Climate Change Uncertainty is a reason for action not inaction

One of the primary motivations for the Uncertain World – from its conception to this website – is to explain the perils of uncertainty, how it undermines confidence, planning, strategy and adaptation and by extension every sector of modern society from farming to finance.  Uncertainty is risk.  And the greater that uncertainty, especially with respect to its intersection with complex ecological, biogeochemical and social systems, the greater the risk.  Others, however, use uncertainty as a means to confuse and prevaricate: ‘The science is not settled.’  ‘We don’t know enough to act.’  This blog co-written with my friend and colleague Steve Lewandowsky tackles that issue (originally published in The Conversation).

Former environment minister Owen Paterson has called for the UK to scrap its climate change targets. In a speech to the Global Warming Policy Foundation, he cited “considerable uncertainty” over the impact of carbon emissions on global warming, a line that was displayed prominently in coverage by the Telegraph and the Daily Mail.

Paterson is far from alone: climate change debate has been suffused with appeals to “uncertainty” to delay policy action. Who hasn’t heard politicians or media personalities use uncertainty associated with some aspects of climate change to claim that the science is “not settled”?

Over in the US, this sort of thinking pops up quite often in the opinion pages of The Wall Street Journal. Its most recent article, by Professor Judith Curry, concludes that the ostensibly slowed rate of recent warming gives us “more time to find ways to decarbonise the economy affordably.”

At first glance, avoiding interference with the global economy may seem advisable when there is uncertainty about the future rate of warming or the severity of its consequences.

So let’s do nothing. WSJ

But delaying action because the facts are presumed to be unreliable reflects a misunderstanding of the science of uncertainty. Simply because a crucial parameter such as the climate system’s sensitivity to greenhouse gas emissions is expressed as a range – for example, that under some emissions scenarios we will experience 2.6°C to 4.8ºC of global warming or 0.3 to 1.7 m of sea level rise by 2100 – does not mean that the underlying science is poorly understood. We are very confident that temperatures and sea levels will rise by a considerable amount.

Perhaps more importantly, just because some aspects of climate change are difficult to predict (will your county experience more intense floods in a warmer world, or will the floods occur down the road?) does not negate our wider understanding of the climate. We can’t yet predict the floods of the future but we do know that precipitation will be more intense because more water will be stored in the atmosphere on a warmer planet.

This idea of uncertainty might be embedded deeply within science but is no one’s friend and it should be minimised to the greatest extent possible. It is an impetus to mitigative action rather than a reason for complacency.

Uncertainty means greater risk

There are three key aspects of scientific uncertainty surrounding climate change projections that exacerbate rather than ameliorate the risks to our future.

First, uncertainty has an asymmetrical effect on many climatic quantities. For example, a quantity known as Earth system sensitivity, which tells us how much the planet warms for each doubling of atmospheric carbon dioxide concentration, has been estimated to be between 1.5°C to 4.5ºC. However, it is highly unlikely, given the well-established understanding of how carbon dioxide absorbs long-wave radiation, that this value can be below 1ºC. There is a possibility, however, that sensitivity could be higher than 4.5ºC. For fundamental mathematical reasons, the uncertainty favours greater, rather than smaller, climate impacts than a simple range suggests.

Second, the uncertainty in our projections makes adaptation to climate change more expensive and challenging. Suppose we need to build flood defences for a coastal English town. If we could forecast a 1m sea level rise by 2100 without any uncertainty, the town could confidently build flood barriers 1m higher than they are today. However, although sea levels are most likely to rise by about 1m, we’re really looking at a range between 0.3m and 1.7m. Therefore, flood defences must be at least 1.7m higher than today – 70cm higher than they could be in the absence of uncertainty. And as uncertainty increases, so does the required height of flood defences for non-negotiable mathematical reasons.

And the problem doesn’t end there, as there is further uncertainty in forecasts of rainfall occurrence, intensity and storm surges. This could ultimately mandate a 2 to 3m-high flood defence to stay on the safe side, even if the most likely prediction is for only a 1m sea-level rise. Even then, as most uncertainty ranges are for 95% confidence, there is a 5% chance that those walls would still be too low. Maybe a town is willing to accept a 5% chance of a breach, but a nuclear power station cannot to take such risks.

Finally, some global warming consequences are associated with deep, so-called systemic uncertainty. For example, the combined impact on coral reefs of warmer oceans, more acidic waters and coastal run-off that becomes more silt-choked from more intense rainfalls is very difficult to predict. But we do know, from decades of study of complex systems, that those deep uncertainties may camouflage particularly grave risks. This is particularly concerning given that more than 2.6 billion people depend on the oceans as their primary source of protein.

