A solo developer has launched Submarius, a dashboard originally built for diving conditions that now aims to solve a vexing data gap: the lack of a reliable API for water clarity, or "viz" as divers call it. The project, shared on Hacker News, uses a state-augmented ensemble Kalman filter to estimate clarity from satellite imagery, attempting to filter out noise like clouds and shallow reef reflections.
The creator argues that water clarity is among the most consequential factors for dive experiences, yet no standardized method exists to retrieve real-time data globally. Submarius represents an open attempt to crowdsource a solution, with the developer openly asking the community for feedback on improving the underlying model.
Technical details posted on the project's clarity page explain the ensemble approach, which the developer describes as a technique to handle messy satellite data. The project is in early stages; the Hacker News post garnered 7 points and zero comments at time of writing.
The developer is actively seeking collaboration on the algorithm, noting that cloud cover often reads as turbid water, while shallow reefs cause abnormally high readings. No timeline for a public API or commercial product has been announced.
One potential issue: without community validation or peer review, the filter's accuracy remains unproven across diverse global dive sites. The developer's request for alternative methods underscores the novelty of the approach.