← Back to the indicators

Understanding the foehn

The foehn is southern air crossing the Alps. The easiest way to picture it is to think of water.

1. Air behaves like water

A low to the west pushes southern air against the Alps. The air hits the barrier like water against a dam. It rises, it cools, and it drops its moisture: it rains in Ticino.

What pushes is the pressure difference between the south and the north of the Alps: higher in the south, lower in the north. Air runs from high to low the way water runs downhill. That is the ΔP on the graphs on the foehn page. Beyond about 4 hPa a foehn becomes possible: measured, there is foehn at Visp four times out of ten.

Once over the crest, the air comes back down. Descending, it compresses and warms, by about 1 °C every 100 metres. It arrives warm, dry and fast. That is why a foehn day is mild and crystal clear north of the Alps while it rains in the south.

foehn wall+high pressurelow pressureΔP south − north: what drives itSOUTH · Ticinothe air rises, it rainsNORTH · Valaisthe air descends, warm and dry

2. The Rhone valley is a bathtub

Before the foehn breaks in, the Rhone valley is full of cold air. That cold air is heavier, so it stays at the bottom, like water in a bathtub.

The foehn passes over it first. On the ground you feel nothing: the sky clears, the wind aloft picks up, but the launch is calm. That is the most deceptive moment.

Then the foehn pushes. It drives the cold air down-valley, from Brig towards Lake Geneva. When the bathtub empties at your height, the foehn reaches the ground. All at once.

Brig · VispSierre · VercorinSionMartignythe foehn comes in hereit pushes the cold air that way

3. Why Visp should worry you at Vercorin

South foehn enters the Rhone valley over the Simplon, at Brig and Visp. Then it works down the valley westwards: Sierre, Sion, Martigny.

So if Visp is already in foehn, with a steady, warm, dry easterly, the front is on its way to you. At Vercorin the wind can go from calm to over 40 km/h in a few minutes. You can be in the air when it breaks in.

An easterly at Visp is not always foehn, and the foehn does not always come down as far as Sierre. It is the other signs, taken together, that make the difference.

4. The foehn does not fill the Rhone valley all at once

Back to water. Through a bend a river does not flow evenly across its width: it hugs one bank. Air does the same coming down the Rhone.

It hugs the right bank first: Montana, Crans. Then it widens towards the middle. And if it strengthens further, it ends up filling the whole width, as far as the left bank, where Vercorin is. It does not move from one bank to the other: it grows.

Direct consequence: Montana sees the foehn more often than Vercorin does. At Vercorin, over time, you end up believing it never gets in. That is false. It gets in less often, and that is exactly what makes the day it does get in dangerous: nobody is expecting it.

It gets worse. If the foehn reaches the left bank, that means it has filled the entire valley. Put another way, by the time it finally reaches Vercorin, the whole valley is already in it.

In practice: look at the far bank before you launch. If the foehn is already established at Montana, it can reach you.

the RhoneMontana · Crans (right bank)Vercorin (left bank)the foehn breaks inBrig · Visptowards Sion, Martigny

5. The signs

  • Easterly at Visp, steady, with strong gusts.
  • Rain in Ticino while it stays dry here: the air is unloading in the south.
  • Positive ΔP between the south and the north of the Alps, beyond about 4 hPa.
  • Southerly at 700 hPa above the crest, what drives the whole thing.
  • Temperature rising and humidity collapsing on the valley floor, at the same time.
  • Exceptional visibility, crystal-clear air, mountains that look close.
  • Foehn wall on the crest, lenticulars, clouds that sit still.

6. The risks in free flight

  • When it breaks in, it is brutal. Not a gradual build: a few minutes.
  • The foehn is often up there before it is on the ground. A calm launch tells you nothing about what is happening 200 metres higher.
  • Strong turbulence, rotors in the lee of crests and terrain. Inside one you are a dead leaf in the wind: you are no longer the one flying.
  • Shear at the boundary between the cold air and the foehn: it changes fast, over very little height.
  • Going backwards. If the wind exceeds the wing's speed you go backwards, into the turbulence.
  • « The foehn never gets into Verco. » It gets in less often. That is not the same thing.

What the measurements say

Everything below comes from our own records, not from a textbook. Every figure states what it is calculated on and over what period: a number without a method is only an opinion.

1. When: the foehn is a spring affair

Foehn hours at Visp in 2018, month by month. 670 hours over the year, 197 of them in April alone, and zero in July.

Visp, 2018. Full MeteoSwiss method (Arbeitsbericht 223), humidity included: 2018 is the last year our records held it.

What to take from it: in summer, an absence of foehn proves nothing about how you are reading the instruments. It is the season that is quiet. A whole summer of measurements cannot set a foehn threshold: we learned that by trying.

1b. Not all valleys are alike

Same year, same method, applied to every valley station in MeteoSwiss Table 1. Visp is the most exposed in Switzerland; Sion, 60 km down-valley, gets a quarter of it.

2018, full method. Güttingen (25 h) left out for legibility.

What to take from it: « there is foehn in Switzerland » means nothing. There are foehn valleys, and the upper Valais is one of them.

2. Where: right bank against left bank

The most important figure on this page. During a foehn measured at Visp, here is the share of the time each site actually feels the wind at home.

Oct 2024 to Jul 2026. A site counts as « lit » when its wind is inside its own foehn sector and above 10 km/h.

What to take from it: the foehn comes down the right bank first. Bella-Lui and Montana feel it two times in three; Vercorin and the Crêt du Midi almost never, and for Vercorin no wind sector even separates it from ordinary weather. That is where the trap is: « there is never any foehn here » is an accurate observation and a false conclusion. Vercorin's anemometer will not warn you. Montana's will, and it is seven kilometres away.