Similarly, warming of Arctic permafrost could promote the growth of CO2-sequestering plants, the release of warming-accelerating methane, or both. Warm worlds with very high levels of carbon dioxide did exist in the very distant past and these earlier worlds provide some insight into the response of the Earth system; however, we are accelerating into this new world at a rate that is unprecedented in Earth history, creating additional layers of complexity and uncertainty.

Uncertainty does not imply ignorance

Increasingly, arguments against climate mitigation are phrased as “I accept that humans are increasing CO2 levels and that this will cause some warming but climate is so complicated we cannot understand what the impacts of that warming will be.”


This argument is incorrect – uncertainty does not imply ignorance. Indeed, whatever we don’t know mandates caution. No parent would argue “I accept that if my child kicks lions, this will irritate them, but a range of factors will dictate how the lions respond; therefore I will not stop my child from kicking lions.”

The deeper the uncertainty, the more greenhouse gas emissions should be perceived as a wild and poorly understood gamble. By extension, the only unequivocal tool for minimising climate change uncertainty is to decrease our greenhouse gas emissions.



Steve and I with many brilliant colleagues elaborated on these in a special publication of the Royal Society edited by Tim Ballard, Steve and myself – Uncertainty as Knowledge.  The introduction is now available online for free and elaborates on these critical issues.  And do watch the accompanying video!



The Pliocene – The last time Earth had 400 ppm of Carbon Dioxide

One of the main approaches for better understanding our future Uncertain World, both with respect to minimising that uncertainty but also identifying the limits to our knowledge, is to look to the past.  A few years ago, we passed 400 ppm of CO2 in the atmosphere, a level that we have not experienced for about 3 million years, during the Pliocene epoch.  That coincided with Bristol hosting the 2nd International Conference on the Pliocene, providing an opportunity through this piece, originally published on The Conversation, and its embedded video to discuss how we try to unravel ancient climates to better understand the future.  I have updated some of the text, reflecting some of the changing intellectual landscape.  To provide some immediate context, below is the latest record of current atmospheric carbon dioxide concentrations from the Mauna Loa Observatory (i.e. The Keeling Curve).

How can air bubbles trapped in ice for millions of years, or fossilised fern fronds, or the chemical make-up of rocks that were underwater in the distant past provide us with an inkling of our future?

The answer lies in these clues provided by studying the Pliocene epoch, the span of geological time that stretched from 5.3 to 2.6 million years ago. This period of Earth’s history is interesting for many reasons, but one of the most profound is that the Earth’s atmosphere apparently contained high concentrations of carbon dioxide. Our best estimates suggest concentrations of about 300-400 parts per million (ppm) – much higher than concentrations of 100 years ago, but the same or lower than today after centuries of industrialisation and fossil fuel burning.

So studying the Pliocene could provide valuable insight into the type of planet we are creating via global warming. Our researchers at the Cabot Institute recently released a video on the topic, which has coincided with pronounced flooding across the UK [in Winter and Spring 2014] and renewed attention focused on our weather and climate. There is little doubt that increased carbon dioxide concentrations will cause global warming. The key questions are how much, and with what consequences.

One of the key lessons from Earth history is climate sensitivity. Climate sensitivity can be expressed in various ways, but in its simplest sense it is a measure of how much warmer the Earth becomes for a given doubling of atmospheric carbon dioxide concentrations.

This is well known for the Pleistocene, and especially the past 800,000 years of Earth history, a period for which we have detailed temperature reconstructions and carbon dioxide records derived from bubbles of gas trapped in ancient ice cores.

During that time, across several ice ages, the planet’s climate sensitivity showed warming of about 2.5-3°C for a doubling of carbon dioxide, which falls in the middle of the range of predictions given by models. Ice core records, however, extend back no more than a million years, and this time period is generally characterised by colder climates than those of today.


A section of an ice core of from 16,000 years ago. National Ice Core Laboratory

If we want to explore climate sensitivity on a warmer planet, we must look further back into Earth history, to times such as the Pliocene.

Reconstructing atmospheric carbon dioxide concentrations without relying on ice cores is admittedly more challenging. Instead of directly measuring the concentration of carbon dioxide in gas bubbles, we must rely on indirect records. For example, carbon dioxide concentration influences the number of stomata (pores) on plant leaves, and this can be measured on the fossils of ancient leaves. Alternatively, there are a number of geochemical tools based on how carbon dioxide affects the pH of seawater, or how algae take up carbon dioxide as they photosynthesise – these are recorded in the chemical composition of ancient fossils.

[For more info on how we reconstruct atmospheric carbon dioxide, especially in times pre-dating our ice core records, see this fantastic website maintained by Gavin Foster and friends.]

 Atmospheric CO2 from AD 1000 to AD 2018 (right) from a mix of ice core records and measuresments of the astmosphere from Mauna Lao.  On the left is a compilation of ice core CO2 (red) and boron isotope based estimates (blue).  Note the age scales are different but y-axis is the same. See this document for references.


