Weather
Six weather models, prepared for your summit — instead of the value at a grid point that applies to the valley next door. How each figure comes about is set out openly in the method paper.
What decides is not the weather in the valley but what happens at your elevation:
- Wind and gusts
- whether you stay able to act on the ridge
- Apparent temperature
- clothing, reserves, turnaround point
- Fresh snow
- how much is added up there
- Freezing level
- rain or snow, icing
- Sunshine duration
- firn window and dry rock
- Thunderstorms
- the window for the summit
It is not the wind that decides a tour, it is the gust
The mean wind above, the strongest gust of the block in brackets below — often the number that decides a tour. The colour follows the gust: orange from 60 km/h, red from 80. Which air layer counts depends on the summit height.
What the thermometer says, and what the body feels
What the thermometer measures is not what reaches you on the ridge — the wind makes the difference. Highs and lows are not averaged: a model forecasting a frosty night should not vanish into the mean.
How much snow arrives up there
Most models place their grid point hundreds of metres below the summit, where rain falls instead of snow. How much arrives up top is worked out for the summit height.
The moment rain turns to snow
The grey line is the rain–snow boundary: red bars above it, blue below. Every altitude band is judged on its own — snow at the summit, rain further down, in the same time window. The number row stays the familiar freezing level. It sits higher, because falling precipitation cools on the way. Only altitude bands the models actually resolve are shown — depending on the mountain that is two or three.
Thunderstorm risk as a single number
Five quantities, one number: how thunderstorm-prone the situation is. Red from 4. Not a lightning probability, but a pointer to which time window is better avoided.
What the weather services say themselves
Above the forecast, the official warnings of the national services run alongside — for the warning area your location falls into, rather than as a blanket overview for an entire country.
How certain is any of this?
The probability puts the amounts above it into context — the same centimetres mean something different at high probability than at low. Instead of a zero it reads „under 1 per cent“: a zero would read as a promise, and the number does not carry one.
Ten days, four blocks per day
For the longer view, four six-hour blocks a day. From day three the global models weigh more — from day six the regional Alpine models are out of range.
Where the numbers come from
Six models, three of them regional — with a mesh of 2.2 to 6.5 km instead of 13 to 25 km. The fine resolution captures valleys and ridges, the coarse ones reach further into the future — which is why both run side by side.
In the app they can be switched on and off individually — the numbers then show what the selected ones yield. Missing model values do not count as zero — they drop out of the averaging. One exception is the thunderstorm risk: two of its input quantities come from a single model only — it therefore always computes across all models that have data.
- ICON D2
- DWD · 2.2 km · 48 hours
- ICON EU
- DWD · 6.5 km · 5 days
- AROME
- Météo-France · 2.5 km · 48 hours
- ECMWF IFS
- ECMWF · 25 km · 10 days
- AIFS
- ECMWF · 25 km · 10 days
- GFS
- NOAA · 13 km · 10 days
1600 mountain stations for context
Current measurements come with the forecast: stations above 800 m from 19 networks across the Alpine region. The nearest one sits in the app directly below the forecast.
It shows what is actually being measured on the mountain right now — conditions on site alongside the forecast. Elevation, terrain shape and aspect of the station can differ from the summit.
- 🇨🇭 Switzerland
- SLF IMIS · MeteoSwiss
- 🇦🇹 Austria
- GeoSphere Austria
- 🇩🇪 Germany
- DWD
- 🇮🇹 Italy
- Bolzano · Trento · Piedmont · Lombardy · Veneto · Aosta Valley · FVG
- 🇫🇷 France
- Météo-France
- 🇸🇮 Slovenia
- ARSO
And how it is calculated is out in the open
Between the output of a weather model and a number for a named summit lie several steps: selection, weighting, conversion, correction. The method paper describes it in full — with sources, with every constant that was set, and with an assessment of how well each individual quantity holds up.
Read the method paper (PDF, German)Data sources: Model data via our own Open-Meteo instance — ICON EU and ICON D2 (Deutscher Wetterdienst), AROME (Météo-France), IFS and AIFS (ECMWF), GFS (NOAA). Measurements from SLF IMIS, MeteoSwiss, GeoSphere Austria, DWD, Météo-France, ARSO and the regional services of South Tyrol, Trentino, Piedmont, Lombardy, Veneto, Aosta Valley and Friuli Venezia Giulia. Official warnings via MeteoAlarm. The forecasts shown are from 9 September 2026.