EPA researchers are developing predictive models to relate nutrient loads, land use, land cover, socioeconomic factors, and climate to the frequency, location, and severity of cyanobacterial blooms in lakes of the United States. Improving scientific approaches will help decision makers, watershed managers, and community leaders make informed decisions.

Predictive models for water quality and indices of cyanobacteria

EPA's computational ecologists have used machine learning methods to predict chlorophyll a concentrations in lakes. These methods rely on large data sets, emphasize open science practices, and are well suited to extract information about the potential drivers of water quality. Because chlorophyll a is the type of chlorophyll found in cyanobacteria, it is widely used as a measure of cyanobacteria concentrations to determine water quality.

Additional Resources

Software and Analyses: 

• LakeTrophicModelling [GitHub] Exit
• Lakemorpho [GitHub] Exit

Publications:

• Ecology of Harmful Algal Blooms Publications
• Modeling Lake Trophic State: A Data Mining Approach
• Estimating Summer Nutrient Concentrations in Northeastern Lakes from SPARROW Load Predictions and Modeled Lake Depth and Volume

Blog Posts:

• Open Science and Cyanobacterial Research at EPA
• Modeling Cyanobacteria Ecology to Keep Harmful Algal Blooms at Bay
• SPARROWs, Lakes, and Nutrients?