FogWindow

FogWindow predicts fog formation and dissipation at street level for the next 0–12 hours, focusing on airports, ports, highways, and industrial sites where visibility drives real costs. It blends public NWP model output with local topography, land/water masks, and live observations (METARs, road weather stations, nearby IoT sensors) to produce hyperlocal visibility probability, expected onset/clear times, and confidence bands. Users can set geofenced alerts for “visibility < 1 km” or “LVP likely” and get a concise explanation of the drivers (dewpoint spread, wind shift, inversion risk). The product is not a general weather app; it’s a decision tool for operations that lose money when fog surprises them. MVP can start with a limited region and a small number of high-value locations, then expand as data coverage and verification improve.