Oxford-led project uses machine learning to study the ‘wild card’ of climate change
The effects of climate change are already being felt across the world – and with devastating consequences.
Not only are they getting more powerful, but they are happening at an increasingly quickening rate.
Just this week a United Nations official issued a stark warning, saying that climate crisis disasters capable of causing death, displacement and suffering now occur at a rate of about one a week.
While the big events, like the cyclones in Mozambique and the drought in India make headlines, there are continually smaller events happening the world over, Mami Mizutori, the UN secretary-general’s special representative on disaster risk reduction, told the Guardian. These events may not get the same level of attention but they are just as devasting to the affected communities.
Mizutori emphasised that small-scale climate crises are happening much faster and more frequently than previously predicted, so the urgency to tackle climate change is immense and growing every day.
While so many climate scientists, after decades of intensive study, have a solid grasp on the workings of global warming and the recommended methods to slow it down, there is one area that still remains somewhat of a question mark.
The US National Science Foundation (NSF) calls this weather feature the “wild card of climate change.” It’s something that is so omnipresent that many of us simply take them for granted. We all learn about them in primary school, and yet they still remain somewhat of an enigma when it comes to climate. So what is this mysterious entity, you ask? Quite simply, it’s clouds.
“It is a little-known but significant fact that about 70 percent of the Earth’s surface is covered by clouds at any given time,” explains the NSF. “But not all clouds are the same; different types of clouds affect the Earth’s climate differently. While some types of clouds help to warm the Earth, others help to cool it.”
“Currently, all of the Earth’s clouds together exert a net cooling effect on our planet. But the large and opposing influences of clouds on the Earth’s climate begs the question: What will be the net effect of all of the Earth’s clouds on climate as the Earth continues to warm in the future? Will clouds accelerate warming or help offset, or dull warming?”
A new network of academics and industry, funded by the European Commission, is hoping to answer these questions.
Led by Oxford University, the group will train PhD students in machine learning skills to determine the role of aerosols and clouds on global warming.
Called iMIRACLI (innovative MachIne leaRning to constrain Aerosol-cloud CLimate Impacts), the project brings together big names from both the academic and the industrial worlds, including Amazon, the UK MetOffice, and a consortium of nine universities.
Philip Stier, Professor of Atmospheric Physics at Oxford University, is taking the lead on the project which will start with students taking a summer school held at the famous university. It will also fund 15 PhD students across Europe, with three of them directly supervised in Oxford – two in Physics, one in Statistics.
Each student will have a climate science and a machine learning supervisor as well as an industrial advisor. All students will have secondments to their industrial partners, as well as to the co-supervisor.
“Machine Learning has the potential to unlock unique in-depth understanding of the climate system from vast climate datasets. However, this requires a new generation of experts with substantial knowledge of both climate and data science,” explains Stier in a press release.
“We will train and shape a new generation of climate data scientists, with a solid foundation in climate science and a competence in the latest machine learning techniques.”
While the project is led by Oxford, the training network also includes the University of Leipzig, Stockholm University, ETH Zurich, the University of Edinburgh, Universitat de Valencia, University College London, the German Aerospace Center (DLR), Ecole Polytechnique Federale de Lausanne and the University of Jena, together with the non-academic partners Amazon, The Alan Turing Institute, MetOffice, Iris.ai, GAF AG and FastOpt.