Li et al. (2025) Improved Madden–Julian oscillation simulation using the modified moist physical parameterizations for a global climate model

Identification

• Journal: Climate Dynamics
• Year: 2025
• Date: 2025-12-12
• Authors: Xiaohan Li, Yanluan Lin, Xiao Zhou, Yiran Peng, Xiaomeng Huang
• DOI: 10.1007/s00382-025-07834-1

Research Groups

• State Key Laboratory of Climate System Prediction and Risk Management/Earth System Modeling Center, Nanjing University of Information Science and Technology, Nanjing, China
• Department of Earth System Science, Ministry of Education Key Laboratory for Earth System Modeling, Institute for Global Change Studies, Tsinghua University, Beijing, China

Short Summary

This study investigates the impact of modified moist physical parameterizations on the global climate model, CIESM, regarding Madden–Julian oscillation (MJO) simulation. The new moist physical schemes significantly improve MJO propagation characteristics and mean-state fidelity by enhancing zonal asymmetry in moist static energy tendency and optimizing cloud-convection interactions.

Objective

• To investigate the impact of modified moist physical parameterizations (Gauss-PDF cloud macrophysical, Single-ice microphysical, and modified ZM convective scheme with convective microphysics and stochastic process) on Madden–Julian oscillation (MJO) simulation in the CIESM global climate model.
• To identify the physical mechanisms through which these modifications improve MJO propagation and mean climate state without degrading mean-state fidelity.

Study Configuration

• Spatial Scale: Global climate model (CIESM) with approximately 1° resolution and 30 hybrid levels up to 1 hPa. Analysis focused on tropical regions (e.g., 15° S–15° N, 30° E–150° W for precipitation frequency; 10° S–10° N for MJO spectra and moist static energy analysis), including the South Pacific Convergence Zone (SPCZ), Intertropical Convergence Zone (ITCZ), equatorial Indian Ocean, Maritime Continent, and western Pacific.
• Temporal Scale: Boreal winter (November–April) for mean state and MJO analysis. 15-year CMIP-type (piControl) simulations were conducted, with the last 10 years used for evaluation. MJO-related signals were extracted using a Lanczos band-pass filter for periods of 20–90 days.

Methodology and Data

• Models used:
  • Community Integrated Earth System Model (CIESM), developed based on the Community Earth System Model v1.2.1 (NCAR-CESM1.2.1) and Community Atmospheric Model Version 5 (CAM5).
  • Modified moist physical parameterizations in CIESM:
    • Gauss-PDF cloud macrophysical scheme (replaces CAM5's Park et al. (2014) scheme).
    • Single-ice microphysical scheme (replaces CAM5's Morrison and Gettelman (2008) scheme, combines cloud ice and snow into a single prognostic variable).
    • Modified Zhang and McFarlane (ZM) convective scheme (incorporates two-moment microphysical processes within deep convective clouds and the Plant and Craig (2008) stochastic process).
  • Control run (CTLCAM5) used default CAM5 physical parameterizations (original ZM convective scheme, UW turbulence and shallow convection, Park14 macrophysics, and MG08 microphysics).
  • Intermediate experiments: EXPstratus (Gauss-PDF and Single-ice schemes), EXPconvmicrop (EXPstratus + convective microphysics), EXPCIESM (EXPconvmicrop + stochastic process).
• Data sources:
  • Reanalysis: ERA-5 (horizontal wind speed, temperature, vertical pressure velocity, specific humidity, and geopotential height).
  • Satellite/Observation: Integrated Multi-satellite Retrievals for GPM (precipitation), Hadley-National Oceanic and Atmospheric Administration/OI merged data product (sea surface temperature).
  • Prescribed greenhouse gas concentrations, aerosols, and other external forcings followed CMIP6 design.

