Developing a forecasting model for ocean visibility through crowdsourced data

The DiveViz platform being developed through this SBIR will communicate accurate real-time and forecasted ocean visibility reports. It is a prognostic tool that will lead to more coordinated and timely responses to high visibility events for the recreational, commercial, scientific, and military diving communities and improve advanced activity planning for all ocean activities. The primary impediment to visibility forecasting is a total absence of a single, global repository for daily visibility data and, alarming, there is no publicly available historical source of visibility data. DiveViz uses predictive analytics and machine learning to address the numerous influences that must be accounted for in visibility forecasting, including physical factors like currents, wind, waves, storm patterns, near shore rainfall, upwelling strength, and thermohaline circulation, as well as biological factors including algae blooms, nutrient inputs, and human activity, but without any historic ocean visibility data researchers cannot begin this analysis. DiveViz will engage the recreational diving community to crowdsource the first dataset of visibility reports and in collaboration with the Scripps Institute of Oceanography, begin refining the forecasting algorithm. This proposal supports the NOAA’s goal of increased engagement, awareness and understanding of the ocean by facilitating public access, commercial enterprise, scientific expeditions, and military training and operations.