Phenology of phytoplankton blooms in the East Sea (Sea of Japan): remote sensing analysis

Abstract

The phytoplankton seasonality in the East Sea controls the recruitment success of higher trophic levels and drives the carbon export production. To quantify the variability of this seasonality, we used the long-term time series of chlorophyll-a concentration derived from satellite ocean color data (1998-2014). A shifted Gaussian function was used to estimate bloom initiation, peak timing, duration, termination and peak magnitude taking into account uncertainty in bloom detection. In the East Sea, the bloom initiation was inversely related to bloom duration in both spring and autumn, thus early blooms tended to last longer. This trend was more predominant in spring than autumn. Furthermore, the phytoplankton seasonality showed spatially different patterns. The variability of seasonality indicators in the southern area is much higher than that in the northern area. Unlike the southwestern area of the East Sea, in the northern area, the bloom in spring was advanced and lasted longer, whereas the bloom initiation in autumn was delayed and the bloom duration was shortened. Among the physical factors, the average wind stress during February and March is a stronger predictor for peak timing in spring. Changes in bloom termination in autumn also presented an obvious response to physical forcing, especially the date that the decreasing rate of photosynthetically available radiation (PAR) was the maximum. We discuss the implicll-a concentration derived from satellite ocean color data (1998-2014). A shifted Gaussian function was used to estimate bloom initiation, peak timing, duration, termination and peak magnitude taking into account uncertainty in bloom detection. In the East Sea, the bloom initiation was inversely related to bloom duration in both spring and autumn, thus early blooms tended to last longer. This trend was more predominant in spring than autumn. Furthermore, the phytoplankton seasonality showed spatially different patterns. The variability of seasonality indicators in the southern area is much higher than that in the northern area. Unlike the southwestern area of the East Sea, in the northern area, the bloom in spring was advanced and lasted longer, whereas the bloom initiation in autumn was delayed and the bloom duration was shortened. Among the physical factors, the average wind stress during February and March is a stronger predictor for peak timing in spring. Changes in bloom termination in autumn also presented an obvious response to physical forcing, especially the date that the decreasing rate of photosynthetically available radiation (PAR) was the maximum. We discuss the implic