Dark Mode
United States Weather Activity Data
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This dataset offers a countrywide collection of 8.6 million weather events across 49 states in the United States, spanning from January 2016 to December 2022. It captures a diverse range of weather phenomena, including regular occurrences like rain and snow, as well as extreme conditions such as storms and severe cold. The data originates from 2,071 airport-based weather stations nationwide. Each event is defined as a spatiotemporal entity, associated with specific locations and times. Event types covered include Severe-Cold (temperatures below -23.7 degrees Celsius), Fog, Hail, Rain (light to heavy), Snow (light to heavy snowstorms), Storm (wind speeds at least 60 km/h), and Other Precipitation.
Columns
- EventId: A unique identifier for each recorded weather event.
- Type: The specific category of weather event, such as 'Rain' or 'Snow'. This column has 7 unique types, with 'Rain' being the most common (58%) and 'Fog' (23%) also being significant.
- Severity: Describes the intensity or severity of an event where applicable. 'Light' is the most common severity (60%), followed by 'Severe' (20%).
- StartTime(UTC): The precise start time of an event, recorded in the Coordinated Universal Time (UTC) timezone. The data ranges from 1 January 2016 to 1 January 2023.
- EndTime(UTC): The precise end time of an event, recorded in the Coordinated Universal Time (UTC) timezone. This also ranges from 1 January 2016 to 1 January 2023.
- Precipitation(in): The total amount of precipitation, measured in inches, if the data is available for the event. The mean precipitation is 0.09 inches, with a maximum recorded value of 1104.13 inches.
- TimeZone: The US-based timezone corresponding to the event's location (e.g., US/Central, US/Eastern). 'US/Central' is the most frequent timezone (41%).
- AirportCode: The code of the airport station from which the weather event was reported. There are 2,071 unique airport codes.
- LocationLat: The latitude (GPS coordinate) of the airport-based weather station reporting the event. Latitudes range from 24.6 to 48.9, with a mean of 38.8.
- LocationLng: The longitude (GPS coordinate) of the airport-based weather station reporting the event. Longitudes range from -125 to -67.8, with a mean of -91.9.
- City: The city associated with the address record of the airport-based weather station. There are 1,717 unique cities, with 'Jacksonville' being the most common. A small percentage of records have missing city information.
- County: The county associated with the address record of the airport-based weather station. There are 1,100 unique counties.
- State: The state associated with the address record of the airport-based weather station. The dataset covers 48 unique states, with 'TX' (Texas) being the most common (7%), followed by 'MN' (Minnesota) (5%).
- ZipCode: The postcode in the address record of the airport-based weather station. There are a large number of unique postcodes, ranging from 1022 to 99362. A small percentage of records have missing postcode information.
Distribution
This data product is typically distributed as a CSV (Comma Separated Values) file. The dataset size is 1.09 GB and it contains 8.6 million weather events structured across 14 distinct columns. Most columns have 8.63 million valid records, with 'City' having 8.61 million and 'ZipCode' having 8.56 million valid records.
Usage
This dataset is ideal for research and academic applications involving large-scale geo-spatiotemporal data analysis. It can be particularly useful for:
- Short and long-term pattern discovery related to weather phenomena.
- Developing and testing weather prediction models.
- Analysing climate trends and extreme weather event frequencies.
- Geospatial analysis of weather impacts across different US regions.
- Studies on the temporal and spatial distribution of specific weather event types.
Coverage
The dataset provides countrywide coverage across 49 states in the United States. The temporal scope ranges from January 2016 to December 2022, offering seven full years of weather event data. All data is sourced from a network of 2,071 airport-based weather stations located nationwide, providing detailed geo-spatiotemporal information for each event.
License
CC BY-NC-SA 4.0
Who Can Use It
This dataset is exclusively intended for non-commercial, research, or academic applications. Ideal users include:
- Researchers studying meteorology, climatology, or environmental science.
- Academic institutions for educational purposes or scientific projects.
- Data scientists and analysts engaged in spatio-temporal data mining and pattern recognition for weather-related insights.
- Anyone needing a substantial dataset for geo-spatiotemporal modelling and analysis of US weather.
Dataset Name Suggestions
- US Weather Events 2016-2022
- United States Weather Activity Data
- Airport Weather Observations USA
- North American Weather Event Log
- US Climate Event Records
Attributes
Original Data Source: United States Weather Activity Data
QUALITY