Probabilistic Modeling and Evaluation of Surf Zone Injury Occurrence along the Delaware Coast

Beebe Healthcare in Lewes, DE collected along the DE coast surf zone injury (SZI) data for seven summer seasons from 2010 through 2016. Data include, but are not limited to, time of injury, gender, age, and activity. Over 2000 injuries were recorded over the seven year period, including 116 spinal injuries and three fatalities. These injuries are predominantly wave related incidents including wading (41%), bodysurfing (26%), and body-boarding (20%). Despite the large number of injuries, beach associated hazards do not receive the same level of awareness that rip currents receive. Injury population statistics revealed those between the ages of 11 and 15 years old suffered the greatest proportion of injuries (18.8%). Male water users were twice as likely to sustain injury as their female counterparts. Also, non-locals were roughly six times more likely to sustain injury than locals. In 2016, five or more injuries occurred for 18.5% of the days sampled, and no injuries occurred for 31.4% of the sample days. The episodic nature of injury occurrence and population statistics indicate the importance of environmental conditions and human behavior on surf zone injuries. Higher order statistics are necessary to effectively assess SZI cause and likelihood of occurrence on a particular day. A Bayesian network using Netica software (Norsys) was constructed to model SZI and predict changes in injury likelihood on an hourly basis. The network incorporates environmental data collected by weather stations, NDBC buoy #44009, USACE buoy at Bethany Beach, and by researcher personnel on the beach. The Bayesian model includes prior (e.g., historic) information to infer relationships between provided parameters. Sensitivity analysis determined the most influential variables to injury likelihood are population, water temperature, nearshore wave height, beach slope, and the day of the week. Forecasting during the 2017 summer season will test model ability to predict injury likelihood.