These means of drawing estimates come with larger margins of error, but they still provide key insights into climate sensitivity on a warmer Earth. Recent research indicates that these various carbon dioxide estimates of Pliocene carbon dioxide levels are converging, giving added confidence from which to derive estimates of climate sensitivity. In particular, it seems an increase of carbon dioxide from about 280ppm (equivalent to that before the industrial revolution) to about 400ppm in the Pliocene resulted in an Earth warmer by 2°C.  The below figure shows the sea surface temperatures reconstructed for the Pliocene using a range of chemical and biological proxy data (a) and the difference (anomaly) between those temperatures and those of the modern pre-industrial world, i.e. before we started adding carbon dioxide to the atmosphere in significant quantities (b); notice how much hotter the oceans were, especially at high latitudes. From the Proceedings of the Royal Society.


Image result for Pliocene SSTs


This next figure is derived from an ensemble of climate models, which allows an extrapolation between data and therefore a comparison of temperatures on land and the modern pre-industrial world.  Again, note how much higher temperatures were at high latitudes… but also continental interiors.  We are seeing a manifestation of this now, with elevated temperatures occurring all over the globe but some areas experiencing much more dramatic warming.  This work is from the amazing PlioMIP project and this figure is specifically from the PAGES website, adapted from Haywood et al., 2013.


Image result for images of pliocene warmth terrestrial


Taking into account other factors, this suggests a climate sensitivity of about 3°C, which confirms both the Pleistocene and model-based estimates. It also suggests that we have yet to experience the full consequences of the greenhouse gases already added to the atmosphere, let alone those we are still putting into it.  [And finally, it suggests that there is a risk that we have already surpassed the agreed limits of the Paris Climate Agreement.]

So then, what was this much warmer world like? First of all, it was not an inhospitable planet – plants and animals thrived. This should not be a surprise – in fact, the Earth was much warmer even further back into the past. The changes in the climate we are inducing is a problem for us humans, and for our societies, not the planet we’re on. [And that is particularly evident, as I re-post this blog.  We are experiencing a global heatwave that is causing forest fires (where associated with aridity), but it is also impacting infrastructure and the economy, warping rail lines, disrupting work patterns, driving up electricity usage. It is also causing deaths which raises a particularly acute and challenging question – are we and the ecosystems on which we depend prepared for the speed of this rapid global warming?  Organisms and ecosystems had millions of years to evolve in a manner that allowed them to thrive in the Pliocene and previous greenhouse climates.]

Second, the Pliocene was a rather different world. For example, higher global temperatures were associated with a climate that was also wetter than at present. That provides important corroborating evidence for models that predict a warmer and wetter future.

Perhaps most striking, sea level in the Pliocene appears to have been between 10 to 40 metres higher than today, indicating that both the Greenland Ice Sheet and Antarctic Ice Sheet were markedly smaller. To put that into context, the Met Office has already commented on how flooding in the UK has been affected by sea level rise of 12cm over the last 100 years, and will be exacerbated further by another 5-7cm by 2030.

We must be careful in how we extract climate lessons from the geological record, and that is particularly true when we consider ice sheet behaviour. One widely discussed concept is ice sheet hysteresis. This is a fancy way of saying that due to feedback mechanisms, it could be easier to build an ice sheet on Greenland or Antarctica than it is to melt one. If hysteresis is a force stabilising our current ice sheets, then it may be that a planet with today’s carbon dioxide levels of 400 ppm will not necessarily have a sea level 20 metres higher than that of today – as it was during the Pliocene. On the other hand if hysteresis is rather weak, then the question is not whether we will see such a massive sea level change, but how long it will take to arrive (probably hundreds or even thousands of years).

Most importantly, the collective research into Earth history, including the Pliocene, reveals that Earth’s climate can and has changed. It also reveals that climate does not just change randomly: it changes when forced in ways that are relatively well understood – one of these is the concentration of carbon dioxide in our atmosphere. And consequently, there is little doubt from Earth’s history that transforming fossil carbon underground into carbon dioxide in the air – as we are doing today – will significantly affect the climate we experience for the forseeable future.

[Gerald Haug delivered the keynote for the Pliocene Conference and his outstanding public lecture is available here. With an introduction from the founder of the Organic Geochemistry Unit, Geoff Eglinton.]

[For a more fulsome discussion of how Warm Climates of the Past can hold Lessons for the Future, please check out our Special Royal Society Volume on the topic.  Led by my Bristol colleague Dan Lunt but with lots of friends.]

