You are here:
Analysis of high-frequency water quality data collected during a cyanoHAB event on an inland, multi-use reservoir
Citation:
Allen, Joel, C. Nietch, AND M. Elovitz. Analysis of high-frequency water quality data collected during a cyanoHAB event on an inland, multi-use reservoir. Presented at U.S. Algal Toxin Conference 2015, Akron, OH, April 28 - 30, 2015.
Impact/Purpose:
This presentation will demonstrate current analysis of a cyanoHAB in an inland multi-use reservoir.
Description:
We describe here an effort to use high frequency data collected from online, continuous monitors coupled with field collected data to describe the temporal relationship between suspected HAB drivers and observed cyanoHABs and cyanotoxin production to provide insight on the necessity and timing of the use of expensive engineering solutions to mitigate risks associated with cyanotoxins.Partnering with Clermont County, Ohio's Drinking Water Utility, two data collection sites were developed, one at the intake structure of the BMW Treatment Plant and another on a buoy located 300 meters southeast of the intake structure. A variety of monitoring equipment were used to measure intake and surface water continuously. Data presented here include in-vivo fluorescence algal abundance data at the division level collected using a BBE Algae Online Analyzer (AOA), and chlorophyll-a and phycocyanin signals using a YSI multiprobe. Regular water quality sampling for nutrients, algal quantification and taxonomic identification was also performed.Time series analysis suggests that the bloom observed in June of 2014 had three phases: a period of relatively low phycocyanin values and diurnal periodicity, a period of moderate phycocyanin values and increasing diurnal amplitude, and finally, a peak period with the highest phycocyanin values and diurnal amplitude. These phases can be categorized in state-space allowing for classification of the bloom status. Transitions from one state-space to another may be useful for understanding potential short-term changes leading to a cyanoHAB event. Results of prediction efforts using exponential smoothing and neural networks will also be discussed.
