A globally verified coastal glare estimation tool

Abstract

The reflection of light off a rippled water surface often produces glare for an observer at the beach. For applications such as coastal video monitoring, surf lifesaving and coastal research it is useful to know when glare will be present and how strong it will be. Open-source software was developed to predict the proportion and intensity of glare from the water surface visible in a field of view at any time and day of the year. The algorithm for estimating glare links together physics, geometry and the software library “Astropy”, which provides the positioning of the Sun relative to the Earth. The algorithm was verified with still images taken at different latitudes and bearings around the world with an average glare region overlap accuracy of 40% and average glare region size similarity accuracy of 60%. Glare graphs produced by the software were verified with yearlong camera footage with an average time period overlap accuracy of 60%. The algorithm was particularly accurate for times of strongest glare. This new tool can be used for automating glare avoidance for continuous coastal monitoring systems, informing lifeguards when glare will be obstructive to surveillance and optimising the timing and positioning of coastal research experiments requiring video capture. • Open-source software estimates when coastal observations are affected by glare. • Verified with phone images globally and for a full year in Australia. • Can be used to improve lifeguard surveillance, coastal monitoring systems and field experiments.