01 February 2023

Will Steffen’s crucial climate ideas on “Hothouse Earth”, tipping cascades and non-linearity

By David Spratt 

The eminent Australian climate scientist, and former Labor government advisor and head of climate at ANU, Will Steffen, who died early this week from complications following cancer surgery, will be remembered for some of the big, crucial ideas he and his collaborators contributed to the understanding of the Earth System, particularly planetary boundaries, climate tipping point vulnerabilities and cascades, risk and nonlinearity, and the “Hothouse Earth” scenario. 

Particularly in the last few years, Steffen was very clear and forthright in communicating the threat and the dynamics of the climate system, and the trajectory towards collapse:

"Given the momentum in both the Earth and human systems, and the growing difference between the ‘reaction time’ needed to steer humanity towards a more sustainable future, and the ‘intervention time’ left to avert a range of catastrophes in both the physical climate system (e.g., melting of Arctic sea ice) and the biosphere (e.g., loss of the Great Barrier Reef), we are already deep into the trajectory towards collapse … That is, the intervention time we have left has, in many cases, shrunk to levels that are shorter than the time it would take to transition to a more sustainable system.”

More recently, Steffen and others asked:

“Could anthropogenic climate change result in worldwide societal collapse or even eventual human extinction? At present, this is a dangerously underexplored topic. Yet there are ample reasons to suspect that climate change could result in a global catastrophe.”

And in a recent book chapter, "The Earth System, the Great Acceleration and the Anthropocene", he recognised that climate change was already dangerous:

"It is clear from observations of climate change-related impacts in Australia alone – the massive bushfires of the 2019–2020 Black Summer; the third mass bleaching of the Great Barrier Reef in only 5 years; and long-term cool-season drying of the country’s southeast agricultural zone – that even a 1.1 °C temperature rise has put us into a dangerous level of climate change.” 

Steffen was the key scientific climate advisor to the 2007-2013 Australian Labor government.  In that role I often disagreed with his political advocacy, and told him so. Will’s advocacy stuck pretty rigidly to the 2°C, 450-ppm-CO2 goal that was de rigueur in global and national policy-making circles in those days, even as the science was already clear that it was a dangerous goal, and Jim Hansen had persuasively outlined the case for a goal of under 350 ppm, which had been adopted around the world by some more progressive parts of the climate movement. 

But Steffen was far from alone amongst scientists around the world who become jammed between politics and science, in the struggle for balance between research knowledge and wanting, or sounding, to be politically relevant to governments either in denial or unwilling to grasp the size of the task.

At the same time, Steffen was keenly aware in 2008 that “the scientific community is underestimating the speed at which the climate is changing”, warning that the world's sea levels could rise by up to four metres by the end of the century.

And whilst Ross Garnaut’s 2008 report for Labor on climate change had contemplated the 450 ppm and 550 ppm goals (with short-term warming of around 2°C and 3°C respectively), Steffen in a 2009 report for the climate change department had recognised that 550 ppm (a doubling of the pre-industrial CO2 level) could lead to 6°C of warming:

“... when ice extent is treated not as a forcing but as a climate system response, so that both fast and slow feedbacks are included, the climate sensitivity approximately doubles (Hansen et al. 2008). Therefore, the eventual temperature rise in response to a doubling of CO2 is at least 3C and likely up to 6C, depending on the behaviour of the slow feedbacks.” 

Many scientists and collaborators have already expressed their sorrow at the death of Will Steffen and paid tribute to his legacy, for example here and here and here

Steffen’s passing is a great loss to climate understanding and advocacy in Australia, and to the research community at a global level.  I will remember his work particularly for publications over the last fifteen years that dealt with system-level analysis of climate change, boundaries, risks and dynamics.  In no particular order, some key papers and contributions in which he was either lead author or a co-author include the following:

1. “Climate tipping points — too risky to bet against” (2020)

Authored by Timothy Lenton, Johan Rockström, Owen Gaffney, Stefan Rahmstorf, Katherine Richardson, Will Steffen and Hans Joachim Schellnhuber, this paper was the first peer-reviewed paper (to my knowledge) to recognise the climate emergency as constituted by existential risks: “We are in a climate emergency… this is an existential threat to civilization” and “the evidence from tipping points alone suggests that we are in a state of planetary emergency: both the risk and urgency of the situation are acute”.

