Date of Award


Level of Access Assigned by Author

Campus-Only Thesis

Degree Name

Master of Science (MS)


Electrical and Computer Engineering


Habtom Ressom

Second Committee Member

Mohamad T. Musavi

Third Committee Member

Bruce Segee


This thesis presents the use of computational intelligence-based techniques for retrieving two oceanic parameters, namely, vertical light attenuation coefficient, and chlorophyll a concentration and a physiological parameter, photosynthetic efficiency. Photosynthetic efficiency is a measure of plant stress that can be used to monitor the health of marine plants such as seagrasses. In this thesis, neural network-based models are developed to estimate photosynthetic efficiency from spectral reflectance measurements. Light attenuation coefficient is an important parameter that indicates the light attenuated by the water column and has been used as an indicator of seagrass health. This thesis presents neural network-based models for estimation of light attenuation coefficient fiom water quality parameters in the hdian River Lagoon, FL. The neural network models developed for estimation of photosynthetic efficiency performed better compared to the other regression models such as simple linear regression models, model trees, and support vector regression models. A clustering-based technique is proposed in this thesis for estimation of chlorophyll a fiom spectral reflectance measurements. The proposed approach performed better compared to the global models in retrieving chlorophyll a concentration from spectral reflectances measured at different locations that involve various water types. The performance of the computational intelligence-based techniques developed in this thesis proves the ability of these techniques to capture complex and non-linear relationships. Hence, they offer great promise in retrieving relevant but difficult-to-measure oceanic and physiological parameters from other related parameters that are relatively easy-to-measure or that could be made available more frequently.