Towards a Predictive Model for Fecal Bacteria Contamination at Rincon Public Beach, Puerto Rico

Beach water quality has become a frequent problem in Puerto Rico. Just within this year, 85% of the beaches sampled by the Environmental Quality Board exceeded, at least once, the allowable limits for fecal indicator bacteria (FIB). The current operational strategy to assess swimming conditions consists of detecting FIB from in situ collected samples. This process requires at least 18 hours of incubation, which by the time a result is produced may not be representative of the actual state of the site. Furthermore, due to the laborious nature of this process, the common sampling frequency is scarcely once a week, which does not detect abrupt events with periods of hours to days. This project focuses on developing a statistical model for the presence of FIB at Rincon Public Beach using existing site-specific meteorological, oceanographic, and microbial data as input for EPA’s Virtual Beach software package. Preliminary results show 3-hr rain accumulation consistently correlating with the log₁₀ of bacteria colony forming units (CFUs), suggesting that surface water runoff may play a crucial role on beach water quality. A multiple linear regression model using rain accumulation, water levels, wind, and wave characteristics is currently under development. Once validated, this model will be able to provide daily forecasts of beach water quality based on CariCOOS’ operational SWAN and WRF models, among others.