The paper pointed to “biosphere tipping points which can trigger abrupt carbon release back to the atmosphere.. Permafrost across the Arctic is beginning to irreversibly thaw and release carbon dioxide and methane… the boreal forest in the subarctic is increasingly vulnerable.”  

The authors said that: 

“the clearest emergency would be if we were approaching a global cascade of tipping points that led to a new, less habitable, ‘hothouse’ climate state. Interactions could happen through ocean and atmospheric circulation or through feedbacks that increase greenhouse-gas levels and global temperature. Alternatively, strong cloud feedbacks could cause a global tipping point.  We argue that cascading effects might be common. Research last year analysed 30 types of regime shift spanning physical climate and ecological systems... This indicated that exceeding tipping points in one system can increase the risk of crossing them in others. Such links were found for 45% of possible interactions. In our view, examples are starting to be observed.”

The conclusion to be drawn from all of this is that the threat, the risks, are overwhelming. And it is a slap to conventional economic modelling of climate risks, when they write that:  

“If damaging tipping cascades can occur and a global tipping point cannot be ruled out, then this is an existential threat to civilization. No amount of economic cost–benefit analysis is going to help us. We need to change our approach to the climate problem.”

2. “Trajectories of the Earth System in the Anthropocene” (2018)

With Steffen as lead author, this paper known colloquially as the “Hothouse Earth” paper, and one that captured the public imagination, was downloaded 270,000 times in just the first few days after publication.

A “Hothouse Earth” scenario is described, in which system feedbacks and their mutual interaction could drive the Earth System climate to a point of no return, whereby further warming would become self-sustaining (without further human perturbations). This “Hothouse Earth” planetary threshold could exist at a temperature rise as low as 2°C, possibly even in the 1.5°C-2°C range.

Steffen also said that: “even if the Paris Accord target of a 1.5°C to 2.0°C rise in temperature is met, we cannot exclude the risk that a cascade of feedbacks could push the Earth System irreversibly onto a ‘Hothouse Earth’ pathway.”  

In other words, even 1.5°C of warming could be so dangerous that further warming would become self-sustaining; a warning seemingly ignored by policymakers and most advocates.

3. Climate Endgame: Exploring catastrophic climate change scenarios (2022)

This paper published last year by ten authors, including Steffen, cannot be underestimated in bringing together a high-level analysis of the need for climate research to focus on the worse-case high-end possibilities:

  • “Prudent risk management requires consideration of bad-to-worst-case scenarios.” We know that temperature rise has “fat tails”: low-probability, high-impact extreme outcomes. Large uncertainties about dangerous surprises “are reasons to prioritize rather than neglect them.” Thus “a thorough risk assessment would need to consider how risks spread, interact, amplify, and are aggravated by human responses”. 

  • “Climate damages are likely to be nonlinear” and result in an even larger risk tail. There are feedbacks in the carbon cycle and potential tipping points that could generate high greenhouse concentrations that are often missing from models.”  There are even more uncertain feedbacks, which, in a very worst case, might amplify to an irreversible transition into a “Hothouse Earth” state [including] “recent simulations suggest that stratocumulus cloud decks might abruptly be lost at CO2 concentrations that could be approached by the end of the century, causing an additional ∼8 °C global warming. Large uncertainties about dangerous surprises are reasons to prioritize rather than neglect them.”
  • Feedbacks and future warming: Declining emissions does not rule out … extreme climate change [due to] feedbacks in the carbon cycle and potential tipping points that could generate high greenhouse concentrations that are often missing from models. Examples include Arctic permafrost thawing that releases methane and CO2 , carbon loss due to intense droughts and fires in the Amazon, and the apparent slowing of dampening feedbacks such as natural carbon sink capacity. These are likely to not be proportional to warming… instead, abrupt and/or irreversible changes may be triggered at a temperature threshold.   Particularly worrying is a “tipping cascade” in which multiple tipping elements interact in such a way that tipping one threshold increases the likelihood of tipping another.

