CoastColour Global Full Mission Dataset and Spatial high Resolution Case Study
Brockmann, Carsten1; Brotas, Vanda2; Doerffer, Roland1; Pinnock, Simon3; Ruddick, Kevin4; Santer, Richard5; Sathyendranath, Shubha6
1Brockmann Consult GmbH, GERMANY; 2University Lisbon, PORTUGAL; 3ESA ESRIN, ITALY; 4MUMM, BELGIUM; 5Adrinord, FRANCE; 6Plymouth Marine Laboratory, UNITED KINGDOM

The MERIS Full Resolution data archive spans from mid 2002 until April 2012 and comprises more than 130TB of raw data. In 2012 this complete archive has been processed with the dedicated coastal processing developed in the CoastColour project, an ESA Data User Element (DUE) project. The dataset covers all coastal zone of the Earth.

During the first years of the CoastColour project the algorithms for atmospheric correction and retrieval of inherent optical properties, water constituents (chlorophyll concentration, suspended sediment and yellow substance) have been developed and extensively tested. They were applied to the 27 site of champions users and for the time frame 2005 - 2012. In 2012 the algorithms were refined and the complete archive, now covering the complete mission time range and all global coastal areas, has been reprocessed.

The key elements of the CoastColour algorithms are:

  • the atmospheric correction and in-water retrieval are based on a coupled ocean-atmosphere radiative transfer model
  • the parametrisation of the model includes clear water up to very turbid waters with Chlorphyll concentrations up to 100mg/m3 and suspended sediment concentrations up to 1000g/m3
  • the atmospheric correction was successfully validated even above very turbid waters which are highly reflective in the near infrared
  • the in-water retrieval is combined with an optical water type classification

    The CoastColour algorithms are transferable to Sentinel 3 and preparatory work to transfer them to Sentinel 2 has been started. A dedicated study has been performed to apply the algorithms to current sensors. Data of two sites (Chesapeak Bay, USA, and Korean coast) are being studied using special campaign data from SPOT-4 during its de-orbiting in spring 2012, accompanied by other data, such as historic Landsat 5 and possibly RapidEye and LDCM data. These instrument provide comparable spatial resolution but significantly worse spectral characteristics (number of bands, position of bands, saturation radiance, SNR). However, at the time of abstract writing it is expected to demonstrate the possibility of retrieving suspended matter concentrations from these data streams and to follow gradients and fronts at this spatial resolution close to the coast. All CoastColour data are freely available from the CoastColour web ( Access is provided interactively through a MERCI interface, and in an automated way through ftp and openDAP.