In-Situ Classification of Bloom-Forming Phytoplankton by Imaging Multivariate Optical Computing (IMOC)

点击翻译按钮获取中文摘要

The PIs propose to develop a simple system for the in situ identification and detection of three major phytoplankton bloom species in real time. The system will be based on simultaneous analysis of their size and their fluorescence excitation spectral characteristics using a new method called imaging multivariate optical computing (IMOC). This method will produce an optical "bar code" to discriminate between phytoplankton taxa. IMOC works by imaging through an optical filter uniquely designed for spectral analysis for a specific application such as phytoplankton discrimination. It requires that the investigators develop a database of phytoplankton fluorescence excitation spectra under varying growth conditions and employ canonical discriminant analysis to identify one or more discriminant functions to identify phytoplankton taxa. The PIs will use IMOC to combine instant spectroscopic analysis with a streak-camera-type microscopic imaging system to provide real-time or near-real-time classification. The method is ideal for field deployable instrumentation, making it an ideal addition to ocean observing platforms.Broader ImpactsIdentification of marine phytoplankton is a critical area in biological oceanography. Successful implementation of this research could go far towards lowering the ambiguities in the identification of marine phytoplankton, which could easily be automated for use from buoys and other remote platforms. An educational outreach program will be pursued through an existing program ("Rising Tide") that links high school teachers and undergraduate interns in summer research projects. Interns from South Carolina HBCU institutions would be targeted for this experience. Commercial development/private sector partnerships with academia would likely result from a successful project.