The Visual Crossing Weather Data platform provides the ability to easily access weather data sets including weather forecast data, weather history data and historical weather summary data. The data is available for download through the Weather Data Download page and is also available through the Weather Data API.
This document explains the weather data variables that are available for the weather forecast and historical weather record datasets.
General weather data structure
We aim to keep the data set structure as consistent as possible between the three main categories of data – weather forecast, historical weather observations and historical weather summaries. This allows you to easily switch between using data for weather forecast and historical records within the same data and can be useful in many applications. All the data sets include the following characteristics:
Data Set Structure
All data sets are a simple table structure which is available in multiple formats including plain text delimited such as CSV, JSON and ODATA. The data includes a single row of headers information indicating the information in the column. These column headers will either be in a human readable header that may be translated or in a single, short form which is not translated. The short form is typically used for Weather API usage to make it easier to consume the data. When using the Weather API, the shortColumnNames parameters indicates whether or not short column names will be used.
Hourly and Daily Dataset Aggregation
All data within the a system is calculated at the hourly data level for accuracy and consistency. When requests are made for longer time periods, such as daily weather forecast or monthly historical summaries, the hourly data is aggregated using average (mean), sum, minimum or maximum functions. The individual column documentation below explains the aggregation method used.
The units of the data are specified by the overal unit group requests such as metric, US units etc. This unit group defines the unit of measure used for each column type – all columns of the same variable type will have the same unit. For example, celcius will be used for all temperature related quantities when the unit group is set to metric.
Data sets will typically include information about the location, period or data time. If the requested data includes address, the address information is included in the return value. In addition, the latitude and longitude of the requested locations are returned using decimals degrees. The longitude value are zero based at the prime meridian. Values west of the prime meridian are negative and east are positive.
For more information on how dates and times are returned in the Weather API results, see Dates and Times in the Weather API.
Empty values (or null values) are used within the data set to indicate an absence of data, such as missing weather information or unknown data. They are not used to indicate a zero value. For example, an unknown precipitation value will be marked as empty or null. A zero amount of precipitation will be indicated by the value zero.
Weather element availability
We aim to include all weather elements such as temperature, wind, precipitation etc for all locations. Unfortunately not all locations report all weather elements. In those cases the columns will be included in the data but include null values. We continue to expand the availabilty of weather elements as we add additional data sets and data processing.
Core weather columns
These columns are common to most weather data sets including forecast, historical observations and historical summaries. The API setting of ‘shortColumnNames’ when set to true will utilize the column names found below in parenthesis. (&shortColumnNames=true)
Temperature (temp, mint, maxt)
There are three main columns for indicating temperature – Minimum Temperature, Maximum Temperature and Temperature. These will return data in Fahrenheit, Celsius or Kelvin depending on the unit group. The three temperature columns indicate the minimum, maximum and average (mean) temperature for the time period in question. For example, if you request daily data then you will receive the data for minimum, maximum and mean temperature for each day. If you request data for the shortest available timeframe for that request (typically one hour), the minimum, maximum and mean are returned with the same value. All data sets will include all three columns even if the values are the same to ease consumption of the data sets and keep data set consistency.
Wind (wspd, wdir, wgust)
Wind data includes wind speed, wind gust and wind direction.
Wind speed and wind direction (wspd, wdir)
The wind speed and wind direction indicate the mean wind speed for the location and time period requested. The units of the data will be in miles per hour, kilometers per hour or m/s depending on the unit group.
Wind gust (wgust)
Wind gust is the maximum wind speed measures over a short amount of time (typically less than 20 seconds). Note that a wind gust requires the measured short term wind speed to be significantly more than mean wind speed. Typically, the wind speed should be 10 knots more (11mph or 18kph). When the wind gust does not meet these criteria, a null/empty value is returned.
