Why elevation and 850 hPa are not enough
The 850 hPa level is a pressure surface, not a fixed altitude. Levels such as 925, 850, 700 and 500 hPa, when data are available, help show how temperature and humidity change with height. Levels below the ground do not describe the air above the location and must be excluded. All these levels are not always available, and a few readings cannot reveal every warmer layer between them. A cold value at 850 hPa, or a near-surface temperature below freezing, is not enough to distinguish snow, rain and freezing precipitation.
T850 describes temperature at one atmospheric level. A simple linear scenario uses T(z) = T850 + Γ × (Z850 − z) / 1,000, with heights in metres and Γ in °C/km. At Γ = 6, moving 100 m higher corresponds to 0.6 °C lower temperature within that scenario. The 4, 6 and 8 °C/km columns test sensitivity; they are not a calibrated forecast or fixed atmospheric constants.
The dry-bulb freezing level differs from the wet-bulb zero level and the actual snowfall limit. Source: NOAA/NWS, vertical profiles and winter precipitation .
Elevation, inversions and local terrain
Temperature usually decreases with height, but the rate varies with the weather. A simple adjustment of 6.5°C per km cannot describe every situation. On clear, calm nights, cold air can collect in valleys and enclosed basins, making the valley floor colder than a slope above it. This is a temperature inversion. Snow cover also affects daytime heating and night-time cooling, while cloud and wind can change these local contrasts.
NOAA/NWS: temperature inversions
NW or NE wind: what changes for snow?
Drier air can cool more through evaporation of precipitation, but it can also evaporate much of it before it reaches the ground. This mechanism may matter in a northwesterly flow; direction alone does not guarantee low humidity or snow. Check dew point, wet-bulb temperature and the actual air flow. NWS case study: dry air, wet-bulb temperature and precipitation-type changes .
Aegean sea-effect snow: can a northeasterly wind enhance snowfall?
Yes. Cold air crossing warmer water can acquire heat and moisture, helping clouds and precipitation develop. The air track, wind changes with height, depth of the cold layer and uplift over terrain influence where precipitation is concentrated. Sea effect is the marine counterpart of lake effect. Study of marine enhancement of southern European snowfall in January 2017 .
The study of storm Elpis in January 2022 linked Attica’s snowfall to the large-scale circulation and terrain. Simulations with warmer sea temperatures increased snow mainly at higher elevations, while a cooler sea favoured longer-lasting snow cover lower down. Northeasterly winds and marine moisture therefore do not have one fixed effect everywhere. Patlakas et al., University of Athens / HNMS study of Elpis .
Can it snow above 0 °C?
Yes. Near-surface temperature is only part of the picture. Flakes pass through layers with different temperatures and humidity. Evaporation and melting cool them, while a deep warm layer can turn them into rain. A single thermometer reading cannot determine whether it will snow. Source: NOAA/NWS, winter precipitation types .
Humidity, dew point and wet-bulb temperature
Dew point describes how far air must cool to reach saturation. Wet-bulb temperature estimates the temperature it can approach through evaporative cooling. In saturated air both equal air temperature. Lower humidity favours more evaporative cooling, but very dry air can also evaporate precipitation before it reaches the surface. Source: FAO, atmospheric humidity .
Millimetres of water and centimetres of fresh snow
Millimetres describe precipitation as water; centimetres describe the thickness of fresh snow. For a given snowfall water equivalent, cm = mm × snow-to-liquid ratio / 10. For example, 10 mm falling entirely as snow gives 7 cm at an assumed ratio of 7:1, or 10 cm at 10:1. These are different density assumptions, not guarantees. If some precipitation falls as rain, only the snow portion enters this conversion. Source: NOAA/NWS, snow-to-water ratios .
Fresh snowfall is different from snow depth on the ground
Falling snow can melt, settle, compact or be redistributed by wind. The surface may be warm or already covered by older snow. As a result, 10 cm of fresh snowfall does not necessarily increase snow depth on the ground by 10 cm. The amount falling, the snow that remains and an on-site measurement are different pieces of information.
14-day snow and weather forecast for Agios Stefanos · Greece weather forecast with map · Satellites and radar · Historical station measurements
Sunshine, night-time cooling, wind and melting
Melting depends on the heat snow gains and loses, not just on air temperature. During the day, absorbed sunlight can warm it; clean white snow reflects more sunlight than older or dirty snow. At night, the surface often cools by emitting radiation, particularly under clear skies. Wind transfers heat and moisture and redistributes snow, while the ground can supply or absorb heat. Snow can therefore persist differently on a shaded slope, a sunny surface and a road, even at the same air temperature.
USDA NRCS: snowmelt and energy exchange
Calculation method and limitations
Wet-bulb temperature is solved numerically from the air–water energy balance and station pressure. Saturation vapour pressure and dew point are referenced to liquid water (Magnus, 17.62 / 243.12). The liquid-water thermodynamic wet-bulb equation is used; below freezing it approximates supercooled water, not an ice bulb. Pressure is estimated from the standard atmosphere when no observation is supplied. Energy-balance equation: ASHRAE / PsychroLib
Centimetres remain conditional: precipitation in mm × snow fraction / 100 × ratio / 10. A 10:1 ratio is a reference, not a locally calibrated default. The 5:1 and 15:1 examples show sensitivity, not a statistical confidence interval. Precipitation type, the ground and the event’s evolution determine what remains.
How this relates to Xionia forecasts
The calculator explores your own assumptions. Xionia forecasts show model amounts and, when needed, a snowfall estimate for the particular location marked ≈. The combined forecast uses the eligible models available for each variable and period; the models contributing to temperature may differ from those contributing to snowfall.
Compare amounts for the same period and location. One model hour does not describe a whole snowfall event, and the day's minimum temperature may occur outside the precipitation period. A blank means information is missing; zero is an available zero amount. Estimated values should remain distinguishable from native model quantities and observations.
What can the supplied atmospheric levels show?
The calculator examines surface wet-bulb temperature separately from levels where temperature, humidity and height are all supplied. It estimates zero crossings and the total depth of positive wet-bulb temperature by linear interpolation between these points. Unsampled warm layers can be missed; the result is not a verified snowfall elevation. Sims & Liu (2015) .
The 0.5 and 1 °C wet-bulb thresholds are two published benchmarks. When they disagree, the result remains transitional. A warm layer triggers a melting or mixed-precipitation indication. A 2025 study found that even more complex surface-based methods struggle near freezing. Jennings et al. (2025) .
The GFS button imports the station’s next available hourly forecast, including the 925/850 hPa levels. Hourly precipitation covers the interval ending at the displayed time. The import uses one model, without the blended forecast’s calibration. It does not supply ground or sea temperature. Source: GFS through Open-Meteo .