How to predict climate tipping points

The UK’s Advanced Research and Invention Agency - or ARIA - recently announced that it is allocating £80 million to a new programme that aims to create early warning systems for climate tipping points. These tipping points get their name because - if crossed - they can trigger abrupt and irreversible onward changes to the Earth’s climate. ARIA hopes to bring together the best and brightest scientists in the world to predict when they might happen. To find out more, I went to see Sarah Bohndiek, the programme director at ARIA and a professor of biomedical physics at the University of Cambridge. What, I wanted to know first, are those climate tipping points?

Sarah - For example, we might think about the melting of an ice sheet that would contribute significantly to sea level rise, or we might think about the collapse of an ocean circulation that could shift the position of the jet stream, change the the weather in various parts of the world, which would lead to problems with agriculture, biodiversity impacts on food security. These things are happening on timescales that are still unfolding.

Chris - And is the purpose of the project to identify what those tipping points are? To actually make a shopping list of them so we know what to look for and therefore can size up the risk.

Sarah - In our program, we are looking at how we might understand when those tipping points would occur, over what timescales they might unfold, and what the consequences of crossing them would be. At the moment, we have a pretty good understanding of which parts of the Earth's system are prone to undergoing this kind of tipping between different stable states. But what we are lacking is the ability to detect when this is coming. We know that some systems are undergoing significant changes at the moment, but the question is do these changes commit us to, for example, the entire loss of an ice sheet? And we have two major challenges that are preventing us from getting to that point. One is the data. If we think about weather forecasting, we can do that with pretty good accuracy because we have an abundance of data. We are very good at knowing when a heat wave is coming, when a major storm is coming. If we try to think about when a tipping point is coming. Often these things are in incredibly harsh environments. If we're talking about ice sheets, we're talking about extremely cold and remote places, which are very hard to make measurements in. So we have a real scarcity of data, not just on a day-to-day basis, but also long-term records going back decades or centuries. Earth is such a complex and variable system. If we want to detect these changes, then we would need to have a really long record of data from which we could detect these subtle changes that we are approaching a tipping point. The other problem is a modelling problem. We have extremely sophisticated climate models that teach us about how the climate is changing over time, how different systems interact with each other. But these are hugely computationally challenging. Despite their incredible sophistication, they don't always include all of the processes that we would need in order to describe these tipping processes. Understanding how we're cross tipping points. On top of that, they're very computationally expensive. Being able to calculate a trajectory for how the climate might evolve over the coming decades or centuries takes an incredible amount of computational power. As a result of that, it's also quite hard to integrate observations that we are getting every day around the world. We are setting out to try and see whether it would be possible to set up an early warning system that unites modelling and measurements and helps us to understand better when these tipping points are likely to arise.

Chris - How does this work in practice though? Do you kind of orchestrate and bring together people who are working on this kind of problem all around the world, or are you single handedly going to try and take on this challenge? What does it look like in practice?

Sarah - I'm certainly not going to manage it single handedly. This is really a global problem and needs global talent to come together. As a program, we are going to coordinate scientists from across the world. We can fund companies, we can fund academics and universities. We can really reach out to people who wouldn't ordinarily have the opportunity to work together. We are particularly interested in bringing new people into the field of climate science. So I myself am not a climate scientist. I'm a medical physicist and I've come into this field with the blank slate that Aria has given me and the mandate to set up a program. And together with my colleague, Gemma Bale, who's co-directing the program, we've looked at this field with fresh eyes and come up with interesting ideas that we think are really important for tackling some of these problems. And we really believe that there are a huge number of people out there with expertise in new sensing and new imaging that could come and help tackle the climate crisis.

Chris - It's really interesting that the policy has taken you in this direction. We've basically arrived at a situation where we have researchers, scientists using scientific insight, but to run a program like this and have resource behind them to fund other people to get involved instead of relying on all those individuals to, to come together independently under some kind of government policy or something is that's quite an interesting and innovative way of doing things.

Sarah - It's a really key part of the Aria model is to put people first, then projects, and that starts with recruiting us as program directors and giving us that blank slate to go out into the community and find out what are the big problems of today and how can we reach for the edge of the possible to try and tackle them. In talking to the climate science community, we really identified these gaps in observations in harsh environments and also these challenges with building up the next generation of models. And we felt that we could really push this forward if we were able to build up a program around early warning of tipping points that integrated both of those things together. So having that scientist lense means that we can understand what some of the challenges are in building new sensors. What does it take to get a new measurement at the deepest depth of the ocean? What does it take to take a picture with a camera if you are out in the Arctic Circle, for example? Those are things that we've both worked on in our backgrounds as physicists, so we have an understanding of some of those challenges, but we've never thought about them in the context of climate science before.