Modelling the Coastal Fecal Bacteria Contamination in the Eastern Mediterranean Region: Lebanon as a Case Study

Coastal water quality preservation is very important for the well-being of both the local community and the marine environment. Fecal Bacteria (FC) contamination of coastal water is a serious environmental threat. Therefore, the analysis of FC is used to assess coastal waters for potential contamination of biota (e.g., shellfish) and threats to public health and is considered a very relevant pollution-indicating method for coastal waters. For that reason, a data-driven model was created based on several water quality parameters monitored throughout the Lebanese coastline. The created Artificial Neural Network (ANN) managed to capture well the tendency of the FC dynamics (R=0.86 between the real and predicted data), while the sensitivity analysis revealed that the salinity is the most influential parameter on the FC parameter. Additionally, the water samples were categorized into safe and non-safe, and the ANN managed to classify the water quality data samples with high accuracy into these two categories (F1 score=94.24%). Therefore, the created ANN can act as a complementary management tool regarding coastal/bathing water safety indication.