Evidence Coverage

Where Atlas has evidence, and where it does not

Atlas reasons over whatever evidence its connectors can actually retrieve for a location. Coverage is uneven by design and by honesty: this page shows exactly where that evidence is strong, partial, or not yet there.

Coverage map

Drag to pan, scroll or pinch to zoom, double-click to zoom in. Shaded rectangles are the connectors' own coverage bounding boxes; unshaded areas fall back to global sources only.

Strong
Partial
Insufficient
Source reliability grade (fixed per source, Admiralty-style): A primary national/federal survey or agency data · B curated academic catalogue or regional agency product, peer-reviewed or licensed provenance · C index/inventory listing or derived signal, one step removed from primary measurement · D statistical or modeled proxy surface predicting a property at unmeasured locations · grades run A (best) to F (worst); no source on this page is graded below D today.

Why we publish this

Atlas is only as good as the evidence it can retrieve. The binding constraint on what Atlas can responsibly say about a site is evidence coverage, not reasoning: the reasoning step is the same everywhere, but the sources feeding it are not.

We would rather show exactly where we are strong, and say plainly where we are not, than imply uniform global coverage we do not have. A screening tool that overstates its own coverage is more dangerous than one that is candid about its gaps, because the reader cannot tell the difference from the output alone.

This page is the coverage map behind that judgment. It lists, region by region, which sources Atlas actually queries, how reliable each source is, and what is known to be missing. The shaded areas above are the actual geographic bounding boxes the connector code checks before it will even attempt a query, not illustrative markers: a boundary here is drawn from the same numbers the code runs on.

Regions

Loading coverage data...