Detection of Karenia brevis blooms on the west Florida shelf using in situ backscattering and fluorescence data

Using shipboard data collected from the central west Florida shelf (WFS) between 2000 and 2001, an optical classification algorithm was developed to differentiate toxic Karenia brevis blooms (>10⁴ cells l^?1) from other waters (including non-blooms and blooms of other phytoplankton species). The identification of K. brevis blooms is based on two criteria: (1) chlorophyll a concentration ≥1.5 mg m^?3 and (2) chlorophyll-specific particulate backscattering at 550 nm ≤ 0.0045 m² mg^?1. The classification criteria yielded an overall accuracy of 99% in identifying both K. brevis blooms and other waters from 194 cruise stations. The algorithm was validated using an independent dataset collected from both the central and south WFS between 2005 and 2006. After excluding data from estuarine and post-hurricane turbid waters, an overall accuracy of 94% was achieved with 86% of all K. brevis bloom data points identified successfully. Satisfactory algorithm performance (88% overall accuracy) was also achieved when using underway chlorophyll fluorescence and backscattering data collected during a repeated alongshore transect between Tampa Bay and Florida Bay in 2005 and 2006. These results suggest that it may be possible to use presently available, commercial optical backscattering instrumentation on autonomous platforms (e.g. moorings, gliders, and AUVs) for rapid and timely detection and monitoring of K. brevis blooms on the WFS.