Modelling the Arctic: How Climate Scientists Simulate the Polar Future

Climate models — mathematical representations of the Earth's atmosphere, ocean, land surface, and ice — are the primary tools through which scientists project future Arctic change. The latest generation of models, participating in the Coupled Model Intercomparison Project Phase 6 (CMIP6), include increasingly sophisticated representations of sea ice physics, ocean-ice interactions, and polar atmospheric dynamics. They are also confronting an uncomfortable reality: previous generations of models systematically underestimated the rate of Arctic warming that has actually occurred — and the reasons for this underestimation are informing improvements in the current generation.

How Climate Models Work

A global climate model divides the Earth's surface into a three-dimensional grid of cells — typically 50-100 kilometres horizontally and with multiple layers vertically through the atmosphere and ocean. For each cell, at each time step, the model calculates exchanges of heat, moisture, momentum, and chemical species with adjacent cells, using equations derived from physical principles. The model is initialised with observed conditions and then run forward in time, driven by specified greenhouse gas concentrations. Arctic-specific components include sea ice thermodynamics (calculating how ice grows and melts), ice dynamics (calculating how ice moves and deforms under wind and ocean stress), and snow-ice-albedo feedbacks.

Why Models Underestimated Arctic Warming

Observations consistently show that the Arctic has warmed 2-4 times faster than the global average — a phenomenon called Arctic amplification. Most previous-generation climate models reproduced Arctic amplification but underestimated its magnitude. Research has identified several contributing factors: insufficient representation of cloud-radiation feedbacks in the Arctic atmosphere, underestimation of the ice-albedo feedback strength, poor representation of vertical mixing in the Arctic Ocean (which affects how much heat from the Atlantic reaches the surface), and inadequate spatial resolution to capture small-scale dynamical processes that significantly affect the regional energy balance.