GDPS

1. Overview of GDPS v9.0.0

  • Implemented on June 11, 2024, at the Canadian Meteorological Centre (CMC).

  • Major technical improvements in data assimilation, minor updates to atmospheric and ocean-ice forecasting.

  • Provides high-resolution global forecasts at 15 km resolution with 84 vertical levels.

  • Runs four times daily (00, 06, 12, and 18 UTC).

  • Uses a fully ensemble-derived 4DEnVar assimilation scheme, making forecasts more accurate.

  • Supports global applications, including extreme weather prediction, climate monitoring, and environmental modeling.

2. Atmospheric Data Assimilation & Objective Analysis

The data assimilation process ensures real-time atmospheric observations are incorporated into the model to improve predictions.

2.1 Key Features

  • Model: Version 5.2 of the Global Environmental Multiscale (GEM) model.

  • Assimilation Method: Uses a Four-Dimensional Ensemble-Variational (4DEnVar) approach.

    • Combines variational data assimilation with ensemble-based background error covariances.

    • Hourly analysis increments improve forecast accuracy.

    • Observations are assimilated in 15-minute bins over a 6-hour window.

    • Uses Radiative Transfer for TOVS (RTTOV v13) for processing satellite radiance data.

    • Uses 25 km resolution ensembles from GEPS 8.0.0 for background error calculations.

2.2 Data Sources & Observations Assimilated

GDPS integrates data from various sources to refine its global forecasts:

  1. Satellite Radiance Data

    • Microwave Sensors: AMSU-A, MHS, ATMS, SSMIS, MWHS-2.

    • Infrared Sensors: AIRS, IASI, CrIS.

    • Geostationary Imagers.

  2. Ground-Based & Other Observations

    • Radiosondes (weather balloons)

    • Surface stations (SYNOP, METAR, BUOY, SHIP)

    • Aircraft-based weather observations

    • GPS-based refractivity data

    • Atmospheric Motion Vectors (AMVs) from satellites

    • Scatterometer winds for oceanic forecasting

2.3 Quality Control & Bias Corrections

  • Satellite Radiance Bias Correction

    • Observational bias is corrected based on last 7 days of analysis data.

    • Uses inter-channel correlation error matrix for radiance data processing.

    • Dynamic bias correction for GPS-based refractivity observations.

  • Other Enhancements

    • Improved use of radiosonde BUFR data.

    • Background-error covariances updated using a scale-dependent localization approach.

    • Incremental 3D-Var assimilation for ozone predictions.

3. Forecast Model Configuration

3.1 Key Features

  • Model: GEM v5.2

  • Numerical Grid: Global coverage at 15 km resolution.

  • 84 vertical levels, with top level at 0.1 hPa.

  • Iterative-implicit semi-Lagrangian scheme with a 450-second timestep.

  • Prognostic Variables Include:

    • 3D winds, temperature, surface pressure, cloud condensate, ozone.

    • Precipitation, cloud cover, humidity, UV indices.

3.2 Physical Parameterizations

  • Deep Convection: Updated Kain & Fritsch scheme for high-resolution forecasting.

  • Shallow Convection: Bechtold mass-flux scheme.

  • Microphysics: Two-moment cloud microphysics scheme.

  • Boundary Layer Mixing: Based on Turbulent Kinetic Energy (TKE).

3.3 Land Surface & Hydrology

  • Uses a four-layer surface model for land, water, sea ice, and glacier.

  • Surface variables include soil temperature, snow depth, sea ice thickness.

  • Coupled with ocean models for dynamic sea surface temperature updates.

4. Ocean & Sea Ice Prediction System

GDPS includes a coupled ocean-sea ice prediction component, which enhances climate and marine forecasts.

4.1 Key Features

  • Ocean Model: NEMO 3.6 (Primitive Equations)

  • Sea Ice Model: CICE 6.2.0

  • Resolution: ¼ degree (~25 km horizontal resolution) with 50 vertical levels.

  • Uses updated turbulence mixing models for accurate ocean forecasts.

  • Assimilates global sea ice and ocean observations (AVISO, CLS, CMC SST Analysis).

  • Sea ice thickness and fraction dynamically computed.

5. Computational Performance

  • Runs on 3,600 cores, completing analysis in ~18 minutes.

  • Uses a nested-grid strategy for high-resolution sub-regional forecasts.

  • Parallel computing optimizations reduce processing time.

6. Applications and Use Cases

GDPS supports global decision-making across multiple domains:

  • Extreme Weather Forecasting: Hurricanes, heatwaves, cold outbreaks.

  • Climate Monitoring: Long-term trends in temperature, precipitation, and ocean currents.

  • Air Quality & Environmental Prediction: Predicts air pollutants and wildfire smoke movement.

  • Energy & Water Management: Supports renewable energy integration by forecasting wind, solar, and hydrological conditions.

  • Agriculture & Food Security: Provides early warnings for droughts, frosts, and heat stress on crops.

7. Summary & Importance

The GDPS v9.0.0 is a state-of-the-art global weather prediction system that significantly improves forecasting for extreme weather events, environmental risks, and climate change adaptation. With advanced data assimilation techniques, high-performance computing, and an integrated ocean-sea ice model, it serves as a critical tool for scientists, policymakers, and emergency responders worldwide.

8. References & Further Reading

Conclusion

GDPS v9.0.0 represents a major leap in global weather forecasting capabilities. Its ability to integrate real-time observational data and produce highly accurate multi-day forecasts makes it a powerful tool for climate resilience, disaster preparedness, and sustainable resource management.

PreviousNumerical: DeterministicNextRDPS

Last updated

Was this helpful?