Main Results

• Mean State Improvements:
  • The modified moist physical schemes preserve mean-state fidelity, reducing mean precipitation biases over the South Pacific Convergence Zone (SPCZ) and mitigating the double Intertropical Convergence Zone (ITCZ) problem.
  • Global mean precipitation for EXPCIESM was 3.43 × 10⁻⁸ m/s, closer to the observed 3.07 × 10⁻⁸ m/s, compared to CTLCAM5's 3.55 × 10⁻⁸ m/s.
  • CTLCAM5 exhibited a "high frequency and low intensity" precipitation bias; EXP runs reduced the frequency of weak convective precipitation (< 3.47 × 10⁻⁷ m/s) and increased large-scale precipitation.
  • EXPCIESM demonstrated a more realistic precipitation-humidity relationship, reducing weak precipitation in dry regimes and increasing intense precipitation in high humidity regimes.
• MJO Simulation Improvements:
  • EXP_CIESM significantly improved the simulation of MJO's key characteristics, including the wavenumber-frequency spectrum, eastward propagation speed (approximately 5 m/s), and propagation distance across the Maritime Continent to the western Pacific.
  • The pattern correlation coefficient (PCC) for MJO-filtered precipitation propagation improved from 0.743 (CTLCAM5) to 0.917 (EXPCIESM) compared to observations.
  • Enhanced MJO-band squared coherence and realistic phase lag between precipitation and zonal winds (U850, U200) were observed.
  • The eastward-to-westward propagation power ratio (E/W) for precipitation improved from 1.68 (CTLCAM5) to 3.56 (EXPCIESM), and for U200 from 1.05 to 2.45.
• Physical Mechanisms:
  • The new moist physical schemes enhanced zonal asymmetry in the column-integrated moist static energy (MSE) tendency anomalies, with stronger negative anomalies (drying and cooling) to the west of MJO convection.
  • Gauss-PDF macrophysics and Single-ice microphysics strengthened top-heavy latent heating and effectively exported MSE west of the MJO convection.
  • They captured mid-level evaporative cooling anomalies over the western Pacific (~150° E), promoting a broader scale of MJO-related zonal circulation anomalies.
  • Convective microphysics further enhanced top-heavy stratiform heating.
  • The stochastic process in the ZM deep convection reduced artificial weak precipitation in dry environments and increased intense precipitation in high humidity regimes, amplifying MJO precipitation.
  • Improved vertical tilt structures of MJO-related temperature and moisture anomalies, including low-level moistening to the east and drying to the west, were observed.

Contributions

• Demonstrates a successful strategy to improve Madden–Julian oscillation (MJO) simulation in a global climate model (CIESM) without compromising the mean climate state, addressing a long-standing trade-off problem in climate modeling.
• Provides a detailed process-oriented diagnosis, linking specific modifications in moist physical parameterizations (Gauss-PDF macrophysics, Single-ice microphysics, convective microphysics, and stochastic convection) to the observed improvements in MJO propagation dynamics and the alleviation of mean state biases.
• Highlights the critical role of stratiform cloud schemes in facilitating efficient coupling between large-scale condensation heating and low-frequency circulation anomalies through enhanced top-heavy heating.
• Emphasizes the importance of appropriate interaction between deep convection and stratiform clouds, and the role of hydrometeors in convective detrainment, for promoting MJO propagation.
• Quantifies the impact of these parameterization changes on key MJO diagnostics, including moist static energy budget terms, wavenumber-frequency spectra, and propagation characteristics, offering insights for future model development.

Funding

• National Natural Science Foundation of China, Grant Numbers: 42205160 and 42125503.

Citation

@article{Li2025Improved,
  author = {Li, Xiaohan and Lin, Yanluan and Zhou, Xiao and Peng, Yiran and Huang, Xiaomeng},
  title = {Improved Madden–Julian oscillation simulation using the modified moist physical parameterizations for a global climate model},
  journal = {Climate Dynamics},
  year = {2025},
  doi = {10.1007/s00382-025-07834-1},
  url = {https://doi.org/10.1007/s00382-025-07834-1}
}