A great deal gets claimed about machine-made snow and very little gets calculated. We gathered the available figures, broke them down to a unit you can picture, and cross-checked them. Two of the three calculations come out cleanly. The third does not, and that is the interesting part.
The starting figures
Seilbahnen Schweiz, the Swiss cableway association, gives the following for technical snowmaking in Switzerland:
| Quantity | Value | As of |
|---|---|---|
| Electricity | 65 GWh per winter | 2022 |
| Water | 13 billion litres (13 million m³) | 2022 |
| Snowmakeable piste area | around 121 km² (12,100 ha) | 2022 |
| Share of pistes that are snowmakeable | 54 per cent | 2022 |
For comparison: in 1990 it was one per cent. Austria and Italy are at 70 to 90 per cent today, Germany and France at 25 to 39.
Calculation 1: the electricity, and it adds up
An older University of St. Gallen study put the energy demand in 2009 at 31,500 kWh per snowmade kilometre of piste per year. More efficient equipment has roughly halved that since, to about 17,000 kWh per kilometre.
Let us check that against the headline figure: 65 GWh divided by 17,000 kWh gives about 3,800 snowmade kilometres of piste. Switzerland has roughly 7,000 kilometres of piste, of which 54 per cent are snowmakeable – about 3,780. The two numbers meet to within half a per cent. Two independent routes, the same answer: this order of magnitude is sound.
What does 17,000 kWh mean? A Swiss four-person household uses about 4,500 kWh of electricity a year:
Calculation 2: the water, and here it jams
Now the same with water. Thirteen million cubic metres spread over 12,100 hectares of snowmakeable area gives about 1,075 m³ per hectare per season.
That figure can be checked against physics. Machine-made snow has a density of roughly 400 to 500 kg/m³, considerably more than natural snow. Allowing for the usual losses to evaporation and wind drift, one cubic metre of water yields roughly two cubic metres of machine snow. A base layer of 30 centimetres over one hectare therefore needs 3,000 m³ of snow, or about 1,500 m³ of water. Add the top-ups over the season and you land at 3,000 to 4,500 m³ per hectare.
That is three to four times more than the headline figure allows. And indeed a figure of over 4.5 million litres – 4,500 m³ – per snowmade hectare per season circulates for 2019. Both numbers cannot be describing the same thing.
The resolution, and who supplies it
The most honest answer comes from the research side itself. Fabian Wolfsperger of the SLF says of the water figure that it is “difficult to quantify”, because complete data collection across ski areas is not always possible.
That reconciles the numbers in two ways, which do not exclude each other:
- Snowmakeable is not snowmade. The 121 km² is the area that could be covered technically. In a normal winter considerably less is actually covered, and rarely everywhere to full depth. Thirteen million m³ divided by 4,500 m³ per hectare would give about 2,900 fully covered hectares – roughly a quarter of the snowmakeable area.
- The headline figure may be too low. If not every area reports fully, 13 million m³ is a lower bound, not a measurement.
What we can stand behind: per hectare of piste actually covered, 3,000 to 4,500 m³ of water is used – consistent with the physics and with the 2019 field figure. The frequently quoted Swiss total of 13 million cubic metres should be treated with care and probably refers to less area than the word “snowmakeable” suggests.
Per kilometre of piste (at roughly 3.2 hectares per kilometre) that means 10,000 to 14,000 cubic metres of water per kilometre per season. That is a 50-by-25-metre swimming pool, filled five to seven times.
Calculation 3: what it means for the climate
This is where it gets uncomfortable for both camps. At the Swiss consumer electricity mix of around 128 grams of CO₂ equivalent per kilowatt hour, 65 GWh works out to a good 8,000 tonnes of CO₂ equivalent per winter for all Swiss snowmaking.
That sounds like a lot. For comparison: a return flight Zurich–New York produces about one tonne per passenger. All Swiss snowmaking therefore equals roughly 8,000 transatlantic flights – a single well-filled day of air traffic.
None of which excuses anything, but it puts the proportions in place. In the climate balance of skiing the biggest item is not the snow, it is how the guests get there. Treating machine-made snow as the decisive climate lever means arguing about a few per cent and leaving the rest undiscussed.
The real problem with snowmaking is a different one, and a local one: water withdrawal from mountain streams and reservoirs in the season when flows are lowest anyway, and the land and landscape consumed by storage ponds and pipelines. Those are regional questions, not atmospheric ones.
And the one limit money cannot buy past
Snowmaking needs cold. Specifically, a wet-bulb temperature of about minus 3 degrees or lower – a value combining temperature and humidity. At minus 2 degrees with high humidity no snow gun runs, however much water and power stand ready.
This is exactly where warming bites. Not by melting the snow away, but by shortening the nights in which production is possible at all. And so the number that will decide the 2035 ski winter is not water use per hectare but the count of snowmakeable nights in November.
How that relates to weather and climate, and why those are two different things, is set out in our explainer on the subject.
As of 4 August 2026. The conversions to piste kilometres, hectares and CO₂ are our own calculations from the stated inputs and should be read as orders of magnitude, not measurements.
Sources: Seilbahnen Schweiz (electricity, water, areas, shares, as of 2022), cited via SRF; Institute for Systemic Management and Public Governance, University of St. Gallen (energy demand per piste kilometre, 2009, halved since); assessment of data quality by Fabian Wolfsperger, SLF; Wikipedia on snow cannons for consumption values and historical shares; CO₂ factor for the Swiss consumer mix per the KBOB reference value.
