by Judith Curry
The Royal Society Discussion Meeting on Handling Uncertainty in Science, held 22/23 March 2010, played a seminal role in motivating me to investigate uncertainty in the climate debate.
The Proceedings of the Royal Society has a special issue for papers from the Meeting. The list of papers, abstracts and full text links (for a few of the papers) are found [here].
Table of Contents
TN Palmer and PJ Hardaker: Introduction Handling Uncertainty in Science [full text]
RM May: Science as organized scepticism [abstract]
HR Brown: Curious and sublime: the connection between uncertainty and probability in physics [abstract]
I Stewart: Sources of uncertainty in deterministic dynamics: an informal overview [abstract]
DJ Spiegelhalter and H. Riesch: Don’t know, can’t know: embracing deeper uncertainties when analyzing risks. [abstract]
J Slingo and T Palmer: Uncertainty in weather and climate prediction. [full text]
PJ Webster and J Jian: Environmental prediction, risk assessment and extreme events: adaptation strategies for the developing world. [full text]
D Aikman et al: Uncertainty in macroeconomic policy-making: art or science? [abstract]
JR Krebs: Risk, uncertainty and regulation [abstract]
G Wells, S Williams, SC Davies: The Department of Health perspective on handling uncerties in health sciences [abstract]
R Penrose: Uncertainty in quantum mechanics: faith or fantasy? [abstract]
P Campbell: Understanding the receivers and the recepton of science’s uncertain messages. [abstract]
O Peters: The time resolution of the St Petersburg paradox [full text]
Palmer and Hardaker
Some gems from Palmer and Hardaker’s Introduction:
However, we can go further than this. Let us accept that we cannot be certain whether or not anthropogenic emissions of greenhouse gases will lead to unquestionably dangerous changes to climate in the next century. However, consider the question: how probable would such dangerous climate change have to be for it to warrant some mitigating action now, to limit anthropogenic emissions? By ‘unquestionably dangerous’, we could mean the complete loss of the Amazonian rainforest owing to shifting rain patterns, or of large parts of Bangladesh becoming uninhabitable owing to persistently intense monsoons, storm surges and substantial sea-level rise, or of permanent Sahelian drought of the type seen in the 1980s? How probable before taking mitigating action can be justified: 50, 10, 1 or 0.1 per cent?
This brings us to the second part of the problem: making decisions in the light of uncertain scientific input. Better decisions can be made using predictions that have a properly quantified estimate of uncertainty, than using over-confident predictions with no estimate of uncertainty. But decisions can only be made easily if one can value the different probabilistic alternatives. In many situations, this may be a relatively simple economic matter. If wind speed exceeds 20 m s−1, a wind turbine typically cannot operate safely and therefore must be shut down. A probabilistic forecast of wind, including the probability that the wind speed will exceed 20 m s−1, can be converted into a rational and objective decision into how much electricity a wind farm should contract to produce (bearing in mind that the costs of buying electricity on the spot market should wind speeds exceed 20 m s−1, can be substantial compared with profits when wind speeds are within the operating range). But how do we value the loss of the Amazonian rainforest, or of large parts of Bangladesh, or of prolonged Sahelian drought? Clearly, it is not a purely economic calculation, but involves issues more directly related to human suffering and the destruction of things that we hold intrinsically dear to us. Estimating value in this generalized sense, including the thorny (and ultimately ethical) issue of whether the suffering of future generations should be somehow discounted, is clearly an extremely challenging issue for all of us. Nevertheless, these challenges should not deflect scientists and governments alike from ensuring that we are doing all that is humanly possible; firstly, to estimate uncertainties in future climate change as accurately as possible, and secondly to reduce these uncertainties—a large element of which lies in improving the computational representations of the equations of climate—wherever we can.
Slingo and Palmer
This paper provides an excellent overview of uncertainty in weather and climate predictions. From the concluding remarks:
This paper has considered how Lorenz’s theory of the atmosphere (and ocean) as a chaotic, nonlinear system pervades all of weather and climate prediction and how this has influenced the development of probabilistic ensemble prediction systems on all forecast lead times. It has also shown that the sources of uncertainty are not confined to the initial conditions, the basis of the Lorenz model, but that model uncertainty plays a critical role on all time scales.
