The Southern Ocean

• The map opens centered on the south pole with the prime meridian at the top, which puts the Weddell Sea and the Antarctic Peninsula to the upper left, Prydz Bay to the right and the Ross Sea below — the arrangement Antarctic sea-ice figures are conventionally drawn in. It is also the plane's own orientation, so this page asks for no rotation at all, where the Arctic asks for 135°E.
• Up, under the map, turns it: the longitude you choose runs up the screen. The page reloads when you let go of the slider, because the projection is fixed when the map is built — it comes back where you were looking, at the zoom you were at, with every layer as you left it, rotated. Your orientation is kept, and a link you copy carries it.
• Sea ice is drawn over the sea-surface temperature rather than instead of it: below 15% concentration the ice layer paints nothing, so the water shows through and the edge of the pack is where the two meet. Around Antarctica that edge moves further in a year than anything else on this map.
• The ice shelves are drawn as shelves, in their own paler tone — the Ross and the Ronne-Filchner are floating ice with ocean beneath them, not land, and a coastline that stopped at the grounding line would put about 850,000 km² of them on the wrong side of the shore.
• Only the vector coastline is offered as a basemap here. The tiled bathymetry and street maps are pictures of Web Mercator and cannot be redrawn on this plane, so they are not in the list.
• Most platform layers are empty this far south: the hurricane, buoy, tide-gauge and tropical-array networks all sit in other oceans. Argo floats is the one worth switching on — the Southern Ocean array is the densest thing reporting here.
• Your layers, colors and view are kept for this page on its own — separately from the Arctic and from the global map, so the three do not overwrite each other.

Sea ice concentration and thickness, currents, temperature and salinity come from the US Navy ESPC-D-V02 forecast; NOAA PSL OISST v2.1 supplies the observed temperature, its anomaly against the 1991–2020 normal, and its own ice concentration. ECMWF IFS open data supplies 10 m wind, 2 m air temperature and total cloud cover, and ECWAM the wave height. The land, the ice shelves and the glaciated areas are Natural Earth; GEBCO Compilation Group supplies the depth contours. The platforms in the switcher are the Argo array via the Ifremer Argo GDAC, ocean gliders from the US IOOS Glider DAC and three other - [Rolling Deck to Repository](https://www.rvdata.us/) — the US academic research fleet registry, which names the vessels the GTS reports as bare call signs - [SAMOS (Florida State University)](https://samos.coaps.fsu.edu/) — one-minute underway positions from most of the academic fleet, which is what makes a vessel track a course rather than an 11 km staircase; about half a day behind, so the last hours of each track are still the GTS - [Schmidt Ocean Institute](https://schmidtocean.org/) — RV Falkor (too) relayed from the ship herself, at one minute and full precision - [SondeHub (CC BY-SA 2.0)](https://sondehub.org/) — radiosondes currently aloft, from a volunteer network of receivers rather than the WMO, so it sees about 40% of the world's twice-daily launches and almost none outside Europe, North America, Japan and Australia national data centers, NOAA uncrewed surface vehicles from NOAA PMEL and NOAA OSMC, the National Data Buoy Center's moored buoys, NOAA CO-OPS tide gauges, and the GTMBA tropical moored arrays — of which only Argo has much presence in the Southern Ocean, the rest being northern or tropical networks listed because the layers are one switch away. The National Hurricane Center's storm forecasts and wind speed probabilities are in the switcher for the same reason, and cover no water shown here. The Joint Typhoon Warning Center's do reach these longitudes: it warns on southern-hemisphere cyclones, though they stay well north of the ice. The Japan Meteorological Agency's typhoons are northern and are listed for the same reason as the rest. Everything but the seafloor is refetched hourly.

Five of the map's usual sources cannot appear on this page, and the projection is the reason: a raster tile is a picture of Web Mercator and cannot be redrawn on a polar plane. So the tiled basemaps — Esri Ocean, OpenStreetMap and GEBCO's own bathymetry rendering — are not offered here, and neither are the EMODnet Bathymetry shoreline, the Marine Regions (VLIZ) maritime boundaries, nor NOAA/NWS NEXRAD radar. The satellite cloud imagery, which comes from NASA GIBS / NOAA GOES, is absent for a different reason: a geostationary satellite sits over the equator and cannot see a pole, so GIBS publishes no GOES layer in either polar projection at all. GEBCO's depth contours are unaffected: they are geometry rather than pictures, so they reproject like the coastline does.