A Benchmark Dataset for Validation of MERIS and MODIS Water Products Using Autonomous Buoy Data.
Vanhellemont, Quinten1; Greenwood, Naomi2; Ruddick, Kevin1
1RBINS-MUMM, BELGIUM; 2CEFAS, UNITED KINGDOM
Ocean colour remote sensing is becoming well-established for the monitoring of coastal waters and other marine science applications. However, validation of satellite-derived products remains problematic, as simultaneous matchups of in situ data and cloud-free satellite data are costly and difficult to obtain with ship-based measurements. Optical instruments on autonomous platforms can provide many more matchups, typically one per cloud-free image. For moderate resolution ocean colour sensors such as MERIS/ENVISAT (2002-2012) and MODIS/Aqua (2002-present), this means at least one matchup per cloud-free day at mid-latitudes, giving tens of matchups per year and hundreds over the lifetime of these satellites. SmartBuoys are an example of such autonomous buoys, and have been deployed by CEFAS for many years, some for over a decade. These buoys record several parameters multiple times per hour, with only short disruptions between deployments.
We present a dataset of turbidity (T), Photosynthetically Available Radiation (PAR) at different depths, and fluorescence (F) from three buoys in coastal waters, two in the North Sea, and one in the Irish sea, that has been combined with marine reflectance spectra and several standard L2 products from MERIS and MODIS, including multiple processing versions. The merged dataset contains several hundreds of matchups between in situ and satellite. It provides a powerful benchmarking tool for validating satellite products and retrieval algorithms for turbidity and PAR attenuation. The potential of this benchmark dataset is demonstrated by comparison of a few such algorithms.