The Origins of the Uncertain World

In late 2014, the Cabot Institute was in deep consultation with artists, colleagues, businesses and political leaders about our contribution to Bristol EU Green Capital 2015.  Given the breadth of Cabot, we were keen to contribute in diverse ways, especially around sustainability solutions and the range of environmental challenges we face, from plastics in the sea to procuring safe, sustainable food.  However, 2015 was also a fantastic chance to discuss climate change, its causes and impacts and how Bristol and the wider world would have to adapt – especially given that 2015 would culminate with the COP21 climate negotiations in Paris.  At the same time, we wanted to examine climate change through a somewhat different lens than had been done in the past.  Uncertainty was that lens. We wrote this at the end of 2014 announcing the Uncertain World as our framework for discussing these issues during 2015 and beyond.  It went on to inform Bristol’s strong commitments to climate change and its Resilience Strategy.

Originally posted on the Cabot Institute blog, this was our statement of intent.


Over the next 18 months, in collaboration with Bristol Green Capital 2015 artists, civic leaders and innovative thinkers, the Cabot Institute will be participating in  a series of activities in which we examine how human actions are making our planet a much more uncertain place to live.
Fifty years ago, between 1962 and 1966, J. G. Ballard wrote a trio of seminal environmental disaster novels: The Drowned World, The Burning World and The Crystal World.  These novels remain signposts to our future, the challenges we might face and the way people respond to rapid and unexpected change to their environment. In that spirit and coinciding with the Bristol Green Capital 2015, we introduce The Uncertain World, a world in which profound uncertainty becomes as much of a challenge to society as warming and rising sea levels.

J.G Ballard’s The Drowned World
taken from
For the past twenty years, the University of Bristol has been exploring how to better understand, mitigate and live with environmental uncertainty, with the Cabot Institute serving as the focus for that effort since its founding in 2010.  Uncertainty is the oft-forgotten but arguably most challenging aspect of mankind’s centuries-long impact on the environment.  We live our lives informed by the power of experience: our own as well as the collective experience of our families, communities and wider society. When my father started dairy farming he sought advice from my mother’s grandfather, our neighbours, and the grizzled veterans at the Middlefield auction house. Experience helps us make intelligent decisions, plan strategically and anticipate challenges.

Similarly, our weather projections, water management and hazard planning are also based on experience: tens to hundreds of years of observation inform our predictions of future floods, drought, hurricanes and heat waves. These records – this experience  – can help us make sensible decisions about where to live, build and farm.

Now, however, we are changing our environment and our climate, such that the lessons of the past have less relevance to the planning of our future.  In fact, many aspects of environmental change are unprecedented not only in human experience but in Earth history. As we change our climate, the great wealth of knowledge generated from human experience is losing capital every day.

The Uncertain World is not one of which we have no knowledge – we have high confidence that temperatures and sea level will rise, although there is uncertainty in the magnitude and speed of change. Nor should we view The Uncertain World with existential fear – we know that warm worlds have existed in the past.  These were not inhospitable and most evidence from the past suggests that a climate ‘apocalypse’ resulting in an uninhabitable planet is unlikely.

Nonetheless, increasing uncertainty arising from human-induced changes to our global environment should cause deep concern.  Crucial details of our climate remain difficult to predict, and it undermines our ability to plan for our future. We do not know whether many regions of the world will become wetter or dryer. This uncertainty propagates and multiplies through complex systems: how do we make sensible predictions of coastal flood risk when there is uncertainty in sea level rise estimates, rainfall patterns and the global warming that will impact both?  We can make predictions even in such complex systems, but the predictions will inevitably come with a degree of uncertainty, a probabilistic prediction.  How do we apply such predictions to decision making? Where can we build new homes, where do we build flood defences to protect existing ones, and where do we abandon land to the sea?

Methane escaping from Arctic
permafrost. Image:

Perhaps most worrying, the consequences of these rapid changes on biological and chemical systems, and the people dependent upon them, are very poorly understood. For example, the synergistic impact of warmer temperatures, more acidic waters, and more silt-choked coastal waters on coral reefs and other marine ecosystems is very difficult to predict. This is particularly concerning given that more than 2.6 billion people  depend on the oceans as their primary source of protein. Similarly, warming of Arctic permafrost could promote the growth of CO2-sequestering plants or the release of warming-accelerating methane – or both. Warm worlds with very high levels of carbon dioxide did exist in the past and these do provide some insight  into the response of the Earth system, but we are accelerating into this new world at a rate that is unprecedented in Earth history, creating additional layers of uncertainty.
During late 2014 and 2015, the Cabot Institute will host a variety of events and collaborate with a variety of partners across Bristol and beyond to explore this Uncertain World and how we can live in it. How do we better explain uncertainty and what are the ‘logical’ decisions to make when faced with uncertainty? One of our first events will explore how uncertainty in climate change predictions should motivate us to action: the more uncertain our predictions the more we should employ mitigation rather than adaptation strategies. Future events will explore how past lessons from Earth history help us better understand potential future scenarios; how future scenario planning can inform the decisions we make today; and most importantly, how we build the necessary flexibility into social structures to thrive in this Uncertain World.