Harmful algal blooms (HABs) are increasingly affecting lakes and reservoirs across the United States, posing risks to public health, aquatic life, and local economies. These blooms, which often appear as a green film on water surfaces, have become more common and severe due to rising temperatures and higher nutrient levels in water bodies. Exposure to HABs can result in skin irritation, respiratory problems, or more serious illnesses if toxins are ingested.
The U.S. Environmental Protection Agency (EPA) currently provides national forecasts for harmful algal blooms using satellite data and standardized models. These forecasts offer valuable broad-scale awareness but may not reflect the specific conditions that influence bloom development at the local level, such as rainfall patterns, lake structure, or wind dynamics.
Custom state-level forecasting models have been proposed as a solution to address these limitations. Such models use a combination of satellite imagery, sensor data from within water bodies, and localized environmental variables to provide more precise predictions tailored to individual lakes or systems of lakes. This approach enables state and local authorities to better anticipate and manage bloom events rather than reacting after they occur.
According to proponents of custom modeling: “National HAB forecasts will always play a role in understanding the broader picture. But the greatest impact comes when states take that foundation and make it their own.By tailoring HAB forecasts to local conditions, state agencies can protect their water more effectively and ensure communities have the clean, safe water they depend on.”
One example is a partnership with the Florida Department of Environmental Protection (DEP), where a custom model was developed that runs daily instead of weekly. This allowed teams to be more proactive in field testing by splitting larger lakes into smaller zones for targeted sampling—especially near public beaches or drinking water intakes—and by modeling bloom movement within lakes. Unlike EPA’s binary approach of predicting either “bloom” or “no bloom,” Florida’s model predicts continuous variables representing varying bloom conditions so managers can tailor responses accordingly.
The benefits cited for custom predictive modeling include earlier public health advisories for recreation managers; additional time for water treatment facilities to adjust processes; improved coordination among agencies; more efficient allocation of resources; identification of long-term problem areas; and support for developing proactive mitigation strategies.
John Park is scheduled to present on operational forecasting of harmful algal blooms in Florida lakes using a two-stage Bayesian model at the American Geophysical Union’s 2025 Annual Meeting on December 17th.
As climate patterns shift and harmful algal blooms become more frequent, experts emphasize that investing in localized forecasting helps states move from crisis response toward proactive management—protecting public health while conserving resources.
State agencies interested in developing their own HAB forecasting capabilities are encouraged to reach out for guidance on starting pilot projects tailored to their needs.
