← Back to Prototype 2

Research prototype · public data · not validated · not for navigation.

Anomaly scan: Prototype 2's regions (exploratory)

Run 12 October 2026 over 21 regions and 149,754 hexagons, from the build's own tables. Exploratory: the thresholds were set before the run, but the scan was not written down in advance the way the trackline test was. It reports what the data show, never what an anomaly is.

Unidentified objects (USOs): none of the datasets read so far records any. Adding sourced public records is planned (see the end).

Each profile already says "Sources disagree here" where GEBCO and GMRT differ by a median of more than 100 m, or where GEBCO calls the depth predicted while GMRT holds ship data. This scan adds a test relative to depth (so shallow water counts) and asks, for each one, whether anything in the data explains it.

What it found

Region Hexagons Grids disagree Explained Not explained Predicted-only rises
New Jersey shelf 2,470 0 0 0 0
Mid-Atlantic Ridge 25,297 0 0 0 0
Gulf of Maine 7,558 8 3 5 0
Virginia to Hatteras 6,258 3 2 1 0
Blake Plateau 10,918 43 33 10 1
Florida Keys 4,567 36 28 8 0
Gulf of Mexico 11,845 0 0 0 0
Puerto Rico 5,121 24 24 0 0
Southern California 2,075 74 74 0 0
Monterey 1,654 28 28 0 0
Pacific Northwest 12,348 25 24 1 0
Hawaii 7,500 16 16 0 0
Long Island 398 0 0 0 0
West Florida 1,141 0 0 0 0
Texas shelf 2,722 0 0 0 0
Northern California 2,865 11 7 4 0
Gulf of Alaska 9,873 2 1 1 0
Southeast Alaska 4,708 7 7 0 0
Aleutians 6,647 4 4 0 1
Marianas 17,840 9 8 1 9
American Samoa 5,949 13 13 0 1

Not explained by the data read so far

Largest predicted-only rises

What explains most of these, and what comes next

Thresholds: disagreement 50 m and 10% with at least 20 paired cells; steep when the depth range is 3x the difference; mostly predicted at 50%; a rise 500 m and 25% above the six neighbours' median, with all six at sea and at least 200 m of water (added after the first run, when a 44 m "rise" off Hawaii turned out to be the coastline). Code: aethon_p2/anomaly_scan.py.