Nystrom et al. (2025) A Hybrid Four‐Dimensional Variational Data Assimilation System for the Model for Prediction Across Scales (MPAS‐Atmosphere): Leveraging the Joint Effort for Data Assimilation Integration (JEDI)

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Journal of Advances in Modeling Earth Systems
• Year: 2025
• Date: 2025-11-26
• Authors: Robert G. Nystrom, Chris Snyder, Zhiquan Liu, Byoung‐Joo Jung, Jihee Ban, Ivette Hernández Baños
• DOI: 10.1029/2025ms005183

Research Groups

Not specified in the provided abstract.

Short Summary

This study presents and evaluates a global Four-Dimensional Ensemble Variational (4DEnVar) data assimilation system for the Atmospheric component of the Model for Prediction Across Scales (MPAS-A) using the Joint Effort for Data assimilation Integration (JEDI), demonstrating improved meteorological and precipitation forecasts, especially with Hybrid-4DEnVar and all-sky assimilation.

Objective

• To develop and evaluate a global 4DEnVar data assimilation system for MPAS-A using JEDI, assessing its performance against 3DEnVar and the benefits of Hybrid-4DEnVar and all-sky assimilation for improving meteorological and precipitation forecasts.

Study Configuration

• Spatial Scale: Global; analysis resolution of 30 km, ensemble run resolution of 60 km.
• Temporal Scale: Month-long continuous cycling data assimilation experiments; extended forecasts.

Methodology and Data

• Models used: Model for Prediction Across Scales - Atmospheric component (MPAS-A), Joint Effort for Data assimilation Integration (JEDI), 4DEnVar, 3DEnVar, Hybrid-4DEnVar.
• Data sources: All-sky Advanced Microwave Sounding Unit-A (AMSU-A) radiance observations.

Main Results

• Dual-resolution cycling experiments (30 km analysis, 60 km ensemble) performed well, reducing computational cost.
• 4DEnVar updates showed lower mean errors in both observation and model space compared to 3DEnVar experiments during month-long global cycling.
• Hybrid-4DEnVar, which combines flow-dependent ensemble covariance and static climatological covariance, further improved performance over 4DEnVar.
• Assimilating all-sky AMSU-A radiance observations led to additional improvements.
• Extended forecasts initialized from Hybrid-4DEnVar analyses showed improvements in both traditional meteorological fields and precipitation compared to Hybrid-3DEnVar analyses.
• Improvements in precipitation forecasts from 4D methods were most significant in the Southern Hemisphere, consistent with the largest improvements in other meteorological fields.
• Significant improvements in tropical precipitation forecasts were observed in both 3D and 4D experiments assimilating all-sky AMSU-A radiance observations.
• MPAS-JEDI, an open-source community tool, provides 4DEnVar and Hybrid-4DEnVar capabilities and performs well in continuous global cycling experiments.

Contributions

• Presentation and evaluation of a global 4DEnVar data assimilation system for MPAS-A using JEDI.
• Demonstration of the effectiveness and computational efficiency of dual-resolution cycling for data assimilation.
• Quantification of improved forecast accuracy (lower mean errors) of 4DEnVar over 3DEnVar.
• Demonstration of further forecast improvements through the use of Hybrid-4DEnVar and all-sky AMSU-A radiance assimilation.
• Identification of specific regions (Southern Hemisphere, tropics) where 4D methods and all-sky assimilation significantly improve precipitation forecasts.
• Release of MPAS-JEDI as an open-source community-developed tool, making 4DEnVar and Hybrid-4DEnVar capabilities accessible.

Funding

Not specified in the provided abstract.

Citation

@article{Nystrom2025Hybrid,
  author = {Nystrom, Robert G. and Snyder, Chris and Liu, Zhiquan and Jung, Byoung‐Joo and Ban, Jihee and Baños, Ivette Hernández},
  title = {A Hybrid Four‐Dimensional Variational Data Assimilation System for the Model for Prediction Across Scales (MPAS‐Atmosphere): Leveraging the Joint Effort for Data Assimilation Integration (JEDI)},
  journal = {Journal of Advances in Modeling Earth Systems},
  year = {2025},
  doi = {10.1029/2025ms005183},
  url = {https://doi.org/10.1029/2025ms005183}
}