Skip to main content
GuidesECMWF vs the Met Office model: how they compare

ECMWF vs the Met Office model: how they compare

They are not really rivals - ECMWF is the world-class medium-range global model, the Met Office adds high-resolution UK detail, and forecasters use both.

ECMWF is widely regarded as the world's leading global medium-range weather model, strongest for the synoptic pattern several days ahead; the Met Office's Unified Model, and especially its high-resolution UKV, is strongest for local detail over Britain. They are complementary rather than competing, and the Met Office - an ECMWF member state - uses both. The best UK forecast comes from comparing them, not picking one.

Who runs each model?

ECMWF, the European Centre for Medium-Range Weather Forecasts, is an independent intergovernmental organisation supported by its member and co-operating states, with sites in Reading, Bologna and Bonn. Its flagship is the Integrated Forecasting System (IFS), which produces the high-resolution deterministic forecast (often called HRES) and a large ensemble. ECMWF's reputation as the leading global model is long-standing in the meteorological community.

The Met Office is the UK's national weather service. It runs the Unified Model, whose global configuration is often shorthanded UKMO and whose UK high-resolution configuration is the UKV, at around 1.5 km grid spacing. Crucially, the UK is an ECMWF member state, so the Met Office both develops its own model and draws on ECMWF's output - they are collaborators, not enemies.

What each model is built to do

The clearest way to compare them is by the job each is designed for.

DimensionECMWF (IFS)Met Office (Unified Model / UKV)
OperatorECMWF (intergovernmental, Reading/Bologna)Met Office (UK national service)
CoverageGlobalGlobal model plus high-res UKV over the UK
Core strengthMedium-range synoptic patternLocal UK detail; short-range hazards
Resolution edgeStrong global grid and large ensemble~1.5 km UKV over Britain
Best used forDays-ahead planning, storm trackingShowers, convection, coastal and local timing

Where ECMWF tends to win

For the medium range - broadly two to seven days ahead - ECMWF has an excellent track record and is generally the model professionals reach for first when judging the large-scale pattern: where the jet stream will sit, how a North Atlantic depression will track, and whether a spell of settled or unsettled weather is coming. Its large ensemble is a particular strength, because it samples many plausible outcomes and gives a robust read on uncertainty. For big-picture UK weather several days out, ECMWF is a formidable baseline.

Where the Met Office tends to win

Zoom in and the picture changes. The UKV's roughly 1.5 km grid can represent features a coarser global model smooths away: individual showers, thunderstorm initiation, sea breezes, and the way the Pennines or the Welsh mountains wring out rain. For the next day or two over Britain - the timescale that matters for "will it rain on my afternoon?" - the Met Office's high-resolution modelling, tuned specifically for UK conditions and fed by dense local observations, is hard to beat. It is also the model behind the National Severe Weather Warning Service and its yellow, amber and red warnings.

Do they ever disagree - and what then?

Yes, and the disagreements are informative. Because ECMWF and the Met Office use different data assimilation, resolutions and physics, they can diverge on the track of a storm, the timing of a front or how far south snow will reach. When they agree, confidence in the forecast is high. When they split, the honest interpretation is that the atmosphere is genuinely uncertain and you should plan flexibly. Add the American GFS (NOAA) and German ICON (DWD) to the mix and you get an even fuller sense of the spread.

This is precisely what independent comparison does well. checkweather.io lines the Met Office, ECMWF, GFS and ICON up side by side for the UK, converts their level of agreement into a plain-English confidence score, and grades an accuracy scoreboard against recorded weather. Rather than declaring a permanent winner, it shows which models are converging today and how each has been performing lately - the "no black boxes" approach.

So which should you trust?

Both, for different questions. If you are planning several days ahead and want the overall pattern - storm risk, a settled spell, a cold snap - weight ECMWF heavily. If you want to know whether a shower will hit your town this afternoon, lean on the Met Office UKV and the official forecast. And for anything hazardous, follow Met Office warnings and Environment Agency flood alerts regardless of what any single model shows. The professional habit is never to marry one model: read them together, respect their disagreements, and let convergence - not a single confident-looking chart - tell you when to trust the forecast.

What about the other global models?

ECMWF and the Met Office are not the only players, and a serious UK forecast weighs more than two opinions. NOAA's GFS, run in the United States, is free, frequently updated and the engine behind many consumer apps; it is a capable global model even if it is generally considered a step behind ECMWF over the North Atlantic. Germany's DWD runs ICON, another well-respected global and regional system that often handles fronts approaching from the near-continent well. Each brings a slightly different view of the same atmosphere. When ECMWF, the Met Office, GFS and ICON all converge on the same solution, that agreement is a powerful confidence signal; when they scatter, it is an honest warning that the forecast is uncertain. Treating the four as a panel of expert witnesses, rather than crowning one, is how professionals actually work - and it is the reasoning a comparison site formalises into a single confidence score.

The takeaway

ECMWF versus the Met Office is not a fight with a winner. ECMWF is the world-class global medium-range model; the Met Office adds the high-resolution UK detail and issues Britain's official warnings. The two are collaborators, and the smartest reading of UK weather comes from comparing them - alongside GFS and ICON - and treating their agreement as your real confidence signal.

Common questions

Is ECMWF more accurate than the Met Office model?
For global medium-range forecasts several days ahead, ECMWF is widely regarded as the leading model. But the Met Office UKV is stronger for high-resolution local detail over Britain. Neither is simply more accurate - it depends on the timescale and scale you care about.
Does the Met Office use ECMWF forecasts?
Yes. The UK is an ECMWF member state, so the Met Office uses ECMWF output alongside its own Unified Model. They are collaborators, and professional forecasts blend multiple models.
What is the difference between ECMWF IFS and the Met Office Unified Model?
They are separate modelling systems built by different organisations, with different data assimilation, resolutions and physics. ECMWF's IFS is a global medium-range system; the Met Office Unified Model includes a global configuration plus the high-resolution UKV over the UK.
When ECMWF and the Met Office disagree, which should I believe?
Disagreement is a signal of genuine uncertainty. Weight ECMWF for the medium-range pattern and the Met Office UKV for local next-day detail, plan flexibly, and follow Met Office warnings for anything hazardous.
Are there other models worth comparing for the UK?
Yes. The American GFS (NOAA) and German ICON (DWD) are both useful for the UK. Comparing several models reveals the spread of possible outcomes, which is more informative than trusting any single one.

Sources

Compare UK forecasts, live

More guides