ARPEGE 0.25° Explorer

This is the same Météo-France run as ARPEGE 0.1°, drawn over the whole planet. ARPEGE is one global forecast on a stretched, tilted grid — sharpest over France at about 5 km, falling to roughly 24 km at the antipode — and Météo-France publishes several regular interpolations of it. The European page draws the 0.1° one; this draws the 0.25° global one. They are not two models, and the site does not pretend they are.

That was measured rather than assumed. Over a box around France at +24 of the 29 August 00Z run, the two grids agree to 0.021 K in mean 2-metre temperature and 0.05 hPa in mean sea-level pressure, and they put the lowest pressure in the box on the same pascal. Two forecasts that merely agreed would not do that. So when you swipe this against the European domain in Compare, what you are looking at is resolution — what 0.1° resolves that 0.25° smooths away, which over France is mountain valleys, coastlines and convective detail. Swiping against ICON-EU is the other kind of comparison entirely: two national services that genuinely disagree.

What this model has that nothing else on this site does is the rest of the world. Both other feeds stop at a boundary — ARPEGE-EUROPE at 32°W and 42°E, ICON-EU at 23.5°W and 62.5°E — and a French forecast four days out is being decided outside both of them. The blocking high over Greenland, the hurricane recurving into the Atlantic storm track, the Rossby-wave train coming across the Pacific: none of those fit in a European box, and all of them are on this map.

And it is the only model here with a pole in it, which is why the polar page exists at all now. Web Mercator is infinite at 85.05°, so no map on this page can contain a pole at any size — the two caps get their own stereographic render instead. The ARCTIC and ANTARCTIC buttons beside the place picker carry whichever field you are looking at across to them.

The axis is hourly to +48 and three-hourly to +102 — 67 steps, against the European domain's 103. That is Météo-France's own publication axis for this grid rather than a choice made here: there is no +49 or +50 in the feed at all. The compensation is that almost nothing thins within it. On the European domain the fields aloft drop to three-hourly after +48 while the surface stays hourly, so its scrubber changes length depending on which button you press; here the 500 hPa height has a frame at every one of the 67 steps, exactly as the 2-metre temperature does.

Two-metre maximum and minimum temperature are here and are on no other model on this site, because ARPEGE-EUROPE publishes no period extremum at all. They are the warmest and coldest the air gets over the model's own three-hour window, so they appear at every third hour and their scrubber is 34 steps rather than 67. Gust is a per-step maximum on a different window again — one hour, even out in the three-hourly tail — and visibility is the per-step minimum and stops at +48. Each field's timeline is built from the frames that field actually has, so a shorter scrubber is the feed being reported honestly rather than a hole in this site.

Colours and band breaks are identical to ARPEGE 0.1°'s and ICON-EU's, on purpose and more strictly here than anywhere: if a band edge differed between this page and the European one, a reader comparing them would be reading a palette and believing they were reading a grid. Precipitation and snowfall are published as totals since the start of the run and are differenced back to per-step here, so every frame states the interval it covers — one hour to +48, three hours after it.

Data © Météo-France, from their open-data service under the Licence Ouverte / Etalab. Forecast France is not affiliated with Météo-France, and is not a warning service — for warnings see Vigilance Météo-France.