How to Find Accurate Snow Weather Data

April 28, 2026

Snow is one of the more difficult weather variables to measure and interpret accurately.

Unlike temperature or pressure, snow can:

  • Accumulate over time
  • Melt
  • Compact
  • Drift because of wind
  • Vary significantly over short distances
  • Be reported less frequently than other weather observations

For this reason, it is important to understand the difference between snowfall, snow depth, and liquid-equivalent precipitation when using historical or forecast snow data.

Visual Crossing provides snow data through the Visual Crossing Weather API and Visual Crossing Weather Data tools.

If you do not already have an account, you can sign up for Visual Crossing Weather.

The main snow weather fields

The most important snow-related fields are:

snow
snowdepth
precip
preciptype

These fields describe different aspects of a winter-weather event.

Snowfall: snow

The:

snow

field represents the amount of new snow that fell, or is forecast to fall, during the period.

For a daily weather record, snow represents snowfall during that day.

For an hourly record, it represents snowfall associated with that hourly period.

For example:

snow = 3.2

means 3.2 units of new snowfall during the period, using the selected unit group.

In US units, snow is expressed in inches.

In metric and UK units, snow is expressed in centimeters.

Snow depth: snowdepth

The:

snowdepth

field represents the depth of snow already on the ground.

Snow depth is different from snowfall.

For example:

snow = 0
snowdepth = 8

means no new snow fell during the period, but approximately 8 units of snow remained on the ground.

Snow depth can change because of:

  • New snowfall
  • Melting
  • Compaction
  • Wind redistribution
  • Sublimation
  • Rain falling onto an existing snowpack

This means you should not calculate current snow depth simply by adding all previous snowfall values together.

Precipitation: precip

The:

precip

field represents the liquid-equivalent amount of precipitation.

This includes liquid rain and the liquid equivalent of frozen precipitation such as snow and ice.

For example, a snowstorm might report:

snow = 6.0 in
precip = 0.55 in

The snow value describes the depth of new snowfall, while precip describes approximately how much liquid water that precipitation represents.

The ratio between snowfall and liquid-equivalent precipitation can vary significantly between storms.

Dry, powdery snow can produce much greater snow depth from the same amount of liquid water than heavy, wet snow.

Precipitation type: preciptype

The:

preciptype

field identifies the type or types of precipitation associated with the period.

Possible values include:

rain
snow
freezingrain
ice

A weather record can contain more than one precipitation type if conditions change during the period.

For example, a day could include:

["rain", "snow"]

if precipitation begins as rain and changes to snow.

Use preciptype together with precip, snow, and temperature when analyzing winter weather.

Snowfall and snow depth are not interchangeable

A common mistake is to treat snowfall and snow depth as the same measurement.

Consider:

snow = 4
snowdepth = 10

This might mean:

  • Four inches of new snow fell.
  • Ten inches of total snow remained on the ground after the event.

Alternatively:

snow = 4
snowdepth = 3

could occur when melting, compaction, or drifting reduces the depth on the ground.

Always choose the field that matches your question.

Use:

snow

when asking:

How much new snow fell?

Use:

snowdepth

when asking:

How much snow was on the ground?

Why snow is difficult to measure

Snow measurements can vary more than many other weather variables.

Several factors contribute to this.

Wind

Wind can move snow from one location to another.

One measurement site may become partially cleared while another develops a drift.

Compaction

Fresh snow settles and compacts after it falls.

Ten inches of freshly fallen snow may become significantly shallower even if no melting occurs.

Melting

Ground temperature, air temperature, sunlight, and rain can reduce snow depth.

Terrain

Elevation and exposure can produce large differences in snowfall over short distances.

Observation frequency

Some snow observations are reported less frequently than temperature, wind, or pressure.

As a result, high-resolution snow data should always be interpreted in the context of the underlying observation availability.

Historical snow data

Historical snow data is available through the Timeline Weather API and Weather Data tools.

A historical daily request might look like:

https://weather.visualcrossing.com/VisualCrossingWebServices/rest/services/timeline/Denver,CO/2026-01-01/2026-01-31?unitGroup=us&include=days&elements=datetime,tempmax,tempmin,precip,preciptype,snow,snowdepth&key=YOUR_API_KEY

This returns fields useful for analyzing winter weather, including:

datetime
tempmax
tempmin
precip
preciptype
snow
snowdepth

For complete API details, see the Timeline Weather API documentation.

Hourly historical snow data

If the timing of snowfall is important, request hourly records.

For example:

https://weather.visualcrossing.com/VisualCrossingWebServices/rest/services/timeline/Denver,CO/2026-01-15?unitGroup=us&include=hours&elements=datetime,temp,precip,preciptype,snow,snowdepth&key=YOUR_API_KEY

Hourly data can help answer questions such as:

  • When did the snow begin?
  • During which hours was snow reported?
  • When did temperatures fall below freezing?
  • How did snow depth change during the event?

