Climate change – strong action in the presence of uncertainty

Managing climate change is one of the greatest challenges of the 21st century. We are already in a very difficult place with concentrations of greenhouse gases at their highest levels for at least 450,000 years. The science suggests that unless the world takes strong action to reduce emissions over the coming decades the risks of dangerous and disruptive changes in climate are immense. For example, the climate records, from a variety of sources such as ice cores and ocean sediments, suggest that while greenhouse gas concentrations may have changed relatively slowly in the past, climate transitions have appeared very rapidly. If concentrations of greenhouse gases in the atmosphere continue to increase over the coming decades, we risk shifts in climate and impacts appearing very suddenly. This, and other insights from the science, indicates that the risks we are taking with the planet are immense and as such managing climate change is a problem of risk management. To view the problem otherwise is likely to result in misleading conclusions that tend to suggest little or delayed action. Therefore understanding how to comprehend and think about risk and uncertainty is key to greater progress on managing climate change.

There are several types of uncertainty relevant to climate change; as described in the paper by Smith and Stern in 2011 for the Royal Society. The first is “imprecision” or statistical uncertainty where we are able to provide statements of probability around a particular outcome. The second is “ambiguity” (or Knightian uncertainty) where it is not currently possible to provide statements of probability around an outcome or we are uncertain about the probabilities themselves. Third is “intractability” where we know it is possible to make computations relevant to an outcome but it is beyond current mathematical or computational capacity. And last is “indeterminacy” where we are unable to attach precise values to quantities that are relevant for policy-making. Here differing views between people and ethical considerations are relevant.

Understanding how these different types of uncertainty apply to this problem is very helpful; they provide a robust framework for thinking through the issues around climate change. For example, greenhouse gas concentrations (or stocks) have increased from around 280 parts per million (ppm) of carbon-dioxide-equivalent (CO2e) in the mid-19th century to around 445 ppm today. If we continue at current rates of emissions, by the end of this century we would likely add at least 300 ppm, taking concentrations to around 750 ppm CO2e or higher. Here we are able to provide fairly robust probability statements. Such a path would bring somewhere in the region of a 50-50 chance of a warming of more than 5°C on mid-19th century levels, a temperature not seen on Earth for more than 30 million years. Homo sapiens has experienced nothing like this, being present for around only 250,000 years, and our agrarian civilisation only 8,000 or 9,000 years, since the emergence from the last ice age. The commitments embodied in the most recent global agreements to manage climate change, the UNFCCC Copenhagen Accord and Cancun Agreements, if implemented in full, would likely see a rise in global average temperature of 3.5-4°C (50-50 chance), a temperature not seen on Earth for more than 3 million years.

From here we are able to, and should, “speculate” on impacts from such levels of warming. Such warming, particularly warming of 5°C, would likely cause disruption on a huge scale to local habitats and climates, for example through flooding, desertification, erosion and water and food scarcity. Hundreds of millions of people, perhaps billions, would probably have to move, with the associated risks of severe and extended conflict.

Here we must explicitly recognise ambiguity (we are sure a 5°C temperature rise will have great impacts but we are unable to determine precisely how), and also intractability and indeterminacy – none of these predictions can be made with certainty. However, it is clear that the potential risks are huge, and the probabilities where they are available are not small – this is about risk management. Unmanaged climate change will put at risk the great advances in development of the last few decades, which have seen hundreds of millions rise out of income poverty, great improvements in health and life expectancy, and major advances in education and literacy. A high-carbon growth strategy is likely to destroy itself and is not a serious medium-term option.

The case for strong action is clear. The risks are immense. But there is also much uncertainty around action and this must be explicitly considered or we may arrive at the wrong conclusions on the scale of action necessary and timing of emissions reductions. The science can guide us on global emissions paths that will constrain the rise in global average temperature to less than 2°C with a reasonable probability (imprecision). The 2°C limit is the level accepted in international agreements, beyond which the risks of passing climate “tipping” points and other climate feedbacks are greatly increased. As a world we are already at close to 1°C above the mid-19th century level.

An emissions path that would give a reasonable chance (50-50) of staying below 2°C (with a 20% chance of exceeding 3°C – remember, this is a probability distribution), would require annual global emissions to fall from their current level of around 50 billion tonnes of CO2 in 2010, to around 44 billion tonnes in 2020, under 35 billion tonnes in 2030 and to well under 20 billion tonnes in 2050. This would see concentrations of greenhouse gases in the atmosphere peak somewhere above 500 ppm over the coming decades and decline from there.

Reductions in emissions on this scale require nothing short of an energy-industrial revolution across all countries and economic sectors. Research, development and deployment of low-carbon technologies will be crucial if we are to achieve these reductions. It is important here to consider uncertainty around technological advance. The presence of intractability around the consequences and prospects of progress in technology might be taken by some to suggest a “wait-and-see” approach on emissions reductions. That would be a profound mistake and it is crucial that this is clearly understood. First, there is a flow-stock process here, from emissions to increasing concentrations, and it is very difficult to reduce stocks of greenhouse gases on a major scale in the short or medium term (geoengineering is currently not a viable option due to costs and/or risks associated with the various technologies). Second, infrastructure and capital investment involves “lock-in” with much of the relevant capital having a technical lifetime of a few decades. Delay would result in the lock-in of vast amounts of long-lived high-carbon infrastructure, especially in the developing world.

Delay is very dangerous: we are already at a difficult starting point in terms of annual global emissions and concentrations of greenhouse gases and weak action or inaction for a decade would make the necessary emissions reductions and stabilisation of concentrations at acceptable levels much more problematic or impossible.

Managing climate change is a problem of risk management. Risk and uncertainty is core to arguments for action and must be explicitly addressed – the risks we face as a world are immense and delay is very dangerous. Now is the time for strong action.

James is a Research Officer at the Grantham Research Institute on Climate Change and the Environment. The views expressed are his own and do not represent those of the Grantham Research Institute. This work was prepared drawing on various materials from collaborations with Nicholas Stern. I alone am responsible for any errors. 

Image reproduced from and