. Pedersen, T. Bauna, L.G. Seljelv, F. Landmark, R.T. Enoksen, J.H. Andersen, «Earth Observation a Cost-effective Improvement for Oil Spill Monitoring», Paper presented at the 27 th International Conference on Remote Sensing of the Environment, Tromsø, June 1998.