A basin-scale wave or tsunami model needs a global grid. A dredging or coastal flooding study needs the shape of every channel and sandbar.
Datamesh has bathymetry at three scales:
๐๐น๐ผ๐ฏ๐ฎ๐น: GEBCO 2026 at ~450 m, released in April by the Nippon FoundationโGEBCO Seabed 2030 Project. Every edition back to 2008 is there as well.
๐ก๐ฎ๐๐ถ๐ผ๐ป๐ฎ๐น: AusBathyTopo 250m 2026, released in August by Geoscience Australia and the Australian Hydrographic Office, alongside the 2009, 2023 and 2024 grids.
๐๐ผ๐ฐ๐ฎ๐น: NOAA NCEI’s CUDEM at 1/9 arc-second (~3 m). It covers the US coast and territories with land, intertidal zone and seabed in one continuous grid, organised into 32 datasources from the 1,036 tiles NCEI publishes.
All of them work the same way. From the Datamesh Python client, the web app or the API, you give a datasource name and a bounding box, and you get back only that area. Going from a global grid to a 3 m grid of the Golden Gate means changing the name and the box.
We timed a few requests: 13 million cells of GEBCO off south-east Australia, 3.4 million of AusBathyTopo over Bass Strait, and 6.3 million of CUDEM at the Golden Gate. Each came back in under 10 seconds.
The notebook behind those numbers, which also uses GEBCO 2026 to time the 2004 tsunami across the Indian Ocean: https://lnkd.in/eTm_v5ZX
- GEBCO Compilation Group (2026). GEBCO 2026 Grid. https://www.gebco.net
- Geoscience Australia & Australian Hydrographic Office (2026). AusBathyTopo 250m (Australia) 2026 Grid. https://lnkd.in/ePRirXsj
- CIRES (2014). Continuously Updated Digital Elevation Model (CUDEM), 1/9 Arc-Second Resolution Bathymetric-Topographic Tiles. NOAA NCEI. https://lnkd.in/e3Q834H4
#Bathymetry #Seabed2030 #GEBCO #GeoscienceAustralia #CUDEM