Which data source is not typically used to initialize numerical weather prediction models?

• A. Seismic data
• B. Surface observations
• C. Radiosondes
• D. Satellites

Answer: (A)

Initializing numerical weather prediction models requires data that describe the atmosphere’s current state—temperature, humidity, wind, and pressure across different levels. Surface observations provide essential surface values, radiosondes lift profiles up through the atmosphere to capture vertical temperature, moisture, and wind structure, and satellites offer broad coverage of radiances and derived fields that help constrain temperature and humidity in regions where in-situ data are sparse. Seismic data, however, measure ground motion from earthquakes and reflect properties of the Earth’s interior and crust, not atmospheric conditions. They don’t provide the atmospheric state needed for model initialization, so they’re not used in this context.