Skill assessment for an operational algal bloom forecast system.

Area of Science:

• Marine Biology
• Oceanography
• Environmental Science

Background:

• Harmful algal blooms (HABs), caused by Karenia brevis, impact Florida's coast, causing shellfish toxicity and respiratory irritation.
• Operational forecast systems are vital for predicting HAB extent and public health impacts.

Purpose of the Study:

• To analyze the forecasting skill of an operational harmful algal bloom (HAB) system in southwest Florida.
• To evaluate the accuracy of HAB forecasts, including identification, intensification, transport, extent, and impact.

Main Methods:

• Utilized satellite imagery, wind predictions, and a rule-based model for HAB forecasting.
• Validated forecasts against water sampling for bloom extent and transport.
• Assessed respiratory irritation forecasts using lifeguard observations and county-level data.

Main Results:

• Forecast skill could not be resolved at scales finer than 30 km due to resolution limitations.
• Respiratory event forecasts achieved 70% accuracy at the beach level but had high false positive rates (80%).
• Reducing validation resolution to the county level improved accuracy to 78% with 22% false positives.

Conclusions:

• Systematic sampling is essential for accurate HAB impact assessment, even with qualitative data.
• Matching forecast and validation data resolutions is critical for evaluating forecast capabilities.
• The study highlights the need for improved resolution in both HAB forecasting and validation methods.