It is important, however, to distinguish between model uncertainty that arises from imperfect knowledge of the real system, such as the representation of the carbon cycle, and uncertainty that comes from sub-gridscale phenomena that are understood quite well, but are inadequately represented because of the resolution of the model. In weather forecasting, there has been a continuous drive to higher-and-higher resolution with substantial benefits in terms of model performance and forecast skill. Furthermore, recent studies with ultra-high-resolution (approx. 3 km) global models, the so-called cloud system-resolving models, have shown a remarkable ability to capture the multi-scale nature of tropical convection of the type seen in figure 4 . However, the resolution of climate models, still typically 100 km or more, has been constrained fundamentally by a lack of computing resources , even though there is compelling evidence to suggest significant improvements in climate model performance with higher horizontal and vertical resolution in both the atmosphere and ocean .
Finally, Lorenz’s theory of the atmosphere (and ocean) as a chaotic system raises fundamental, but unanswered questions about how much the uncertainties in climate-change projections can be reduced. In 1969, Lorenz  wrote: ‘Perhaps we can visualize the day when all of the relevant physical principles will be perfectly known. It may then still not be possible to express these principles as mathematical equations which can be solved by digital computers. We may believe, for example, that the motion of the unsaturated portion of the atmosphere is governed by the Navier–Stokes equations, but to use these equations properly we should have to describe each turbulent eddy—a task far beyond the capacity of the largest computer. We must therefore express the pertinent statistical properties of turbulent eddies as functions of the larger-scale motions. We do not yet know how to do this, nor have we proven that the desired functions exist’. Thirty years later, this problem remains unsolved, and may possibly be unsolvable.
So how much will uncertainties in climate-change predictions of the large-scale reduce if models are run at 20, 2 or even 0.2 km resolution rather than say 100 km resolution? Equally, we may ask whether there is a certain resolution (e.g. 20 km), where it might be feasible to represent small-scale motions using stochastic equations, rather than trying to resolve them? These questions urgently need answering as the pressures grow on the climate science community to estimate, and if possible reduce uncertainties, and provide more reliable and confident predictions of regional climate change, hazardous weather and extremes.
Nevertheless, however much models improve, there will always be an irreducible level of uncertainty—‘flap of the seagull’s wings’—because of the chaotic nature of the system. Even the climate we have observed over the past century or so is only one realization of what the real system might produce.
Figure 12 shows 2000 years of El Nino behaviour simulated by a state-of-the-art climate model forced with present day solar irradiance and greenhouse gas concentrations. The richness of the El Nino behaviour, decade by decade and century by century, testifies to the fundamentally chaotic nature of the system that we are attempting to predict. It challenges the way in which we evaluate models and emphasizes the importance of continuing to focus on observing and understanding processes and phenomena in the climate system. It is also a classic demonstration of the need for ensemble prediction systems on all time scales in order to sample the range of possible outcomes that even the real world could produce. Nothing is certain.
Webster and Jian
Webster and Jian address applications of weather and climate projections to adaptation. From the abstract:
The uncertainty associated with predicting extreme weather events has serious implications for the developing world, owing to the greater societal vulnerability to such events. Continual exposure to unanticipated extreme events is a contributing factor for the descent into perpetual and structural rural poverty. We provide two examples of how probabilistic environmental prediction of extreme weather events can support dynamic adaptation. In the current climate era, we describe how short-term flood forecasts have been developed and implemented in Bangladesh. Forecasts of impending floods with horizons of 10 days are used to change agricultural practices and planning, store food and household items and evacuate those in peril. For the first time in Bangladesh, floods were anticipated in 2007 and 2008, with broad actions taking place in advance of the floods, grossing agricultural and household savings measured in units of annual income. We argue that probabilistic environmental forecasts disseminated to an informed user community can reduce poverty caused by exposure to unanticipated extreme events. Second, it is also realized that not all decisions in the future can be made at the village level and that grand plans for water resource management require extensive planning and funding. Based on imperfect models and scenarios of economic and population growth, we further suggest that flood frequency and intensity will increase in the Ganges, Brahmaputra and Yangtze catchments as greenhouse-gas concentrations increase. However, irrespective of the climate-change scenario chosen, the availability of fresh water in the latter half of the twenty-first century seems to be dominated by population increases that far outweigh climate-change effects. Paradoxically, fresh water availability may become more critical if there is no climate change.
Moderation note: this thread will be moderated for relevance. Relevant comments will discuss the papers referred to, and any other broader issues associated with uncertainty ignorance. Technical debates about the sun, surface temperature, or whatever should be on the older threads that discuss a topic relevant to your comment.