Evaluating the Hadley Centre’s HadGEM1 Coupled Climate Model: Physics Improvements, Skill Assessment, and Climate Sensitivity
Observation and Context Coupled atmosphere–ocean general circulation models (AOGCMs) are essential tools for predicting future global climate change. However, earlier global models like HadCM3 relied on lower spatial resolution, parameterized virtual salt fluxes instead of true freshwater exchange, and lacked interactive representations of atmospheric aerosols. These limitations produced regional biases, particularly in ocean currents, cloud formations, and sea ice behavior.
Hypothesis Scientists hypothesized that developing a new coupled model (HadGEM1) using higher horizontal and vertical resolutions, a nonhydrostatic “New Dynamics” atmospheric core, interactive aerosol schemes, and multi-category sea ice dynamics would substantially improve the simulation of Earth’s mean climate without requiring artificial flux adjustments.
Experiment and Methodology The Met Office Hadley Centre developed HadGEM1 and compared it directly against observed climatologies and HadCM3:
- Model Configuration: HadGEM1 features an atmospheric grid of 1.25∘ latitude×1.875∘ longitude with 38 vertical levels. Its ocean component uses a 1∘×1∘ grid that sharpens meridionally to 1/3∘ at the equator, with 40 vertical levels.
- Key Improvements: The model incorporates the “New Dynamics” semi-Lagrangian dynamical core, interactive aerosols (sulfate, black carbon, biomass smoke, and sea salt) with both direct and indirect cloud effects, an explicit free-surface freshwater ocean scheme, and the Los Alamos CICE sea ice scheme resolving five thickness categories using elastic-viscous-plastic (EVP) dynamics.
- Testing: The team carried out a multi-century control simulation exceeding 400 years under constant pre-industrial 1860 conditions, an idealized 1%-per-year transient CO2 increase run, and slab ocean equilibrium experiments.
Results and Data
- Overall Model Skill: Evaluated across multiple variables using a weighted Climate Prediction Index (CPI), HadGEM1 scored 2.884 compared to HadCM3’s 3.054 (where lower scores indicate greater skill), driven largely by major reductions in cloud errors.
- Ocean and Sea Ice: The higher ocean resolution successfully opened narrow straits (such as the Makassar Strait Indonesian Throughflow). Arctic sea ice distribution improved substantially, with the thickest ice properly banking against Greenland and the Canadian Archipelago rather than being misplaced in the Beaufort Gyre.
- Remaining Deficiencies: HadGEM1 developed an overly strong equatorial trade wind stress and an exaggerated cold bias in the tropical Pacific. Consequently, El Niño–Southern Oscillation (ENSO) variability was too weak (Niño-3 standard deviation was 0.69 K vs. 0.80 K observed) and its seasonal cycle was poorly phase-locked.
Conclusion and Climate Impact HadGEM1 proved to be a significantly more physically realistic and skillful coupled climate model than HadCM3, demonstrating the benefits of higher grid resolution, dynamic sea ice modeling, and interactive aerosols.
- Climate Impact: Under a doubling of atmospheric CO2, HadGEM1 produced an effective climate sensitivity of 2.8 K and a transient sensitivity of 1.8 K, closely aligning with HadCM3 (3.1 K and 2.0 K, respectively). Despite similar global sensitivity, HadGEM1 revealed distinct regional warming patterns due to improved marine stratocumulus cloud feedbacks. These findings provided essential projections for the IPCC Fourth Assessment Report (AR4) and established a foundation for future Earth System modeling.
Full Citation Johns, T. C., Durman, C. F., Banks, H. T., Roberts, M. J., McLaren, A. J., Ridley, J. K., Senior, C. A., Williams, K. D., Jones, A., Rickard, G. J., Cusack, S., Ingram, W. J., Crucifix, M., Sexton, D. M. H., Joshi, M. M., Dong, B.-W., Spencer, H., Hill, R. S. R., Gregory, J. M., Keen, A. B., Pardaens, A. K., Lowe, J. A., Bodas-Salcedo, A., Stark, S., & Searl, Y. (2006). The new Hadley Centre Climate Model (HadGEM1): Evaluation of coupled simulations. Journal of Climate, 19(7), 1327–1353. https://doi.org/10.1175/JCLI3712.1