Indo-Pacific Control of Climate States and the Overturning Circulation

The School of Earth and Atmospheric Sciences Presents Dr. Jess Adkins, California Institute of Technology

Indo-Pacific Control of Climate States and the Overturning Circulation

The tropical oceans absorb more heat from the atmosphere than they release. The resulting energy surplus is transported polewards by warm surface currents, resulting in warming of polar regions and sustaining the buoyancy loss required to close the global overturning circulation. However, there is no general understanding of how this fundamental link between low and high latitudes varies with climate state. 

Here, we use an unprecedented suite of fully-coupled climate simulations, equilibrated for thousands of years across a wide range of CO2 levels, to both highlight and explain marked differences in overturning structure and heat transport between hemispheres across climate states, from cold glacial, to warm modern-like, and hot ‘future’ CO2 levels. Changes are related through a key feature of these simulations: as CO2 increases, there is a monotonic increase in net heat uptake within the Indo-Pacific basin, that must be balanced by heat transport to the other ocean basins.  As the climate warms, the heat loss in the Indo-Pacific mid-latitudes decreases while the tropical uptake increases.  

Heat export to higher latitudes in other basins balances this excess heat and drives the overturning circulation.  However, high latitude changes evolve asymmetrically with hemisphere. Between glacial to modern analogues, heat loss in the North Atlantic intensifies, in step with an intensification of the Atlantic-sourced overturning branch. Between modern analogues and hot climates, heat loss and overturning increases around Antarctica. Basin geometries impose different dynamical regimes of heat transport in each hemisphere — advective in the North Atlantic and eddy diffusive in the Southern Ocean. We propose that the efficiency of each mechanism is climate-state dependent and offer a simple, dynamically-based scaling relationship to explain the observed phasing.

Event Details


  • Thursday, November 5, 2020 - 11:00am to Friday, November 6, 2020 - 11:59am

Virtual seminar



For More Information Contact