However, snow observations can have a lower native observation frequency than some other weather elements.

An hourly value should therefore not automatically be interpreted as meaning that a physical snow-depth measurement was made every hour.

Forecast snow data

The Timeline Weather API also provides forecast snowfall and snow-depth information where available.

A forecast request can include:

snow
snowdepth
precip
precipprob
preciptype

For example:

https://weather.visualcrossing.com/VisualCrossingWebServices/rest/services/timeline/Denver,CO?unitGroup=us&include=days,hours&elements=datetime,temp,precip,precipprob,preciptype,snow,snowdepth&key=YOUR_API_KEY

Forecast snowfall can be useful for:

  • Transportation planning
  • Snow removal
  • Event planning
  • Facility operations
  • Agriculture
  • Travel
  • Winter-risk monitoring

Use precipprob only for forecasts

The:

precipprob

field represents the forecast probability of measurable precipitation.

It does not represent snowfall amount.

For example:

precipprob = 80
snow = 4.5

describes an 80% probability of measurable precipitation along with forecast snowfall of 4.5 units.

Do not interpret:

precipprob = 80

as meaning an 80% chance of receiving exactly the predicted snowfall amount.

Daily snow values

Daily snow records aggregate the underlying period according to the meaning of each field.

Current Visual Crossing Weather Data documentation defines:

snow

as the total new snowfall for the daily period.

snowdepth

represents snow on the ground, with the daily value representing the daily snow-depth summary.

This distinction is important when creating monthly or seasonal snow statistics.

Calculate total snowfall

To calculate total snowfall over a period, sum the snow values.

For example:

total January snowfall =
sum(daily snow values for January)

This is appropriate because snow describes new accumulation during each period.

Do not sum snow depth

Do not calculate:

sum(snowdepth)

to determine seasonal snow accumulation.

Snow depth is a state variable describing the snow present on the ground.

For snowpack analysis, useful statistics may include:

maximum snowdepth
mean snowdepth
number of days with snowdepth > 0

depending on the analytical question.

Example: analyze a snowstorm

Suppose you want to understand a winter storm.

Useful weather fields include:

datetime
temp
feelslike
precip
preciptype
snow
snowdepth
windspeed
windgust
visibility
conditions

Together these can answer questions such as:

  • When did precipitation start?
  • Was it rain, snow, freezing rain, or mixed precipitation?
  • How much new snow accumulated?
  • How much snow was on the ground?
  • Was strong wind causing blowing or drifting snow?
  • Was visibility reduced?
  • Did temperatures support melting or refreezing?

Snow should rarely be analyzed using only one weather field.

Example winter-weather record

A record might contain:

{
  "datetime": "14:00:00",
  "temp": 28.4,
  "precip": 0.08,
  "preciptype": ["snow"],
  "snow": 0.9,
  "snowdepth": 5.6,
  "windspeed": 18.2,
  "windgust": 29.7
}

This provides much more information than simply saying “it snowed.”

It indicates:

  • Below-freezing temperature
  • Snow as the precipitation type
  • New snowfall
  • Existing snowpack
  • Strong winds that may affect drifting or visibility

Historical snow sources

Historical weather can use multiple available observation sources.

For station-based snow observations, the available observation network varies by:

  • Location
  • Date
  • Weather element
  • Reporting station
  • Observation frequency

Visual Crossing can also use additional remote observation sources where appropriate for other weather elements.

The precise sources available for one element should not automatically be assumed to apply to every other element in the same record.

For example, the best precipitation source may differ from the best source for temperature or snowfall.

Nearby snow observations may be some distance away

Snow observations are less common than basic temperature or pressure observations in some areas.

As a result, the closest station reporting snowfall or snow depth may sometimes be farther from the requested location than stations reporting other weather fields.

This matters particularly in:

  • Mountainous areas
  • Rural regions
  • Remote locations
  • Areas with highly localized snowfall

When snow accuracy is critical, inspect nearby stations and compare their location and elevation with the requested point.

See How Do I Find My Nearest Weather Station?.

Elevation matters

Snowfall can vary dramatically with elevation.

Two locations only a few miles apart can experience very different snowfall if one is significantly higher.

When investigating historical snow data, compare:

  • Requested-location elevation
  • Observation-station elevation
  • Distance
  • Terrain
  • Exposure

Visual Crossing provides advanced station-search controls including:

maxStations
maxDistance
elevationDifference

where applicable.

See Modifying the Weather Station Search Parameters.

Check snow together with temperature

Temperature provides important context when evaluating snow.

For example:

temp = 35°F
snow = 0
precip = 0.20
preciptype = ["rain"]

is very different from:

temp = 28°F
snow = 2.5
precip = 0.20
preciptype = ["snow"]

Even when the liquid-equivalent precipitation is identical, the resulting surface conditions can be very different.

