A new approach to discriminate dinoflagellate from diatom blooms from space in the East China Sea

Dinoflagellate and diatom blooms often occur in the East China Sea (ECS) during spring and summer. Some of the dinoflagellate blooms are toxic, resulting in widespread economic damage. In order to mitigate the negative impacts, remote-sensing methods that can effectively and accurately discriminate between bloom types are demanded for early warning and continuous monitoring of bloom events at large scales. An in situ bio-optical data set collected from diatom and dinoflagellate blooming waters indicates that the two types of blooms exhibited distinctive differences in the shapes and magnitudes of remote-sensing reflectance (R-rs). The ratio of in situ measured R-rs spectral slopes at two spectral ranges (443-488 and 531-555 nm, bands available with the moderate resolution imaging spectrometer (MODIS) sensor), abbreviated as BI (representing bloom index), was found effective in differentiating dinoflagellates from diatoms. Reflectance model simulations, which were carried out using in situ and algal culture data as input, provided consistent results. A classification approach for separating dinoflagellate from diatom blooms in the ECS was then developed: When fluorescence line height (FLH) is doubled over the background level and total absorption coefficient at 443 nm >= 0.5 m(-1), if 0.0 < BI <= 0.3, it suggests a dinoflagellate bloom; if 0.3 < BI <= 1.0, it suggests a diatom bloom. Finally, the approach was applied to MODIS measurements over the ECS, and a series of diatom and dinoflagellate bloom events during April-June 2005 and 2011 were successfully identified, suggesting that the proposed approach is generally valid for the ECS.