Most of the main determinants of water quality either consist of, or are controlled by, particles. Previous water quality research has focused on particular particles in isolation or in binary combinations. In this project, we are taking a holistic approach to the characterization of the particle load in water, focusing on the collective properties of the particles rather than individual components. Because the characterization of particles is often time-consuming, applying an informatics-based approach could speed up the evaluation of water quality and the assessment of treatment effectiveness. Further, the breadth of potential changes that could be detected using this multiplex approach may far surpass the abilities of current approaches to monitor threats to water quality. Among the instruments capable of rapidly detecting and manipulating cells is imaging flow cytometry, which distinguishes cell shape and unique fluorescence properties associated with cell types. Sets of images and corresponding data from a 1.5-year time series of samples from the Columbia River were studied and the different particle properties analyzed. Principal Component Analysis (Empirical Orthogonal Function analysis) was applied in order to reduce the number of variables and identify patterns in particle characteristics when compared to environmental data collected from the observation station. The first three principal components were extracted and the dominant characteristics identified: the most prominent variables are particle size, particle color, and fluorescent qualities (transparency and phytoplankton pigments). Further work will relate these top principal components to specific environmental factors that determine water quality.

Thesis (M.Sc. (Technology of Environmental Management))--Mahidol University, 1979.