Abstract
Nearly annually, blooms of the dinoflagellate Karenia brevis form along the southwest Florida coast leading to a variety of negative impacts, including respiratory irritation (RI) in humans. To limit these impacts, NOAA's National Centers for Coastal Ocean Science (NCCOS) developed a RI model to provide beach-goers with a category-based estimate of RI risk at individual beaches along Florida's Gulf and Atlantic coasts. The RI model is based on: (a) K. brevis cell counts collected at individual beaches; (b) high resolution wind direction and speed forecasts and observations; and (c) point-based beach shoreline orientation used to designate onshore and offshore winds. To test the model logic, an analysis of modeled RI was compared to same-day RI reports, based on the frequency of coughs at individual beaches from the Beach Conditions Reporting System (BCRS). Overall, the model proved to be 88% accurate when K. brevis was present along the southwest Florida coastline from 2006 to 2022. In addition, validation efforts confirmed model assumptions, including: (a) reports of higher RI correlate with higher K. brevis cell abundances; and (b) when cells are present, onshore winds lead to a higher risk of RI. However, individual model categories ("low," "moderate") were less robust. Furthermore, BCRS was not a direct measure of toxic aerosol presence, so some coughing (modeled false negatives) may result from other environmental factors. Together, results suggest the RI model accurately predicts "very low" and "high" risk, but that additional research is needed to better capture environmental conditions when RI is "low" or "moderate."