by Judith Curry
We know that climate change is a problem – but how big a problem is it? We have to answer this question before we can make a good decision about how much effort to put into dealing with it.
Climate Seer James Hansen Issues His Direst Forecast Yet. Jim Hansen has submitted a new paper for publication (can’t find a copy yet). But it is already making a splash in the media. Excerpts:
James Hansen is now warning that humanity could confront “sea level rise of several meters” before the end of the century unless greenhouse gas emissions are slashed much faster than currently contemplated.
This roughly ten feet of sea level rise—well beyond previous estimates—would render coastal cities such as New York, London and Shanghai uninhabitable. “Parts of [our coastal cities] would still be sticking above the water,” Hansen says, “but you couldn’t live there.”
Is Hansen’s forecast plausible? Even Mann is skeptical (see his quote in WaPo article). What does ‘plausible’ mean? What is the plausible worst case scenario for human caused climate change?
Climate Change: A Risk Assessment
I haven’t found climate change risk assessments to be very satisfactory, for a range of reasons. There is a new report out, entitled Climate Change: A Risk Assessment. IMO this is far and away the best risk assessment for AGW that I have seen. That said, it is far from perfect, for reasons described towards the end of the post (largely associated with how much warming we can expect, which is obviously the key issue). But IMO it has appropriately framed the climate risk assessment problem, and its authors (for the most part) don’t seem to have any obvious agenda beyond . . . risk assessment.
From the summary on the Cambridge web site:
This report argues that the risks of climate change should be assessed in the same way as risks to national security, financial stability, or public health. That means we should concentrate especially on understanding what is the worst that could happen, and how likely that might be.
The report presents a climate change risk assessment that aims to be holistic, and to be useful to anyone who is interested in understanding the overall scale of the problem. It considers:
- What we are doing to the climate: the future trajectory of global greenhouse gas emissions;
- How the climate may change, and what that could do to us – the ‘direct risks’ arising from the climate’s response to emissions;
- What, in the context of a changing climate, we might do to each other – the ‘systemic risks’ arising from the interaction of climate change with systems of trade, governance and security;
- How to value the risks; and
- How to reduce the risks – the elements of a proportionate response.
Some excerpts of relevance to formulating the plausible worst case scenario:
In assessing the risk of climate change, the immediate questions for any country anywhere in the world are: How serious is the threat? How urgent is it? How should we prioritise our response, when we have so many other pressing, national objectives – from encouraging economic recovery to protecting our people around the world?
A risk assessment asks the questions: ‘What might happen?’, ‘How bad would that be?’ and ‘How likely is that?’ The answers to these questions can inform decisions about how to respond.
Climate change fits the definition of a risk because it is likely to affect human interests in a negative way, and because many of its consequences are uncertain. We know that adding energy to the Earth system will warm it up, raising temperatures, melting ice, and raising sea levels. But we do not know how fast or how far the climate will warm, and we cannot predict accurately the multitude of associated changes that will take place. The answer to the question ‘how bad could it be?’ is far from obvious.
Consider the full range of probabilities. The biggest risks could lie anywhere in the probability distribution . . . particular importance of not ignoring low probability, high impact risks. It is a matter of judgment how low a probability is worth considering. When a probability cannot be meaningfully quantified, it is usual to consider a ‘plausible worst case’. Again, the question of what is a relevant threshold of ‘plausibility’ is a matter of judgment.
What is the probability of following a high emissions pathway? Based on an analysis of current policies and plans for major countries and regions, it is very likely that the world will continue to follow a medium to high emissions pathway for the next few decades. If goals for reducing emissions in the EU and the U.S. and stabilizing emissions in China are achieved, then the highest emissions scenarios are less likely to occur, especially if India is able to displace a part of its anticipated construction of new coal-fired power plants with renewable energy capacity. But this will only keep emissions on a moderate trajectory, still far in excess of what is required to limit the impacts of climate change below a harmful level.
How likely are we to exceed the temperature thresholds we’ve identified? For any emissions pathway, a wide range of global temperature increases is possible. On all but the lowest emissions pathways, a rise of more than 2°C is likely in the latter half of this century. On a medium-high emissions pathway (RCP61), a rise of more than 4°C appears to be as likely as not by 2150. On the highest emissions pathway (RCP8.5), a rise of 7°C is a very low probability at the end of this century, but appears to become more likely than not during the course of the 22nd century. A rise of more than 10°C over the next few centuries cannot be ruled out.