Precipitation (precip, pop)
Precipitation is the sum of the amount of liquid equivalent rainfall, snow or other precipitation that fell is predicted to fall in the period. Three output variables cover precipitation:
The amount of precipitation that fell or is predicted to fall in the specified time period. The values are indicated in inches or mm.
Chance of precipitation (pop)
The possibility of precipitation for the given time period expressed as a percentage change from 0-100%. This applies to the forecast queries only.
This is the proportion of time for which measurable precipitation was record during the time period, expressed as a percentage. For example, if within a 24 hour day there are six hours of measurable rainfall, the precipitation coverage is 25% (6/24*100). This information is only available for historical weather observation data and historical summaries. Note that where as trace precipitation is considered zero value for the precipitation column, a trace precipitation report will be considered as precipitation for the precipitation coverage column. Therefore it is possible to observe zero reported precipitation with a non-zero precipitation coverage.
Cloud cover (cloudcover)
Cloud cover is the amount of sky that is covered by cloud expressed as a percentage. The cloud coverage is for all altitudes.
Snow and Snow Depth (snow, snowdepth)
Snow is the amount of new snow that has fallen in the time period. Snow depth is the average amount of snow currently on the ground for the time period. Snow depth will increase with additional snowfall and decrease with melting and compaction. Both snow and snow depth are expressed in inches or millimeters.
Relative Humidity, Heat Index and Wind Chill (humidity, heatindex, windchill)
Relative humidity (humidity)
Relative humidity the amount of water vapor present in the air compared the maximum amount possible for a given temperature, expressed as a mean percentage. Human comfort levels are typically found between 30-70%. Values higher than 70% are considered humid. Values lower than 30% are considered dry.
Heat Index (heatindex)
Heat Index is a measure of how hot it feels combining the actual air temperature with the relative humidity. High humidity is associated with discomfort by making the apparent temperature hotter than it would otherwise feel. Values are expressed in the same units as temperature. Heat index values are only calculated when the temperature is greater than 80F (about 26.7C) and the relative humidity is greater than 40%. An empty value is returned outside of these ranges.
Wind chill (windchill)
Wind chill is the measure of how cold it feels combining the actual air temperature with the wind speed. Wind makes temperatures feel colder than when the air is still. Values are expressed in the same units as temperature. Wind chill values are only calculated when the temperature is less than 50F (about 10C) and the wind speed is greater than 3mph (5kph). An empty value is returned outside of these ranges.
Visibility is the distance than can be seen in daylight. This account of weather phenomenon such as haze, mist, fog or smoke. The distance is expressed in miles or kilometers.
Sea Level Pressure (sealevelpressure)
The atmospheric pressure at a location that removes reduction in pressure due to the altitude of the location. This is expressed in millibars.
Weather Type (conditions)
Notable weather conditions reported at a particular location such as any thunderstorms, rainfall etc. The availability of this data is dependent on the weather station that observed the information and is not reported by all weather stations – the absence of a particular weather type should not be considered evidence that the weather type did not occur.
A general information column that may include text based notes such as absence of data for a particular weather station.
For historical observations, indicates the weather stations that were used for creating the observation. This is a list of weather station names, their ID within the Visual Crossing weather data platform and the distance from the requested location to the weather station. For more information on how weather station data is combined, please read more.
Historical Weather Summary Daily Statistics
Historical weather summaries include above weather data columns. In addition, Historical Weather Summary data sets can optionally include detailed daily statistics which provides additional information about the daily values within the data set. This helps identify the typical weather that was experience in the time period rather than extremes.
The daily statistics analyzes the values across the days including also the maximum, minimum and mean and other distribution statistics for each day.
For example if we would like to understand what is the typical daily high temperature for a particular location we can use the the daily statistics. The daily statistics will include the maximum daily high temperature, the mean daily high temperature and the minimum daily high temperature. The maximum daily high temperature will match the maximum high temperature column found in the summary data set. The average maximum temperature will better match our interest for the typical high temperature. The maximum and minimum values highlight the extremes.