Edward Quilty

Department of Forest Sciences
3041-2424 Main Mall, UBC
Vancouver, B.C. V6T 1Z4
Email: equilty@interchange.ubc.ca

Position: Ph.D. Candidate

Study Area: Analysis and interpretation of continuous water quality data systems

For more information on Ed’s work see Aquatic Informatics Inc. home page


Edward Quilty has worked in the field of water quality since 1992, and is currently pursuing a PhD degree. Ed has worked for British Columbia’s Environment Ministry, for the British Columbia Conservation Foundation, for QA Environmental Consulting, a private firm he founded in 1998, and for Aquatic Informatics Inc., a company he co-founded in 2003. His academic interests include environmental informatics and analysis and interpretation of continuous (high frequency) water quality data.

Ed’s thesis work is focused on testing and refining methods for correction, analysis, and interpretation of continuous water quality data. Traditionally, water quality monitoring programs have focused on intermittent and low frequency grab sampling. In this approach, data assessment has generally focused on “mean water quality.” As the technology available for water quality sampling changes there is an increasing emphasis on characterizing how stream ecosystems function in order to manage for long-term sustainability and ecological change. This ecological approach focuses on relationships, dynamics, and systems rather than on amounts. High frequency automated electronic monitoring is ideally suited to gathering the data necessary to characterize how streams and rivers function. However, with high frequency data collection often comes the ‘data-rich but information-poor syndrome.’ Data managers facing tens of thousands of data points may be tempted to revert back to the classical assessment approach. The alternative is to develop methods that allow us to capture and understand the system details, including seasonal and diurnal responses and significant transient events. Ed is employing various signal processing methods, such as wavelet transforms, artificial neural networks, ARMA models, and fuzzy logic clustering, to examine these details.