# The Science of Powder (Part 2): Snow-liquid ratio, a dry-snow line you can measure Snow-liquid ratio quantifies how airy a snowfall is. How it is calculated, how temperature drives it, and why forecast snowfall cannot give it to you. - URL: https://www.ykhokkaido.com/en/blog/powder-science-02-slr - Published: 2026-09-12 - Section: Powder - Series: The Science of Powder (part 2) - Tags: Snow-liquid ratio, Dryness --- Part 1 ended by handing you a tool. Part 2 takes it apart. ## The definition **Snow-liquid ratio (SLR) = depth of snow ÷ depth of water it melts to.** Japan writes it as cm/mm: centimetres of snow per millimetre of water. English-language sources usually write a ratio, where 10:1 means ten centimetres of snow melting to one of water. Same quantity, and the only trap is the factor of ten between the two conventions. Some reference points: - The global average sits near 13:1 - Japan's dry-snow line is 2.0 cm/mm, which is 20:1 - What gets marketed as "champagne powder" usually runs 20:1 to 25:1 and up ## What controls it Mostly **the air temperature while the snow is falling**. That is the conclusion of Roebber and colleagues' widely cited 2003 paper, and the reason the US National Weather Service runs an operational method built on a single temperature predictor. The causal chain is short: colder air → crystals keep their branches instead of fusing → they prop each other apart on landing → the same water builds a deeper pile → higher ratio. Humidity is secondary. Drier air means less riming — fewer supercooled droplets freezing onto the crystal on the way down — and lighter snow. Heavily rimed crystals arrive as dense little pellets and stack accordingly. Wind does not change the ratio, but it does break crystals up and pack them down once they have landed. So a high-ratio snowfall accompanied by strong wind can still leave a hard surface. **These are two separate judgements**, and our model keeps them as separate multipliers for exactly that reason. ## Why you cannot read it off a forecast This is the easiest trap to fall into when you build something on top of weather data, so it deserves its own section. Most weather APIs expose a "snowfall" field. It looks obvious: divide it by precipitation and you have the ratio. But that field is usually not an observation. It is precipitation multiplied by a **fixed ratio**, commonly 7:1. So whether the day is −20 °C or −1 °C, your computed ratio comes out the same. Which means: **SLR has to be modelled from temperature, not derived from the snowfall field.** We use a Kuchera-style operational heuristic — not a rigorous physical model, but an empirical curve that a national weather service actually runs in production, and one that needs only a temperature to produce a usable estimate. Because it is a heuristic, we label it as an estimate everywhere it appears on our own pages: not a measurement, and not an official forecast. ## Three things to take away 1. SLR = snow depth ÷ water depth. The dry-snow line is 2.0 cm/mm (20:1). 2. Temperature during snowfall dominates, humidity is secondary, and wind does not change the ratio but wrecks the surface after landing. 3. The snowfall field in a forecast API usually has a fixed ratio baked in, so it cannot be used to recover SLR. Part 3 goes to the place this series most needs to be careful about: converting SLR into a 0–100 dryness score, and **which half of that conversion is somebody's published fact and which half is our own inference**. For Furano's run and lift data, see [Furano's runs & lifts page](/resorts/furano/guide). ## FAQ ### What is snow-liquid ratio (SLR)? Snow depth divided by the depth of water it melts down to. Ten centimetres of snow melting to one centimetre of water is 10:1. Japan writes it as cm/mm; the Japan Meteorological Agency's dry-snow threshold of 2.0 cm/mm is the same as 20:1. ### Why can't I compute SLR from forecast snowfall? Because in most weather APIs the snowfall field is not observed. It is precipitation multiplied by a fixed ratio, commonly 7:1. Dividing it by precipitation just returns that constant, carrying no information about the day's actual dryness. ### How cold does it need to be for dry snow? Surface temperatures at or below about −8 °C during snowfall favour high ratios, and above roughly −2 °C the ratio falls away quickly toward wet snow. That is a tendency, not a guarantee — wind and cloud conditions change the outcome. ## Sources - Japan Meteorological Agency — dry/wet snow classification by snow-liquid ratio - Roebber et al. (2003), Improving snowfall forecasting by diagnosing snow density, Weather and Forecasting - NWS Kuchera snow-liquid ratio method (operational heuristic)