by Dash, Padmanava

Abstract (Summary)
Robert K. Vincent, Advisor Toxin-producing blooms of cyanobacteria in most freshwater systems have posed a serious global threat throughout the world. Vincent et al. (2004) have developed a set of algorithms that employ LANDSAT 7 ETM+ data and LANDSAT 5 TM data to map cyanobacterial blooms in fresh water lakes from space by quantitatively mapping phycocyanin, the pigment most uniquely associated with cyanobacterial blooms. However, the LANDSAT TM sensor has a 16-day revisit cycle and a spatial resolution of 28.5 m, whereas the SeaWiFS sensor has a daily revisit cycle and a coarse spatial resolution of 1.1 km for LAC (local area coverage) and 4.5 km for GAC (global area coverage). Hence, the SeaWiFS sensor can be applied with higher temporal frequency and can be used to document large-scale changes in the region, due to its larger synoptic coverage. In this research an algorithm has been developed for detecting incipient cyanobacterial blooms with SeaWiFS data that is also based on phycocyanin pigment. An empirical method for atmospheric correction, ratio normalization, has been applied to SeaWiFS data after additive offset of the sensor and atmospheric haze have been corrected by dark object subtraction, which is primarily aimed at making two input images similar with respect to radiometric qualities. It has been observed that ratio normalization becomes more important for coarse resolution sensors such as SeaWiFS (due to their synoptic coverage) than that of sensors like LANDSAT TM. The results were compared to the LANDSAT TM results and observed that SeaWiFS model predicts phycocyanin concentration about as well as the LANDSAT TM spectral ratio model. To ensure that we are mapping phycocyanin and not chlorophyll a (Chl a), a model for Chl a has been developed from SeaWiFS data. The Chl a model was then applied to the same dates of SeaWiFS data to which the phycocyanin model was applied and the two models results were compared, which brought to view that no correlation exists between phycocyanin and Chl a concentration. This cyanobacterial bloom algorithm with SeaWiFS Data can be used to investigate inter and intra-annual variability of cyanobacterial blooms in the entire Lake Erie. The algorithm could serve as a tool for early stage monitoring of cyanobacterial blooms, thereby assisting future assessment of water quality and the aquatic ecosystem health of large fresh water lakes around the globe.
Bibliographical Information:


School:Bowling Green State University

School Location:USA - Ohio

Source Type:Master's Thesis

Keywords:phycocyanin cyanobacteria seawifsr


Date of Publication:01/01/2005

© 2009 All Rights Reserved.