Gabriele Hegerl, Andrew P. Ballinger, Ben Booth, Leonard F. Borchert, Lukas Brunner, Markus Donat, Francisco Doblas-Reyes, Glen Harris, Jason Lowe, Rashed Mahmood, Juliette Mignot, James Murphy, Didier Swingedouw, Antje Weisheimer
Climate decisions that span climate predictions (which run for a few years into the future) and projections (which run for decades) will benefit from consistent information across these two timescales. This paper provides a review of how climate observations are used in both systems. The authors provide both a first perspective that exposes the challenges we will face in using observations along the way to seamless annual to centennial climate projections, and a review on how climate observations are currently employed in the climate science community. The analysis shows that evaluating climate simulations against observations can be effective in reducing uncertainty and coming to more confident projections. The study uses this overview to provide the first perspective on the challenges we will face if we want to include observational information in a seamless climate prediction system.
Climate model simulations are often split into predictions, which run for up to a few years ahead, and projections, which run for many decades. Combining these into a single, seamless future climate prediction would greatly help efforts at mitigating and adapting to climate change. Research is ongoing on how to combine predictions and projections, but to date these approaches have not considered how to combine information gained by both respective research communities from evaluating their simulations against climate observations. Both predictions and projections use climate observations to fine-tune (or constrain) their results, but the different ways these are used makes the two difficult to merge. This paper discusses the ways in which climate predictions and projections use observed climate data and the effect on their results. Understanding the effects of observational constraint may help advance work on combining climate predictions and projections, a key goal of EUCP.
Observational constraints on climate projections have effectively reduced biases in model ensembles in several prior studies and projects. This field is still very much a young one, in the longer-timescale climate projections in particular. Whilst the UK has made use of observational analysis to reduce uncertainty in its own national climate projections (e.g. UKCP) this is not yet common outside the UK. Different approaches are being pioneered, a number of which are showcased, and their differences discussed in this paper. In contrast, shorter-timescale climate predictions are routinely evaluated against observations, but the evaluated skill captures the ability of these predictions to capture short-scale variability, where skill can vary on lead times (how far in advance a forecast is issued) and on the phase of observed variability. This paper illustrates the state of the art in both fields and discusses some of the issues involved if we are going to incorporate the benefits of these evaluations into one seamless system.
This study provides an overview of the current use of climate observations to create more confident climate predictions and projections. Currently climate predictions (that constrain near-future natural variability) and projections (that constrain the climate response to human emissions) are two distinct sources of information provided to users. Climate datasets that provide consistent information across these two timescales are eminently desirable, but so far questions of how to combine observational constraints have not been explored. This review paper exposes some of the challenges that we expect to face when producing more seamless information using predictions and projections. This enables the first discussion on how information from observations, currently used in the distinct systems, can be combined. These efforts are seen as prerequisite to accounting for prediction skill when merging climate predictions and projections into seamless future climate simulations out to 40 years in the future.
A merged system of climate predictions and projections would have several benefits for policymakers. Decisions could be made using a single consistent climate dataset, without having to change between different sets of model outputs that may have different levels of detail or uncertainty. Climate information such as this, consistent with state-of-the-art near-term climate predictions and longer-term projections, will support effective decision-making on mitigating and adapting to the impacts of climate change, especially those decisions that must span these two timescales.
Observations facilitate model evaluation and provide constraints that are relevant to future predictions and projections. Constraints for uninitialized projections are generally based on model performance in simulating climatology and climate change. For initialized predictions, skill scores over the hindcast period provide insight into the relative performance of models, and the value of initialization as compared to projections. Predictions and projections combined can, in principle, provide seamless decadal to multi-decadal climate information. For that, though, the role of observations in skill estimates and constraints needs to be understood in order to use both consistently across the prediction and projection time horizons. This paper discusses the challenges in doing so, illustrated by examples of state-of-the-art methods for predicting and projecting changes in European climate. It discusses constraints across prediction and projection methods, their interpretation, and the metrics that drive them such as process accuracy, accurate trends or high signal-to-noise ratio. We also discuss the potential to combine constraints to arrive at more reliable climate prediction systems from years to decades. To illustrate constraints on projections, we discuss their use in the UK’s climate prediction system UKCP18, the case of model performance weights obtained from the Climate model Weighting by Independence and Performance (ClimWIP) method, and the estimated magnitude of the forced signal in observations from detection and attribution. For initialized predictions, skill scores are used to evaluate which models perform well, what might contribute to this performance, and how skill may vary over time. Skill estimates also vary with different phases of climate variability and climatic conditions, and are influenced by the presence of external forcing. This complicates the systematic use of observational constraints. Furthermore, we illustrate that sub-selecting simulations from large ensembles based on reproduction of the observed evolution of climate variations is a good testbed for combining projections and predictions. Finally, the methods described in this paper potentially add value to projections and predictions for users, but must be used with caution.