Phytoplankton bloom identification in Arizona source waters through a cloud based modeling effort

The Salt River Reservoir System is composed of four lakes (Saguaro Lake, Canyon Lake, Roosevelt Lake, and Apache Lake) that provide drinking water storage along with hydropower to the Phoenix metropolitan area. Additionally, these lakes are used for fishing and other recreational water activities. Blooms of the potentially harmful Golden Alga Prymnesium parvum and the filamentous cyanobacteria Cylindrospermopsis occur typically in spring and summer. This study adds to a continuous monitoring effort by measuring phytoplankton abundance and extracted Chlorophyll a (Chl a) concentration in the surface of  Saguaro Lake’s from 2019 to 2022. Chl a concentrations were also obtained by the Regional Drinking Water Quality Monitoring Program at ASU. The program provides water quality data sets through the Central Arizona Phoenix Long Term Environmental Research online data repository. Chl-a measurements will be compared to Landsat 8 and DoveR images to identify dates when blooms are visible from satellite overpasses. Google earth engine will be used to train machine learning models to identify blooms throughout the reservoir system based on pixel classification. Results from machine learning models such as Support Vector Machine regression and Random Forests will be compared to select the most suitable predictive model.  The purpose of this study is to aid resource managers in identifying blooms of potentially noxious phytoplankton without the associated costs and effort required by in situ sampling.

Presentation Preference: Either

Primary Presenter: Ashley Foster, Arizona State University (ashleynicolefoster16@gmail.com)

Authors:

Susanne Neuer, Arizona State University  (Susanne.Neuer@asu.edu)

Peter Fox, Arizona State University  (Peter.Fox@asu.edu)