UCL Global


How mathematics can contribute to the climate change debate

UCL Professor Ted Johnson explains how a collaboration with Florida State University could help make climate change predictions cheaper and more widely accessible

“Visualisation of the Gulf Stream stretching from the Gulf of Mexico to Western Europe” from the NASA Scientific Visualization Studio

23 September 2021

“Climate change, because of its enormous social, economic and political consequences, reigns as the leading scientific problem of our times,” explains Professor Johnson.

Through a collaboration with Professor WK Dewar from Florida State University (FSU), they have bought together the studies of oceanography and mathematics, focusing on models that concentrate on the behaviour of the oceans: to predict the climate you must predict what the oceans will do.

Having submitted a successful NERC-NSF (Natural Environment Research Council UK and the National Science Foundation USA) funding application for £700,000 to support postdoctoral and doctoral researchers, the investigators hope to make their work, should it be successful, available to the modelling community within the next three years.

This will make this kind of modelling more accessible, building knowledge about climate change and ultimately leading to more accurate decision-making related to future weather patterns.

Why modelling the oceans is so important

Being able to accurately and cheaply predict climate change over the next hundred years could enable countries to prepare against extreme changes in weather patterns which affect water supplies, farming, landmasses and every eco-system on the planet, displacing people and wildlife.

Professor Johnson explains:

“We predict accurately the position of satellites, and the tides, sunrises, sunsets, eclipses and comets because their motion is governed by the equations of physics. It is the same with the atmosphere and the ocean. Their motion is governed by equations.

However, the equations related to oceans are far more complex as the number of processes occurring and the interactions between them is far greater.

Resolving all these interactions exactly is far beyond the power of today’s greatest computers and will remain so for many decades.

We need to be clever and analyse the equations and simplify them, without losing essential physics, so that they are soluble on current computers. This analysis is called modelling.

The atmosphere changes rapidly (over a period of hours or less): just think of weather.

The ocean changes slowly over periods, from months to thousands of years, but holds and transports an enormous amount of heat that helps drive the motion of the atmosphere.

If we wish to predict the average behaviour of the atmosphere over any timescale longer than a week or so, we must predict the behaviour of the ocean. Thus accurate ocean state predictions are vital for climate predictions.”

A collaboration enabled by Global Engagement Funding

Professor Johnson and his colleagues at UCL had been working on dynamically based models of how regions of the ocean nearest to the coast behave: how energy, mass and other important physical quantities are transported rapidly along coasts and perhaps re-injected into the interior.

Independently, Professor Dewar had been considering the importance of motions along coastal boundaries as a mechanism though which energy, injected into the oceans by winds, is removed to maintain an approximate energy balance in the ocean.

There were many areas of overlap but since Professor Dewar was based in Florida, it wasn’t easy to organise a meeting.

However, thanks to Global Engagement Funding (GEF), Professor Dewar was able to visit UCL to meet and formulate ideas for future collaborations. The informality of meeting in person was vital at the initial stage of the collaboration when the direction of any joint research was still unclear. With this basis, subsequent development could be based on Skype conversations and email.

This enabled them to:

  • Formulate a simple model to test their ideas in collaboration with Dr Bruno Deremble (Laboratoire de Météorologie Dynamique (LMD), Paris) whose modelling expertise was invaluable.
  • Publish a joint paper in Ocean Modelling journal:
  • Submit a successful joint NERC-NSF grant application, which will allow visits between FSU, LMD and UCL, support a Postdoctoral Research Assistant at UCL and support a Graduate Research Assistant at FSU.

Professor Johnson said: “The entire collaboration, journal paper and successful grant application would not have taken place without Professor Dewar’s initial visit supported by the Global Engagement Office.”

Next steps for this important research

Their modelling aims to split the ocean interior (where the dynamically important scales tend to be larger) from the ocean boundaries (e.g. coastlines) where motions are faster and on a smaller scale.

This splitting, if successful, will capture the important dynamics of each region without requiring computer power beyond current capabilities.

“If we can show our approach works, then we hope it will stimulate other groups to adopt and improve our methods.

In conjunction with project partners at the UK National Oceanography Centre we hope to produce modules for community climate models that improve their accuracy and lower their computational cost.

This will enable more groups to run more simulations to test more hypotheses about the climate.

At the moment, the high computational cost of simulations means that only the largest groups (perhaps one or two per country) can afford to run comprehensive, long-time simulations. This possibility of multiple, cheap and straightforward simulations attracted the interest of our other project partner, Willis Towers Watson, a London-based global advisory company designing systems to manage risks like climate change.

More tested hypotheses should lead to more knowledge and more accurate decision making.”

More information

For the latest news about UCL’s international activity, partnerships and opportunities, subscribe to the bimonthly Global Update newsletter.