Check wind when evaluating snow depth

Wind can strongly affect reported snow depth.

Useful fields include:

windspeed
windgust

Strong winds can cause:

  • Drifting
  • Blowing snow
  • Uneven snow depth
  • Reduced visibility

This means an individual snow-depth observation may not represent every nearby surface equally.

Check precipitation type for mixed winter events

Many winter storms involve transitions between:

snow
rain
freezing rain
ice

A daily snowfall total alone may hide these transitions.

Hourly data with preciptype can help show whether a storm:

  1. Began as snow.
  2. Changed to freezing rain.
  3. Changed to rain.
  4. Returned to snow.

This can be important for transportation, insurance, and event analysis.

Snow data for remote locations

Snow data can be particularly challenging in remote regions because the observation network may be sparse.

If snow coverage appears limited:

  1. Verify the requested coordinates.
  2. Inspect nearby stations.
  3. Check station elevation.
  4. Review the date and available snow fields.
  5. Consider increasing the station search area where appropriate.
  6. Compare related fields such as precipitation, temperature, and precipitation type.

See Finding Historical Weather Data for Remote Locations.

Snowfall versus liquid-equivalent precipitation

One important concept is the snow-to-liquid ratio.

For example:

10 inches snow
1 inch liquid precipitation

would correspond to approximately a 10:1 snow-to-liquid ratio.

But this ratio is not fixed.

Some storms produce:

5:1

heavy wet snow, while cold dry snow can produce much higher ratios.

For this reason, do not convert precip into snowfall using a fixed 10:1 ratio unless that approximation is specifically appropriate for your application.

Use the returned:

snow

field when snowfall depth is required.

Snow depth cannot be reconstructed reliably from snowfall alone

It may be tempting to calculate:

new snow depth =
previous snow depth + snowfall

but this ignores:

  • Melting
  • Compaction
  • Sublimation
  • Rain
  • Wind redistribution
  • Surface conditions

Use the returned snowdepth field when you need snow on the ground.

Using snow data in Excel or other analytical tools

For spreadsheet analysis, a useful dataset might contain:

datetime
tempmax
tempmin
precip
preciptype
snow
snowdepth

You can then calculate values such as:

total snowfall
maximum snow depth
number of snow days
number of days with snow on ground
largest snowfall event

For example:

snow day =
snow > 0

or:

snow-cover day =
snowdepth > 0

These are different metrics and should be named clearly.

Common snow analysis mistakes

Mistake 1 – Using precipitation as snowfall

precip is liquid-equivalent precipitation.

Use snow for snowfall depth.

Mistake 2 – Using snowfall as snow depth

snow describes new snowfall.

snowdepth describes snow on the ground.

Mistake 3 – Adding snow depth values together

Snow depth is not accumulated across periods.

Mistake 4 – Assuming snowfall is measured every hour

Snow observation frequency can vary.

Hourly output does not necessarily mean that a physical observer measured snowfall every hour.

Mistake 5 – Ignoring elevation

Mountain snowfall can change dramatically over short distances.

Mistake 6 – Ignoring mixed precipitation

Use preciptype to distinguish snow from rain, freezing rain, and ice.

Mistake 7 – Assuming a fixed snow-to-liquid ratio

Snow density varies considerably between storms.

Which fields should I request?

For most snow analysis, start with:

datetime
temp
precip
preciptype
snow
snowdepth

For more detailed winter-weather analysis, add:

feelslike
humidity
windspeed
windgust
visibility
conditions

For example:

elements=datetime,temp,feelslike,precip,preciptype,snow,snowdepth,windspeed,windgust,visibility,conditions

This gives a more complete picture of the event.

Data quality and historical updates

Recent historical weather can change as additional observations become available or upstream data providers make corrections.

This can also apply to snow observations.

If you are analyzing a very recent snowstorm, repeating the query later may return updated information as observation feeds become more complete.

For details, see How Historical Weather Data Is Updated.

Summary

Accurate snow analysis requires understanding several different measurements.

The most important fields are:

snow

New snowfall during the period.

snowdepth

Snow currently on the ground.

precip

Liquid-equivalent precipitation.

preciptype

Whether precipitation occurred as rain, snow, freezing rain, ice, or a combination.

When evaluating snow data:

  1. Use snow for new snowfall.
  2. Use snowdepth for snow on the ground.
  3. Use precip for liquid-equivalent precipitation.
  4. Use preciptype to identify precipitation form.
  5. Consider temperature, wind, elevation, and observation availability.
  6. Use hourly data when timing matters.
  7. Remember that snow observations can be less frequent and more spatially variable than many other weather variables.

For complete field definitions, see the Weather Data Documentation.

For Weather API requests, see the Timeline Weather API documentation.