3. How far down it comes

Same window, but on the valley floor. The foehn very rarely reaches the middle and the lower valley.

Oct 2024 to Jul 2026, during a foehn measured at Visp.

What to take from it: a foehn at Visp does not mean a foehn at Sion. The upper valley takes most of it. And above all, when only the lower valley blows down-valley while Visp does not move, that is not a foehn. Out of 737 samples of that kind, zero.

4. The chain beats a single station

When the strip lights from the top down, here is the probability that there really is foehn:

What is litFrequencyFoehn
Visp alone11,0 %62 %
Visp + Varen5,2 %77 %
Visp + Varen + Sion0,5 %81 %
Lower valley only, Visp calm0,9 %0 %

What to take from it: read the strip from the top down, never the other way. Two stations lighting in order are worth far more than one, and the bottom on its own is worth nothing.

5. Binn and Jeizinen: sensitive, but talkative

The higher a station sits, the earlier it sees the foehn, and the more often it blows from the same sector with nothing going on. Solid bar = share of foehn caught; light bar = share of the rest of the time it lights anyway.

Oct 2024 to Jul 2026, 1116 foehn hours over 152 days.

What to take from it: Jeizinen catches 98 % of foehn, but also lights one time in three for no reason. Chalais is almost never wrong, and misses three foehn out of five. Neither is « better »: you watch the high ones to be warned early, the low ones to be sure. A silent high station is good news; a silent low station proves nothing.

6. The thresholds on the graphs

The markers drawn on the curves on the foehn page are not decorative. What they are worth, measured:

  • ΔP > 4 hPa: above it there really is foehn at Visp one time in two on the Simplon axis (Domodossola - Sion), four times out of ten on the Gotthard one (Lugano - Zurich). The two axes behave almost identically, despite 140 km between them.
  • Gütsch, south component > 25 km/h: happens 4 % of the time. Every hour of steady easterly measured at Visp and at Varen over the period followed such an hour within twelve hours. The reverse is not true: the south often picks up at Gütsch with nothing coming down.
  • Stability 0.75 °C/100 m: above it the valley is mixed enough for a foehn to reach the ground. An ordinary day sits around 0.58: below the dry adiabat is the norm, not an anomaly.

7. What these figures changed

Every measurement on this page changed the indicators or the thresholds. The full list, so that nothing rests on an unverified hunch:

What the measurements sayWhat changed
Two summer months contain almost no south foehn (Altdorf 13 h, Vaduz 4 h, Engelberg 1 h)Dropped our sectors computed on summer data; adopted the ones MeteoSwiss publishes
The down-valley sector differs from station to station: below Martigny it comes from the south-east, not the eastOne sector per station, never a blanket « east »
The night breeze runs down the valley every evening at 2-8 km/hA distinct « down-slope breeze » state, separate from established wind
The lower valley blows 20-30 km/h down-valley with no foehn upstreamThreshold per station: 25 km/h low down, 15 km/h high up
Aigle lit on its own: 0 foehn in 737 samplesA warning when Visp is calm; then Aigle removed from the strip
Visp + Varen together: 77 % foehn, against 62 % for Visp aloneThe strip names the chain: « the foehn is entering the Rhone valley »
Varen: 45-175° catches 78 % of foehn against 61 % for our old sectorVaren's sector widened
Chalais: no wider sector does better (41 % against 39 %)Sector kept, with a note that Chalais is barely exposed, not badly set
Binn and Jeizinen catch 90 % and 98 % of foehn, but light wrongly 44 % and 34 % of the timeKept out of the strip, explained as a watch signal
The potential temperature difference is negative 77 % of the time: that is the normThe « cold layer » box becomes « Rhone valley stability »
A potential temperature difference in kelvin cannot be readShown as a gradient, °C per 100 m, against the dry adiabat at 0.98
4 hPa: 50 % foehn on the Simplon, 42 % on the Gotthard, the two axes are equivalentThreshold kept; the measured figures added to the notes
Gütsch: the south often picks up with nothing coming down (26 % follow through)Presented as a watch signal, never as a forecast
Montana feels the foehn 69 % of the time, Vercorin 3 %A « right bank / left bank » section, and the trap named
Sierre has not reported since May 2020Station hidden everywhere: lists, map, sitemap

One trap that came back three times: taking a single foehn criterion (direction, temperature difference, pressure) and showing it as a verdict. Each one, on its own, describes ordinary weather most of the time. That is why every box now says what it measures, and how often it is wrong.

8. Method, and what these figures do not say

« Foehn at Visp » means, everywhere here, the MeteoSwiss rule for Visp (Arbeitsbericht 223, Table 1): direction gate at Gütsch, sector, minimum wind and gust, potential temperature difference with the crest, minus the humidity criterion, which the holfuy and meo stations do not record. So it is an approximation, and a fairly generous one.

The station comparisons cover 21 months (Oct 2024 to Jul 2026), the window where all the ones we care about were measuring at the same time. The seasonality comes from 2018, the last complete year with humidity. These figures describe what happened, not what will happen: they tell you where to look, never whether you can fly.

9. What the measurements will never say

Everything above comes from stations. A station measures where it was put, and the Valais has already shown us what that costs: at Vercorin foehn is measured 3 % of the time, at Montana 69 %, seven kilometres away. A Vercorin pilot trusting their own station will not see coming what Montana already sees. No amount of data fixes that, because the problem is not precision: it is that nobody is measuring in the right place.

For the Valais we filled the gap with years of flying and pilots who know where to look. For the Reuss, the Rhine, the Linth or the Haslital we have only the numbers, and numbers alone produce exactly the kind of page that looks right and misleads someone. If you fly one of those valleys, what you know is worth more than what we can compute: which launch feels it first, where people say there is never any foehn and there is, what you check before you get the wing out.

Tell us about your valley

← Back to the indicators