Ambee Global Weather Data - Historical | Present | Forecast

**Overview**

Leveraging the best global weather models, including ECMWF, ERA5, CFSR, GFS, and HRRR, combined with Ambee's proprietary models, and augmenting with inputs such as satellite, radar, and observation stations, Ambee's weather data is a homogenized continuum of data that flows seamlessly from the year 1995 to present and into the forecast with forecast views ranging from one hour all the way out to 46 days. All data is available in hourly and daily format and can be mapped to any global location, including a precise lat/lon or any geographically bounded area such as ZIP or Postal Code, DMA, or MSA.

**Use cases**

- Programmatic advertising
- Demand forecasting for weather-sensitive products and services
- Supply chain optimization and logistics planning
- Pharmaceutical demand planning for temperature-sensitive shipments
- Rotation and inventory management for seasonal products
- ESG reporting and climate-related financial risk modeling
- Agricultural optimization and crop yield prediction
- Weather-based automobile routing and navigation
- Insurance risk assessment and catastrophe modeling

**Sample fields**

- temperature: Air temperature measurements in degrees Fahrenheit
- apparentTemperature: Feels-like temperature accounting for humidity and wind effects
- dewPoint: Temperature at which air becomes saturated with water vapor
- humidity: Relative humidity percentage (0-100%)
- pressure: Sea level atmospheric pressure
- windSpeed: Average wind speed
- windBearing: Wind direction in degrees
- windGust: Peak wind gust measurements
- cloudCover: Fraction of sky covered by clouds (0-1 scale)
- visibility: Distance one can see clearly
- uvIndex: Ultraviolet radiation intensity index
- precipIntensity: Precipitation intensity
- precipType: Type of precipitation (rain, snow, etc.)
- ozone: Ozone concentration in Dobson Units

**Tables included**

- Weather_1hr_current_data
- Weather_historical_data
- Weather_historical_data_daily
- Weather_forecast_48hr_data
- Weather_forecast_120hr_data
- Weather_forecast_daily_30days

**Benefits to Databricks Marketplace users**

- Instantly available: Delivered directly for immediate integration into any analytics, machine learning, or Databricks workflow.
- Global coverage: Available for any latitude/longitude or geographically bounded area, including ZIP or Postal Code, DMA, and MSA.
- Sequential time series for predictive forecasting: A homogenized continuum of historical and forecast data, enabling consistent model training, climate trend analysis, and operational forecasting.
- 20+ core parameters: Temperature, feels like temperature, heat index, humidity, dew point, wind speed and direction, precipitation, cloud cover, visibility, solar radiation, UV index, pressure, and more.
- Unparalleled quality: Robust quality controls leveraging ground-based truth data, companion datasets, climatological context, cross-station substitution, and meteorological modeling to address gaps, correct anomalies, and maintain data integrity
- Built for advanced analytics: Compatible with Spark, MLflow, and other Databricks-native tools to accelerate feature engineering, model development, and deployment.