Machine Learning Forecasting of the Ecological and Economic Impacts of Red Tides in the Southwest Florida Coast

Red tides, a form of Harmful Algal Bloom (HAB), pose a severe threat to Florida’s wildlife, fisheries, and coastal communities. These blooms release harmful brevetoxins, which kill marine life and devastate fisheries, diminishing both yields and revenue. Approximately 16 million Florida residents live in counties bordering the coastline and are directly or indirectly affected by red tides through health risks, reduced quality of life, or employment instability. For example, the revenue of fishermen decreased 62.8% during red tide seasons and fisheries have experienced a 13.7% decrease in production for the past decade, with a 26.4% average increase for price per pound. With rising global temperatures, red tides are expected to become more frequent and severe, further damaging Florida’s marine ecosystem and local economy. This paper leveraged advanced predictive modeling to analyze red tide patterns and assess its risks to inform future policy making. Using an LSTM model, this study predicts that Karenia brevis cell counts will increase by 4.5% annually, leading to substantial economic losses to the fishing industry. Using a SARIMAX model, by 2030, it is estimated that (1) the total pounds lost for all fish will be 71,318,313 pounds and (2) the total fish stock value to decrease by $282, 790, 208. After running a Monte Carlo simulation, the estimated value of fish stock lost by 2030 will be $2,545,123,304. This paper also used a burden ratio model to quantify how red tide-related impacts are distributed across income groups based on differences in exposure, vulnerability, and adaptive capacity. Using this model, it was found that by 2030, lower-income communities—who face higher job disruption, limited healthcare access, and greater dependence on affordable seafood—will bear 60% of the red tide-related economic losses, highlighting the need for targeted interventions.