by Judith Curry
The impact of climate change looms large as a deep uncertainty with global consequences. – Khalra et al.
The World Bank has published a new report: Agreeing on Robust Decisions: New Process for Decision Making Under Deep Uncertainty. This is an important report that lays out a new process to help decision makers better manage uncertainty and disagreement, particularly around climate change, by guiding them to the right decision making processes.
The general decision analytic framework that the World Bank has been using is highlighted in two previous posts:
The new aspects of decision making analysis presented in this report will be the subject of a future post at CE. In this post, I summarize the perspective that the World Bank is taking on climate uncertainty, and why it characterizes climate change as a decision problem with deep uncertainty. Excerpts from the new Report:
Continued efforts by climate scientists and others to increase knowledge about the climate and future climate scenarios are valuable. However, uncertainties about climate change and its impacts may increase as scientific inquiry diversifies and deepens. Therefore, decision makers should accept the irreducible uncertainty about the future climate and formulate adaptation and mitigation policies to manage it.
Box 2. Uncertainty in climate change projections
A cascade of uncertainties plagues climate change, and these uncertainties preclude prediction of the precise nature, timing, frequency, intensity and location of climate change impacts. The chain of increasing uncertainty begins with assumptions about the socio-economic characteristics of the global population, which determine the specification of a range of possible emissions scenarios. Estimates of climatic effects depend not only on the scenarios chosen but also on the configuration of the climate model used and existing knowledge of biophysical responses. Additionally, the farther into the future our projections, the greater the uncertainty. Uncertainty is also compounded by geographical resolution: uncertainty increases as the resolution of effects increases, from regional to country to local impacts. Even climate experts are unlikely to agree on a prediction of specific impacts of climate change. Many go even further in rejecting the specification of probabilities for climate change impacts because of the lack of repeated experiments, lack of independent observations, and the fact that all probabilities are conditional on a multitude of socio-economic and other developments.
APPENDIX 1. UNDERSTANDING CLIMATE UNCERTAINTY
This appendix summarizes key issues in understanding climate uncertainty.
Consider one example. Given the vulnerability of water systems to climate change, a Ghanian urban water manager would be wise to ask climate modelers to predict precipitation rates for the next 100 years, instead of relying on historical data. But using a climate model might be dangerously misleading: projections of future precipitation changes in the region are very uncertain. For Ghana, one model (CCSM3) predicts a 20% increase in precipitation, while another (GFDL) predicts a 30% decrease! It would be unwise for our water manager to tailor water management projects to either one of these or any other particular projection.
Such great uncertainties about future climate change stem from three major sources:
- Future emissions of greenhouse gases, which will shape future climate change. Future emissions, in turn, are driven by demographic and socioeconomic trends, technology, values and preferences, policies, which are also deeply uncertain. Scientists have developed emissions scenarios to capture a wide range of potential emissions trends that consider these diverse drivers.
- Scientific uncertainty and modeling limitations. These limitations are a result of our imperfect knowledge of the climate system and of the systems that climate, in turn, affects, such as lakes, glaciers and ecosystems. In particular, “climate sensitivity” refers to the increase in global mean temperature from a doubling of the CO2 concentration in the atmosphere. This sensitivity is uncertain.
- Irreducible natural variability. Global climate variables have their own dynamics linked to the chaotic behavior of the climate system.
These three uncertainties are sometimes referred to as policy, epistemic, and aleatory uncertainty, respectively. Their respective contribution to total uncertainty depend on the timescale and the spatial scale. At a global scale, and over the short term, natural variability and model response play the largest roles, and the emission a very small role; over the long term, the emissions dominate other sources of uncertainty.
It is thus critical not to over-interpret the difference between two climate scenarios run with different emissions or different models. The difference might be caused by aleatory uncertainty, with no significance. To rigorously interpret the difference between two scenarios, it is necessary to use ensembles, i.e. a sufficiently large set of simulations run with the same model and the same emission scenario. The spread of these simulations will represent the effect of natural variability as simulated by the model, and only differences that are robust to this effect can be interpreted as the effect of different emissions scenarios or of different models.
Also, it is critical to recognize that the spread across models do not represent the full uncertainty. All climate models use the same knowledge base and are based on the same basic methodologies. So it is very likely that all models share common biases, making the epistemic uncertainty larger than the differences across models. Thus, testing project robustness, looking outside the range of model results is advisable.
At a regional (or continental) scale natural variability is much more important regionally than globally, emission uncertainty plays a more moderate role, and climate model uncertainty remains large. This suggests that it is much more difficult to predict future climates when looking at one country or one region than globally, regardless of future progress in our understanding of climate change. Natural variability means that the climate signal is more difficult to extract (and – as already mentioned – forecasts of future climate remain out of reach).
There is also large uncertainty from differences between global climate models. The IPCC (2013) provides results from 42 global climate models. The models agree on the very big picture (more warming in high latitude than in low latitude; more precipitations in high latitudes; less precipitation around the tropics; more precipitation around the equator). However, the differences can be huge in some regions (e.g., half of the models predict an increase in precipitation over India; half of the models predict the opposite; and – as a consequence – the “average model” predicts no change, showing the risk of averaging projections).
When looking locally, we usually do not use global climate models. Instead, we use downscaling techniques which can be done with statistical tools or with regional climate models (RCM). Statistical methods use statistical relationships, calibrated on historical data, to relate large-scale drivers – which climate models can reproduce – to local phenomena – which climate models cannot reproduce. Even though our knowledge of the laws of physics helps select potential predictors, this method is not directly based on physical laws. Such statistical methods are computationally efficient and reproduce the current climate well. Statistical models, however, have two main drawbacks: first, they need long series of reliable data; second, even with a sufficiently large data range, it is difficult to know whether a statistical relationship will remain valid in a future climate.
To avoid the problem of validity of historical relationships, one may use physical models, which are based on physical laws. Of course, physical models often require calibration and bias correction, so the distinction between physical models and statistical models is sometimes fuzzy. Examples are Regional Climate Models (RCMs). Thus, statistical analyses are more reliable over the short to medium term, while RCMs are necessary to understand large warming over the long term. Nonetheless, in the long term RCMs remain driven by the input from GCMs, and so they do not resolve uncertainty related to climate variability, for example, that is produced by the GCMs.
In almost all cases, downscaling improves our ability to reproduce the current climate, but it does not reduce the uncertainty on future changes.
JC comment: The World Bank has a fairly aggressive program on climate change [link]. The WB program exists without the need for a high level of certainty about climate change or belief in the projections of global climate models. The WB acknowledges deep uncertainty in our understanding of climate change. But this uncertainty is not a reason for inaction.
Until U.S. and UN policy makers (and other national governments) begin to understand this, we will continue to have gridlock on climate policy, scientists will feel the need to be advocates, climate science will be politicized, and climate scientists will play the manufactured consensus game.
Kudos to the authors of the World Bank Report: Nidhi Khalra, Stefane Hallegate, Robert Lempert, Casey Brown, Adrian Fozzard, Stuart Gill, Ankur Shah.