An architecture for providing forecasts of GNSS obscuration and multipath

Disclosed is a method of providing dilution of precision (DOP) forecasts for GNSS navigation and optionally degree of confidence, for routing of vehicles or alerting humans in vehicles: accessing a 3D map of an area including structure solids and generating cuboids in spaces not contained in the structure solids, and iteratively over time increments, calculating GNSS satellites visible from the cuboids using the 3D map and, using at least the calculated visibility, determining a DOP forecast for GNSS signals observable in the cuboids at the time increments. The disclosed method also includes compressing the calculated DOP forecast spatially and temporally, and distributing the compressed DOP forecast via a content delivery network (CDN), responsive to queries from requestors to an API of the CDN, whereby the requestors' systems can take into account the DOP forecast for routing the vehicles or alerting the humans in the vehicles to a predicted navigation impairment. The area preferably being an urban environment with the solid structures being high rise buildings.