Towards the Early Detection of Gymnodinium catenatum Algal Blooms in the Northern Gulf of California

Abstract

The annual occurrence of harmful algal blooms (HABs) of the dinoflagellate Gymnodinium catenatum in the northern Gulf of California (NGC) during winter and spring has negative ecological, economic, and social impacts on the local coastal population. G. catenatum produces paralytic shellfish toxins, and a robust monitoring program of the species is necessary to sustain mitigation actions against their detrimental effects. Here, we applied the maximum-likelihood classification (MLC) method to classify satellite images from MODIS and Sentinel-3 to evaluate their effectiveness to detect G. catenatum. Different classes associated with the presence of the species were developed from data of two HABs that occurred in 2015 and 2017. Two classes derived from Sentinel-3 data from the 2017 HAB allowed the detection of this species. These Sentinel-3 classes adequately represented the temporal and geographical distribution of G. catenatum in the region and the no-bloom condition during the summer. The concordance between the detection of the Sentinel-3 classes on the west coast of the NGC and the recorded presence of G. catenatum (75% of concordance) in the area indicates that the MLC method could be applied for early detection of the species in the NGC, using Sentinel-3 full resolution images.