Science works best when many minds contribute to solving society’s problems. The roots of our Data Cube Services for Copernicus (DCS4COP) project can be traced back many years through a series of multi-national collaborative projects in which the consortium members have worked together and learned each other’s strengths and weaknesses.
European Union funding has been key to enabling this, particularly through the Framework 6 and 7 Space calls from the period 1997 to 2013. The aim of EU science funding is to encourage the movement of novel ideas generated by scientific research to a wider audience, and eventually to the marketplace, for the benefit of society. Through the early 2000s, several members of the current DCS4COP team worked together on developing new techniques for monitoring intertidal areas in the BIOOPTIS and HIMOM projects.
Working with satellite data was in its infancy at that time. Landsat-7 was launched in June 1999 and the images from its Enhanced Thematic Mapper were eagerly awaited by the scientific community, particularly those working on the coastal zone. Searching and ordering images was, however, a slow and expensive process, with a wait of two months for a compact disc containing the selected scene to arrive. Much has changed over the past twenty years, as the blog posts in this series will reveal!
The start of the satellite information revolution can be traced back to two critical decisions, firstly in 2008, the United States Geological Survey (USGS) allowed free and open access to its Landsat image archive. The US administration also committed to continuity of data with the launch of Landsat-8 in 2013, with greatly improved optics and image resolution. On the European side, plans were put in place for an advanced planetary observing system. Initially called GMES for Global Monitoring for Environment and Security, then later Copernicus, this was a highly ambitious programme to provide new satellite capability together with in situ measurements and forecasting products. These initiatives occurred in parallel with a massive increase in computing power and development of cloud storage (e.g. the broader Digital Revolution).
Recognising the enormous power of satellite data, and Europe’s deficit in this area compared to the US, the Commission has used its science funding to speed up development of its own geospatial industries with targeted calls for proposals using Copernicus data:
“Space-based application at the service of European society (developing satellite observation systems and the GMES services for the management of the environment, security, agriculture, forestry and meteorology, civil protection and risk management)”
The Framework 7 call above was seen as an opportunity by a consortium led by Dr Kevin Ruddick at the Royal Belgian Institute of Natural Sciences (RBINS). Kevin put together a balanced team containing specialists in coastal water quality to produce a prototype satellite image processing service for the new Sentinel satellites. At the stage of proposal writing, none of the Sentinels had been launched, so expertise in previous ocean colour satellites such as MERIS, MODIS and Landsat-8 was needed. But a consortium based on image processing alone would not be strong enough – to his core team of RBINS, Brockmann Consult (Germany) and the Flemish Institute of Technical Research (VITO), Kevin added a strong set of in-situ data providers: the Norwegian Institute for Water Research (NIVA), the Centre for Environment, Fisheries and Aquaculture sciences (Cefas, UK) and Oceanographical Observatory at Villefranche (LOV, France). The University of Hull joined the consortium in 2015.
The common mission of all partners was to seek end-users of the project’s satellite data, and to develop prototype services for their users, from which feedback could be gathered. The project was submitted as HIGHROC, for High spatial and temporal Resolution Ocean Colour Products and Services, and it was funded for four years under the 2013 SPACE call.
Image HIGHROC partners
The first tasks of the HIGHROC project were to gather user requirements. Each partner had one or more users in the maritime sector: from engineering consultancies and industries such as dredging, through to government agencies with responsibility for water quality. Not all requests could be met, but a master list of water-related parameters was selected by the team to form the basis for satellite image processing (Table 1).
Table 1: A list of water quality parameters developed during HIGHROC. Future blog posts will address individual parameters and their uses in more detail. Green: special algorithm developed by HIGHROC; Pink – existing algorithm available; Grey: not available.
The strength of HIGHROC was to concentrate on a core set of satellite-based products, developing new algorithms where necessary, and to validate each product against a very large collection of in situ data (Table 2). Joint cruises and shore-based inter-comparison exercises within the project were regular events in HIGHROC, and helped the team to identify areas of best practice in measuring below the sea surface (Figure 1).
Figure 1: HIGHROC inter-comparison exercise at NIVA in 2015 in which spectral radiometers are aligned to a common light source.
Table 2: In situ ocean-observing systems operated by the HIGHROC consortium and used to validate satellite data products
At the finishing point of HIGHROC, the project had a streamlined set of data-processing chains running on the high-performance computer systems of VITO, RBINS and BC. This enabled the products to be made available to users during the final year of the project.
In parallel, Starlab had participated in the past with several members of the consortium in both ESA and EU Framework 7 Space calls, as MarCoast (Marine & Coastal Environment Information Services, 2005) or AquaMAR (Marine Water Quality Information Services, 2009).
The wide experience in user management, service provision and market development made Starlab join the whole consortium in 2017, and thus completing the team of DCS4COP project.
Our ambition to operationalise service provision, expand its functionality to cater various use cases, maintain the great expertie in the HIGHROC team, and to establish a commercially viable business case has been the starting point for DCS4COP. The Data Cube approach adopted in this current project offers a flexible and powerful solution, which also allows for generalisation to other thematic areas. The further development of the processing environment, and our move towards a Data Cube model, is the starting point for DCS4COP.
Author: University of Hull