Monitoring the marginal seas of Europe would have been an undertaking far larger than the small project team could have accomplished if enough time was also to be devoted to studying the process of extracting relevant pollution related information from the data. Concentrating on the techniques required meant that the project was focussed on test sites where differing hydrodynamic and anthropogenic influences could be used to provide a diverse set of related problems to be investigated by the team. A strategy based on the selection of three test sites, in the Baltic Sea, the North Sea and the north-western Mediterranean, was adopted. Early in the project, the precise boundaries and pollution related objectives were identified and the work in this section introduces and describes key elements from that programme.
Before the detailed data acquisition planning or the hydrodynamic modelling could begin, the first priority of the project was to define objectives and boundaries on the work in each test site, taking into account the characteristics and problems of each area.
The overall principal objectives of the Clean Seas project applies to the work carried out in the Baltic Sea test site. The almost fully enclosed large non-tidal brackish water body of 390 000 km2 is influenced by both the natural conditions as well as the 80 million people living within the large drainage area (2 million km2). The exploitation of different satellite sensors provides a better understanding of natural and man-made pollution in the Baltic. Phosphorous inputs have decreased during the past decade, while nitrogen inputs are difficult to quantify due to their dependency on river run-off, atmospheric deposition and fixation by cyanobacteria, and thereby display a considerable year-to-year variation. In the Kattegat, Baltic Proper and Gulf of Finland nitrogen is generally the limiting nutrient, and additional inputs of nitrogen due to human activities or natural causes will lead to increased algal growth in these areas. The extensive summer bloom of cyanobacteria (blue-green algae) is very hard to monitor by conventional means, while satellite remote sensing techniques provide such a method.
The objective of algal bloom monitoring in the Baltic Sea can be subdivided into
The distribution of algal blooms, and especially the surface accumulations of cyanobacteria have been one of the single most dominant tasks in the Baltic Sea. The efforts have aimed to reveal the spatial dynamics of the blooms, i.e. areas affected and the size distribution as well as the temporal variability on different time scales. By using data from various instruments, the monitoring capabilities of different satellite sensors have been tested. Several instruments are known to be able to detect algal blooms, and especially highly reflective surface accumulations. On the other hand, in a routine monitoring scheme, few satellite-sensor systems will supply good enough data to be truly useful by themselves. An important aim has been the study of combining different sensors to enable better (in several ways) monitoring capabilities. Apart from the detection and monitoring capability, the time lag between data input and information output have also been considered. This time lag depends on data type (processing needs) and availability of data. Even though the objective of the project not has been to provide near real-time monitoring of algal blooms, such work has been carried out. The conclusions and recommendations for further work together with direct results on algal bloom dynamics is the direct outcome of this work.
Under the MARPOL 73/80 convention, dumping of oil in the Baltic Sea is illegal and therefore observations based on the ERS-2 SAR data presented a valuable opportunity for investigation. The distribution of spills in the Baltic has been studied on both the spatial and temporal scale. The investigation has revealed information on areas affected and the size distribution. The temporal variability (day/night and seasonal) of oil spills were studied to detect patterns, trends and supply good background statistics.
As well as direct observations of pollution events, supporting information has been available from the Earth observation data, in particular relating to the dynamics of the area. An example of such a dynamic event was the very high rainfall and subsequent flooding of large areas in Poland, Germany and the Czech Republic during the summer of 1997. As the flooding resulted in large amounts of river transported sediments being carried to the Baltic, remote sensing techniques should be able to detect the plumes.
During an event such as an ongoing algae bloom, several occasions occur when satellite data will be lacking, either due to atmospheric or light conditions, or due to too infrequent registration of the sea surface. Therefore modelling has been a key component within the frame of the project. In the Baltic the main objective has been to help the analysis of the toxic algae bloom dynamics.
