Validation and verification activities in OCEANIDES give an understanding of the relative performance of various oil spill detection approaches in satellite images and of the number of false positives and negatives one could expect, under favourable meteorological conditions, in radar satellite images. Algorithms and procedures for detecting slicks from Synthetic Aperture Radar (SAR) satellite imagery have been benchmarked against each other and against observations from aircraft in order to arrive at a common definition and estimation of false positives and negatives. A clear procedure has been set-up to calculate in a uniform way the probability that a shape seen on an image is indeed an oil slick and, conversely, to define what meteo conditions and SAR mode combination make it likely that a real oil slick would be undiscovered. Two different campaigns aiming at verifying oil slicks observations in satellite images by aircraft were executed. During both campaigns, aircrafts were sent to examine slicks observed in RADARSAT and ENVISAT images downloaded at Tromsø Satellite Receiving Station and analysed in Near Real Time mode. In most cases it was possible to transmit the slick position to the aircraft within an hour from the acquisition. The number of false positive increases, as one might expect, as the operator s confidence diminishes. On the other hand it should be noted that in case of high confidence level (i.e. when all the relevant peculiar characteristics of an oil slick as detected in a radar image are present) the false positive rate is virtually zero. Despite the low number of cases to be analysed and the fact that additional activities might need to be continued in the future in a systematic way to reach firmer conclusions, OCEANIDES has to be considered as an important step forward in the efforts to provide a clearer view on the reliability of satellite observation of oil slicks. The process OCEANIDES has launched needs to be continued and further extended in a systematic way since it is instrumental to establish an operational routine use of satellite observations.
The implemented database collects most of the existing observational data, at European level, concerning oil discharges from vessels. Data are coming from observations made by aircrafts as well as using satellite radar images. It is a tool that, for the first time, put together in a coherent way a large amount of data, which were collected and organised by different entities and according to different standards. The achieved results not only provide the possibility to access all the data from a single source but has also defined a set of common criteria for inclusion in the database. This is the fundamental step to allow a real, unbiased use of the data for a variety of different applications. In addition, a specific statistical methodology has been implemented to deliver added value products. The tool is now a unique source of information in order to perform different tasks, such as the long term monitoring of the evolution of marine oil pollution at European scale, the identification of hot-spots related to the problem, the assessment of the effectiveness of measures taken in specific areas. The JRC is fully committed to further maintain, develop and populate the database and it is currently using the database as a starting point for local/regional studies on the evolution of the marine oil pollution, on request of National Authorities and EC Services. The database is publicly accessible through a viewer on a dedicated web-site and additional information, in GIS compatible format, can be requested.
The work done on the aspect of environmental impact of oil discharged includes the adaptation of the oil-spill/trajectory model to the hydro-dynamical model and dedicated particular efforts on an appropriate procedure to effectively deal with the uncertainties concerning the volume and type of discharged oil. The chosen approach, introducing the concept of life-time of a spill, allowed running a large number of sensitivity calculations. The performed simulations referred to different and realistic scenarios for the North and the Baltic seas and provided some concrete figures of the percentage of oil volume which may reach the shore. The method is based on a statistical approach where a large number of spills have been randomly released and tracked for a given period of time. The tracks of the oil spills have been simulated using velocity fields generated by an advanced hydrodynamical model covering the entire North Sea and Baltic Sea area with a grid size of 3nm. In the vertical the model has used zooming towards the surface to better resolve the surface velocity field. From the hydrodynamic model simulations surface fields have been extracted at hourly resolution and stored in monthly data sets. The lifetime concept has been introduced to overcome the problem of a lack of knowledge about the oil type. Life-times of 6, 24, 72 and 144 hours have been used. Based on the assumptions of the model system described and the simulations performed with this system it can concluded that less than 5% of the randomly released particles actually reach the shore. In all cases, it has been found that only a low percentage of oil can reach the shore. It shall be emphasised that in the study performed the oil-spills are randomly distributed which will not be the case in reality where the spills will be located in the vicinity of the shipping routes. Furthermore, the entire complexity of weathering processes and initial amount of oil has been lumped into a single variable - the life-time of an oil-spill. This has been done because not sufficient input data are usually available. The approach could be applied in specific cases (accidental spills, large deliberate spills) and then provide important information for the decision makers.
The result concerns the development of a harmonised oil spill reporting system capable of integrating, storing and visualising available oil spill data. A system was developed through assessment of current oil spill monitoring and reporting practices in co-operation with those organisations currently involved in this monitoring. Based on this assessment and co-operation, a standardised reporting nomenclature was agreed along with a system to store oil spill data and metadata. Visualisation of the stored oil spill data was enabled using Geographic Information System based web-mapping interface. The design and implementation of this harmonised oil spill reporting system is intended to optimise available information on oil spills, provide a tool for improving the understanding of the nature of the problem of oil spills and enabling the development of further analysis at pan-European level. The first section of the standard nomenclature was designed to include details associated with detected oil slicks, the remaining sections relate to specific information on the satellite and aircraft used, along with details on the accuracy of this information. The nomenclature has been designed to have the potential to record significant detail on detected oil slicks (when available) along with metadata on the source and accuracy of that information. Inclusion of metadata was deemed crucial in order for the user to have some idea of the source and associated accuracy of the information involved.