Bob Kopp

In a new preprint, my co-authors and I critique the increasingly broad way the concept of tipping points is being employed by natural and social scientists in climate research. We propose a correction to an excess of lumping that we think is causing confusion that may distract urgent climate action. We hope to provoke critical reflection that advances climate research and its synthesis, assessment, and communication.

The term ‘tipping point’ – which came to climate science by way of Malcolm Gladwell’s popular book The Tipping Point, and originated in the scientific literature in the 1950s and 1960s in work on neighborhood segregation – was first used in climate research to refer to ‘Earth system tipping points’: abrupt, irreversible, self-amplifying thresholds whose crossing has the potential to shift the planet as a whole into fundamentally different states.

It has since expanded to include ‘negative social tipping points’ with the potential for catastrophic and irreversible societal impacts, and ‘positive social tipping points’, viewed as a way to set in motion rapid and self-sustaining responses, like the adoption of new technologies, as well as ‘adaptation’ or ‘risk’ tipping points, which are simply tolerance thresholds for current strategies.

The current literature makes it seem like the tipping points framing captures the core of a dynamic that is ubiquitous in natural and social systems. We argue otherwise. Rather, we suggest that the ‘tipping point’ concept has become a fuzzy concept used in different ways across disciplinary and system boundaries, much like concepts of ‘sustainability’ – even as it continues to (wrongly) convey the sense of a mathematically precise dynamic.

We critique ‘tipping point’ framings for their insufficiency for describing the diverse dynamics of complex systems; their reductionist view of individuals, their agency and their aspirations in social systems; and their tendency to convey urgency without fostering a meaningful basis for climate action.

We argue for clarifying the scientific discussion of the phenomena lumped under the ‘tipping point’ umbrella by using more specific language to capture relevant aspects (e.g., irreversibility, abruptness, self-amplification, potential surprise) and for the critical evaluation of whether, how and why the different framings can support accurate scientific understanding and effective climate risk management.

In social systems, we suggest that the mental model of a ‘tipping point’ does not align with the multifaceted nature of social change, and that a broader focus on the dynamics of social transformation is more useful than centering the concept of a ‘tipping point.’

Similarly, though the originally adopters of the ‘tipping point’ terminology began using it in the explicit hope of inspiring climate action, we note that multiple social scientific frameworks suggest the deep uncertainty and perceived abstractness associated with many proposed Earth system ‘tipping points’ make them both unlikely to invoke effective action and not useful for setting governance goals.

Temperature-based benchmarks already provide a suitable guide for global mitigation policy targets, but these benchmarks should not be confused with physical thresholds of the climate system (as often happens in the media with the ‘1.5°C tipping point’).

The position we offer is provocative; many who – like I and a number of my co-authors – have contributed over the last couple decades to the literature on ‘climate tipping points’ will strongly disagree with it. We do not intend to disparage this literature, which has highlighted numerous important aspects of both natural and social systems, and certainly do not intend to disparage these researchers (including ourselves!), for whom we have a great deal of respect.

Jon Gertner writes in this week’s NYT magazine about John Moore’s proposals to geoengineering embayments around ice sheets in order to interrupt the destabilizing effect of warm subsurface waters on outlet glaciers. Figure 1 from Moore et al. (2018) summarizes the idea. Some thoughts.

1/ Crazy ideas are great for building and testing our understanding of the Earth system. I remember very clearly from undergraduate global biogeochemistry class with David Archer a quarter century ago his remark to the effect that ‘if you want to permanently (on human timescales) remove carbon from the atmosphere, you can’t just grow forests – you need to take your dead Christmas trees and sink them into anoxic parts of the ocean.’ As an undergraduate, that was great for helping build intuition about the carbon cycle.

(Of course, there are now start-ups pursuing exactly this strategy so I’m not sure quite what the takeaway is from the example…)

Similarly, aerosols are a major source of uncertainty in the Earth’s energy budget, and if researching solar radiation management helps us advance understanding of this term, that is for the good.

2/ If you take the cost estimates Gertner uses seriously – $50 billion to slow loss at Thwaites – the benefit-cost might work out. Thwaites as the potential to contribute about 1 m to global mean sea-level rise on a many-century timescale, though likely more like 1-4 cm over the next hundred years. Our analysis suggests a marginal present value of a 1 cm reduction in sea level through 2100 of about $10 billion. This is an underestimate, since it assumes optimal adaptation and doesn’t consider costs beyond 2100; with no adaptation, costs through 2100 are probably close to 10x higher. So the estimated benefits and costs are within the ballpark of one another.

3/ The scheme targets melting from below, which drives marine ice sheet instability. Hydrofracturing of protective ice shelves, driving by melting from above, may not be much affected. Most ice sheet models still don’t account for the combination of ice shelf hydrofracturing and the instability of resulting marine ice cliffs, which would ought to be taken into account in any modeling-based study of this concept.

4/ This critique is based purely on intuition, but I’m highly skeptical that building a massive, unique megainfrastructure project in the Amundsen Sea Embayment is only going to cost $50 billion. The Big Dig in Boston, fully connected to the supply lines of the Northeast US, cost $20 billion. Yes, it’s particularly costly to build megainfrastructure in the US, but is this crazy geoengineering scheme only 2.5 Big Digs worth in terms of cost? I doubt it.

So should we study this idea from a scientific perspective? By all means, we will learn a lot from doing so. Should we be expecting it to actually be practical? I wouldn’t.

For the new year, I’m trying a new approach to social media posting, using micro.blog. When I feel inspired to say some extended, I’m going to post first to blog.bobkopp.net, which will syndicate to Mastodon and Bluesky, where perhaps they will inspire interaction. This is a test.