Basically, they are using the IPCC AR5 here. The ensemble of climate model simulations for different emissions pathways should be regarded as individual scenarios of future climate change, and probability cannot be meaningfully quantified; see these previous CE posts:
- How should we interpret an ensemble of climate model simulations?
- Worst case scenario versus fat tail
‘Plausible’ versus ‘possible’
I have argued previously that the probabilities and even the likelihoods determined from climate model simulations are misleading. Which brings us back to ‘What is the plausible worst case scenario?’
‘Plausible’ seems a plausibly useful word, but one that is ambiguous. A dictionary definition of plausible: falls within the limits of what might conceivably happen. How is ‘plausible’ different from ‘possible’?
Financial risk management has long grappled with these issues. The ASU-Oxford Plausibility Project has an interesting and relevant essay: How plausible is plausibility as a scenario effectiveness criterion? The essay doesn’t come up with a definitive solution, but addresses the challenges of navigating between the too plausible and too implausible extremes.
I propose the following distinction between ‘possible’ and ‘plausible’:
Plausible: a future scenario that cannot be rejected or falsified based on background knowledge. Scenario falsification was discussed in a previous CE thread. Update: The idea of plausibility further rests on an assessment of uncertainties in the background knowledge.
Possible: a future scenario that is not consistent with background knowledge, but given the uncertainties in our knowledge can be regarded as possible.
Hence a plausible scenario is possible, but a possible scenario may not be plausible. As an example, I regard Jim Hansen’s sea level rise scenario as possible but not plausible (with the caveat that I haven’t read the paper and with the assumption that the paper presents no new observational evidence). Most alarming scenarios fall in this category of possible but not plausible.
Policy responses generally focus on the plausible scenarios, while continuing to assess the plausible worst case scenario and the plausibility of the far-out possible scenarios. There is a practical and philosophical issue of plausible versus possible ‘ruin’ scenarios (see previous CE post).
Plausible worst case scenario
I regard there to be two main elements to formulating the plausible worst case scenario for human-caused climate change:
- The highest plausible representative concentration pathway (see previous CE post)
- The highest plausible equilibrium climate sensitivity (see previous CE post)
The highest representative concentration pathway formulated by the IPCC is RCP8.5. In a previous post Coal and the IPCC, Dave Rutledge argued that RCP8.5 was not a plausible scenario. I just spotted a post by Blair King entitled On RCP8.5 and the business as usual scenario, which also challenges the plausibility of RCP8.5. I suspect there are other challenges to the plausibility of this scenario, but I haven’t really looked.
With regards to climate sensitivity, see my previous post Climate sensitivity: lopping off the fat tail. Excerpt:
Gregor Betz defines modal falsification as follows [BetzModalFalsification]:
Modal falsification: It is scientifically shown that a certain statement about the future is possibly true as long as it is not shown that this statement is incompatible with our relevant background knowledge, i.e. as long as the possibility statement is not falsified.
Nic Lewis’ research arguably falsifies the high values of climate sensitivity determined from instrumental data, owing to problems with the statistical methodology and the forcing data. IMO, Nic’s methodology for determining climate sensitivity from observations and energy balance model represents the best current method.
With regards to climate model determinations, see this paper: The upper end of climate model temperature projections is inconsistent with past warming, by Peter Stott, Peter Good, Gareth Jones, Nathan Gillett, Ed Hawkins, published in Environmental Research Letters (open access) [link]. This paper does not directly address the issue of climate sensitivity. It is obvious from comparing climate model simulations with observations that most climate models are running too hot for the early years of the 21st century.
It is becoming increasingly difficult to defend values of ECS that exceed 3.5C. The threshold warming values identified in the Climate Risk Assessment Report of 4C and 7C by 2150 are possible but they may not be plausible. Since this warming is inferred by climate model simulations, falsification of these scenarios could be accomplished by comparing modeled and observed warming for the period since 1950, and assessing whether the equilibrium climate sensitivity for that model is too high (with a value of aerosol indirect effect that is far too high).
The plausible worst case scenario is arguably where we should focus our efforts (both science and policy). Working to falsify high values of RCP and sensitivity based on the background knowledge that we do have, should be a high priority.
I think the new Climate Change Risk Assessment document is an important step forward in framing how we should approach climate change risk assessment. The appropriate focus is the plausible worst case scenario, which has not been a focus of the IPCC or climate establishment. Instead they have focused on a mean, a likely range, or an alarming possibility.
I have proposed scenario falsification as a way to proceed, in terms of identifying plausible versus possible scenarios. I hope that risk assessors, philosophers and climate scientists can work together and shift their focus to scientifically robust and policy useful strategies for assessing the plausible worst case scenario for human caused global warming.