The North Sea test site was selected as a region where river discharges could be seen from satellite imagery to have an influence on the local sea conditions. Figure 4-1 shows a cloud free image of the region, where the Rhine discharge into the continental coastal waters can be clearly seen. These waters remain as a distinct water mass around the west and north coasts of the Netherlands, continuing east to Denmark. As a major source of freshwater input to the North Sea, the impact of industrial or other pollution within the Rhine could have major consequences for the southern North Sea. Although there are no areas classified as environmentally sensitive, and no coastal sites of special scientific interest along this stretch of coast, the area acts as a spawning ground for several species of commercially important fish species (North Sea CD Atlas). The area further north, around the Wadden See is an internationally recognised habitat for bird life. As an indicator of the types of pollution that the Clean Seas approach of combined remote sensing and modelling might tackle well, it was decided to look at sediment dispersal patterns from the Rhine discharge. As contaminants such as heavy metals are generally associated with sediments, tracing sediment loads could aid in tracing heavy metal contamination.
Figure 4-1. ATSR image of North Sea showing Continental Water Mass
As the most complete and regular remote sensing data source, test periods were selected from the time series of AVHRR images to cover relatively cloud free periods. This was found to be a major limitation on available time periods. The AVHRR data, with higher temporal sampling frequency, is more suitable to use for model boundary conditions than the ATSR data. However this leaves the ATSR data available for validation of the model results, being taken at different times to the AVHRR.
In order to verify the sediment distribution output from the model, the "chlorophyll concentration" data from MOS and SeaWiFS were to be used. In turbid coastal waters such as those along the Dutch coast, the chlorophyll algorithms are corrupted by high sediment loads and become indicators of suspended sediment. In conjunction with the visible channels of AVHRR and ATSR, they give an indication of the sediment load, although this is uncalibrated.
It was also hoped to include at least one SAR image in each test period. Although not used for model assimilation, it was hoped that the model output could be used to in aid in interpretation of the SAR images.
The first period chosen, early in the project when few images were available, was a short period from 25 March - 7 April 1997. During this period there were 3 fairly cloud-free ATSR images. On 24 April, there was a SAR scene showing clear indication of the Rhine Plume. This time period was used principally as a ‘test-bed’ for the model. As the temperatures from AVHRR were poor, due to the high levels of cloud, it was intended to use this period simply to verify the model dynamics for the region.
The second test period, 1-20 August 1997, was the principal period of interest. ATSR images from 4, 7, 12, 17 and 18 August showed the development of fine plume structures close to the Rhine outflow, apparently associated with the outflow, and possibly caused by tidal pulsing of the Rhine Plume. Unfortunately there were no colour data available for this time period. A SAR image on 11 August also showed interesting features along the western edge of the continental water mass.
The final test period, 20-31 May 1998 was chosen late in the project, once SeaWiFS data became readily available. There was a time series of daily SeaWiFS imagery of this period with which to compare the sediment load data output from the model.
Figure 4-2 and Figure 4-3 give a summary of the situation relative to the risk of pollution in the Mediterranean and the Gulf of Lion, as explained in detail in the Clean Seas first annual report (Clean Seas , 1997). Regarding this aspect, the Gulf of Lion has a peculiar situation that relies in three points:
(i) it is the place where the Mediterranean deep waters are made during autumn and winter under the surface action of permanent winds
(ii) the environmental pressure of the inshore and offshore industrial activity is intensive, mainly under the actions of the Barcelona and Marseille zones and the Rhone river effluents. Offshore, the Tarragona offshore oil exploitation area is one of the most important in the basin
(iii) the orientations of the main ship routes the intensity of the traffic are representative of what can be found in the whole basin.
The deep water production is associated with complex mechanisms of air-sea interactions, and of circulation of the water masses, including surface and coastal waters in the Gulf. The state of knowledge about these phenomena had first to be reviewed in the existing literature. The motivation for such attention was two-fold. The deep water production area, also named the ‘blue hole’ must be protected against the risk of any major polluting scenario, given that in such a case the polluted waters would first sink beyond any reachable depth for proper response, then travel for decades though all of the basin, and ending on coastal zones, from time to time, with the help of the upwelling processes. The second motivation was that its complex dynamics (e.g. meandering of coastal currents), associated with numerous instabilities (such as on thermal fronts), is far beyond the capacities the contemporary hydrodynamic modelling techniques.
Regarding all of these aspects, the objectives for the site were:
(i) to assess the state of the knowledge on the area
(ii) to monitor the anthropogenic polluting activity in terms of coastal effluents and offshore spilling
(iii) to monitor the natural processes on hydrodynamic processes and biological activity. The availability of the collected data and analyses had to be ensured, through the development of proper tools (GIS, websites, etc.).
