Skip to content

WeatherForecastSeries

WeatherForecastSeries Model

The WeatherForecastSeries Smart Data Model is an extension of the official WeatherForecast model from FIWARE’s Smart Data Models initiative.
Its main goal is to enable the representation of forecast time series (rather than a single prediction) within a single entity.

Key Modifications

  • Array-based attributes
    Most of the variable types defined in the original WeatherForecast model have been converted into arrays of their base type.
    This structure allows the model to store multiple forecasted values (e.g., temperature, wind speed, precipitation) corresponding to different timestamps for a single point.

  • Inlined definitions
    Some variables originally referenced from Weather-Commons have now been defined directly inside the WeatherForecastSeries schema, improving self-containment and making the model easier to use independently.

  • New timestamps attribute
    A new property, timestamps, has been introduced to specify the exact prediction times corresponding to each value in the variable arrays.
    This makes it possible to directly associate each forecasted parameter with its valid time, simplifying temporal analysis.

Purpose and Advantages

This model enables the aggregation of complete forecast time series (e.g., hourly or daily predictions) in a single JSON object, reducing redundancy and improving data retrieval efficiency.

In the PGTEC project, it is used to:

  • Store weather predictions retrieved from different data sources such as AEMET, Open-Meteo, and Copernicus.
  • Provide these standardized forecasts as input for the TETIS hydrological model, ensuring that all variables share a common format and temporal alignment.
  • Facilitate interoperability between APIs, Airflow workflows, and visualization dashboards.

Example Entity

Below is an example of a WeatherForecastSeries entity representing 3-hourly weather predictions for a given location:

{
  "id": "WeatherForecastSeries:Valencia:2025-03-01",
  "type": "WeatherForecastSeries",
  "location": {
    "type": "Point",
    "coordinates": [-0.3763, 39.4699]
  },
  "timestamps": [
    "2025-03-01T00:00:00Z",
    "2025-03-01T03:00:00Z",
    "2025-03-01T06:00:00Z"
  ],
  "temperature": [14.5, 13.8, 13.1],
  "precipitation": [0.0, 0.2, 1.1],
  "windSpeed": [3.4, 2.8, 4.1],
  "relativeHumidity": [65, 70, 72],
  "source": "AEMET",
  "address": {
    "addressLocality": "Valencia",
    "addressCountry": "ES"
  },
  "dateIssued": "2025-03-01T00:00:00Z",
  "validFrom": "2025-03-01T00:00:00Z",
  "validTo": "2025-03-01T06:00:00Z"
}