4. "A safe operating space for humanity" (2009)

Led by John Rockstrom and Will Steffen, with 27 others (including Hansen and Schellnhuber), this landmark paper proposed that a safe level of atmospheric carbon dioxide should not exceed 350 ppm:

"We propose that human changes to atmospheric CO2 concentrations should not exceed 350 parts per million by volume, and that radiative forcing should not exceed 1 watt per square metre above pre-industrial levels. Transgressing these boundaries will increase the risk of irreversible climate change, such as the loss of major ice sheets, accelerated sea-level rise and abrupt shifts in forest and agricultural systems. Current CO2 concentration stands at 387 p.p.m.v. and the change in radiative forcing is 1.5 watts per square metre. There are at least three reasons for our proposed climate boundary. First, current climate models may significantly underestimate the severity of long-term climate change for a given concentration of greenhouse gases. Most models suggest that a doubling in atmospheric CO2 concentration will lead to a global temperature rise of about 3C°C (with a probable uncertainty range of 2–4.5°C) once the climate has regained equilibrium. But these models do not include long-term reinforcing feedback processes that further warm the climate, such as decreases in the surface area of ice cover or changes in the distribution of vegetation. If these slow feedbacks are included, doubling CO2 levels gives an eventual temperature increase of 6°C (with a probable uncertainty range of 4–8°C). This would threaten the ecological life-support systems that have developed in the late Quaternary environment, and would severely challenge the viability of contemporary human societies."

5. “Human impacts on planetary boundaries amplified by Earth system interactions” (2020)

In this paper, Lade, Steffen et al surveyed and provisionally quantified interactions between the Earth system processes represented by the planetary boundaries and investigated their consequences for sustainability governance and identified a dense network of interactions between the planetary boundaries. The resulting cascades and feedbacks predominantly amplify human impacts on the Earth system and thereby shrink the safe operating space for future human impacts on the Earth system. The three key findings for policymakers were: 

  • Interactions are crucial to understanding the planetary boundaries and humanity’s impacts upon them. For example, biophysically-mediated interactions have almost doubled direct human impacts on the planetary boundaries. 
  • Most interactions we found were amplifying, meaning that impacts on one  planetary boundary lead to increased impacts on other planetary boundaries. Cascading of human actions through multiple components of the Earth system complicates governance of the Earth system. On the other hand, these interactions offer substantial scope for synergies: if impacts on one planetary boundary are decreased, impacts on other planetary boundaries may also lessen. 
  • Interactions between planetary boundaries lead to trade-offs between the  boundaries. For example, interactions between agricultural activity and carbon emissions mean that high levels of both cannot be maintained. On the other hand, these trade-offs offer humanity some freedom in choosing how to navigate to a safe operating space.

6. Model limitations

Will Steffen was clear about the limitations of models and a quantification fetish when dealing with hard-to-project non-linear change and “Hothouse Earth”possibilities, for example in this Guardian article from 2018: 

“I think the dominant linear, deterministic framework for assessing climate change is flawed, especially at higher levels of temperature rise. So, yes, model projections using models that don’t include these processes indeed become less useful at higher temperature levels. Or, as my co-author John Schellnhuber says, we are making a big mistake when we think we can ‘park’ the Earth System at any given temperature rise – say 2°C  – and expect it to stay there”.
“Even at the current level of warming of about 1°C above pre-industrial, we may have already crossed a tipping point for one of the feedback processes (Arctic summer sea ice), and we see instabilities in others – permafrost melting, Amazon forest dieback, boreal forest dieback and weakening of land and ocean physiological carbon sinks. And we emphasise that these processes are not linear and often have built-in feedback processes that generate tipping point behaviour. For example, for melting permafrost, the chemical process that decomposes the peat generates heat itself, which leads to further melting and so on.”

This is important because so much policymaking has been based on a goal of climate getting towards 2°C and then assuming that could become stable, when in fact the paleoclimate history suggest that 2°C is not a point of system stability, but a signpost on the road to a much hotter  3–4°C outcome.

And with William Knorr in The Conversation,  Steffen warned that due to model limitations, we will not know exactly how the climate crisis will unfold until it’s too late.

Finally, Steffen’s capacity to communicate clearly and succinctly is well encapsulated in an interview with The Intercept when he said getting greenhouse  gas emissions down fast has to be “the primary target of policy and economics” with something “like wartime footing”.

That sounded to me like a very short, effective description of the climate emergency.