General Circulation and Earth System Models

The comprehensive numerical models that simulate the coupled atmosphere, ocean, land, and ice, increasingly extended to the carbon cycle and biogeochemistry.

Definition

A general circulation model simulates the large-scale circulation and climate of the atmosphere and ocean by numerically solving the governing physical equations on a global grid, and an Earth system model extends this by interactively including biogeochemical components such as the carbon cycle.

Scope

This topic covers the most comprehensive climate models: atmosphere-ocean general circulation models that solve the equations of fluid motion and thermodynamics on a global grid, and Earth system models that add interactive carbon, chemistry, vegetation, and ice-sheet components. It treats their dynamical cores, the coupling of components, the parameterization of subgrid processes, the trade-offs of resolution and complexity, and the computational demands of running them.

Core questions

• How do general circulation models solve the equations of the atmosphere and ocean?
• How are the components coupled into a single Earth system model?
• What is gained and lost by adding complexity versus resolution?
• What computational and structural choices shape model behavior?

Key theories

Primitive-equation dynamical core

General circulation models integrate the primitive equations, simplified forms of the fluid-dynamical and thermodynamic laws, to simulate the evolving three-dimensional circulation of the atmosphere and ocean.

Coupled Earth system representation

Earth system models couple physical climate components with interactive carbon, chemistry, and vegetation so that feedbacks among them, such as carbon-cycle feedbacks, emerge from the simulation.

Mechanisms

A dynamical core advances the primitive equations on a global grid to compute winds, temperatures, and currents, while physics modules parameterize radiation, clouds, convection, and surface exchanges. The atmosphere, ocean, sea ice, and land components are coupled so they exchange fluxes of energy, water, and momentum, and Earth system models additionally simulate carbon, chemistry, and vegetation so biogeochemical feedbacks arise interactively, all at substantial computational cost.

Clinical relevance

These models are the workhorses for projecting future climate, simulating past climates, and running the coordinated experiments that underpin the IPCC assessments and national climate planning.

History

The first general circulation models emerged at institutions such as Princeton's Geophysical Fluid Dynamics Laboratory in the 1960s, Manabe and Wetherald produced the first three-dimensional carbon-dioxide-doubling experiment in 1975, and successive decades added coupled oceans, sea ice, and eventually the interactive carbon cycle of modern Earth system models.

Debates

Resolution versus complexity

Whether limited computing is better spent on higher resolution to resolve clouds and eddies or on adding Earth system components is an ongoing strategic debate in model development.

Key figures

• Syukuro Manabe
• Warren Washington
• Akio Arakawa
• Joseph Smagorinsky

Seminal works

• manabewetherald1975
• mcguffie2014

Frequently asked questions

What is the difference between a GCM and an Earth system model?

A general circulation model simulates the physical atmosphere and ocean, while an Earth system model adds interactive components such as the carbon cycle, chemistry, and vegetation.

Why do climate models need supercomputers?

They solve the physical equations at millions of grid points and many time steps over long simulations, which requires enormous computing power, especially at high resolution.