Operational monitoring of two species of Dinophycae at the origin of Harmful Algal Blooms (HAB)

Click here to access a presentation of HAB in ppt format


To propose an identification of two dinoflagellates from space, Karenia mikimotoi and Lepidodinium chlorophorum. Both species may generate Harmful Algal Bloom. Based on fuzzy methods, the resulting product is an image with values related to the possibility of occurrence of these species.


The methodology developed during this research has been inspired by the work of Moore et al. (2009, 2012). In those publications, the authors use the spectral properties of the ocean waters to classify them from the clearest to the most turbid one. The first step of this method is the characterisation of the optical signature of the studied species. After that, this signature is compared to the spectrum of each pixel on the studied zone to assess the presence of the phytoplankton. The optical signatures have been set up with reflectance data from known zones of bloom for both phytoplankton species, extracted from the daily standard remote-sensing reflectance of MODIS-aqua dataset at six channels (410, 430, 480, 530, 550 and 670 nm). The Mahalanobis distance (Rencher, 1995) is used to compute the proximity between the optical signature of the phytoplankton and the spectral reflectance of the pixel