As we welcome everyone back after their summer break, EUCP is gearing up for the final few months of the project. It’s exciting to see so many of the plans we made in the early stages of the project realised! Over these next few months, we are focussing on how we can best exploit those scientific breakthroughs and reach those who can use them for wider societal benefit. In this newsletter, we’ll explain some of these breakthroughs and how we are working on the legacy of EUCP.

Decadal predictions (those running up to 10 years into the future) have greatly improved in the last decade thanks to worldwide coordinated efforts (such as WCRP-DCPP). These led to model improvements, advancements in initialization strategies (beginning the model simulation using observed data), and significant increases in the sizes of model ensembles. Thus, decadal predictions are now run operationally across the world (see WMO World Lead Centre) allowing the development of skilful decadal climate services (e.g. see C3S_34c) to aid socio-economic planning and adaptation. This welcome progress calls for further improvements to make decadal predictions even more skilful and capable of providing reliable climate information at the local scale, including predictions on climate extremes and user-oriented indices.

To achieve this goal, EUCP has shown that specific objectives must be met. Similar to seasonal forecasts, decadal predictions suffer from low signal-to-noise ratios, meaning that the respective climate models under-represent the predictable components of climate variability. This can lead to the climate change signal being overwhelmed by natural variability in the predictions. This deficiency is particularly relevant to predicting climate extremes and variability in the Euro-Atlantic sector. Another key problem to be tackled is the still-significant model biases, which not only affect the all-important climate variability patterns and their teleconnections across the world, but also impact climate dynamics after initialization in relation to the so-called model drift. We highlight below three possible avenues for improving decadal predictions.

Increasing the ensemble size of individual model systems, as well as using multi-system ensembles, has been shown to increase the predictive skill for various climatic indices and variables in most parts of the world. As an example, Figure 1 shows how the anomaly correlation coefficient (ACC) for the wintertime North Atlantic Oscillation and high-latitude blocking in the North Atlantic increases monotonically with the ensemble size of the Community Earth System Model – Decadal Prediction Large Ensemble (CESM-DPLE). Clearly, the skill is still far from saturating at ensemble size N=40.

Figure 1: The predictive skill for the CESM-DPLE ensemble-mean measured by the anomaly correlation coefficient (ACC) for high-latitude blocking (HLB) in a and the North Atlantic Oscillation (NAO) in c. Each cell below the diagonal corresponds to a different lead-year range defined by the start and end lead-years. The cyan markers (o) indicate non-statistically significant correlations. In a and c, an X marker indicates the lead-year range with the highest ACC (0.65 for HLB and 0.63 for NAO). In b and d, the respective skill is computed as a function of the ensemble size (averaged for all possible member combinations). Each line corresponds to a different lead-year range. Lines in color correspond to statistically significant correlations for the full ensemble (N = 40) following the same color code as in a and b. The dashed-dotted lines show the skill of the sub-ensemble mean against a single member of the ensemble (averaged for all possible combinations). Reproduced from Athanasiadis et al. (2020).

Such skill increases with the ensemble size occur thanks to noise cancellation accomplished via the ensemble averaging, which is more effective for larger ensemble sizes. The predictable component of the variability, in principle common to all members, survives the averaging, thus leading to higher skill.

On the other hand, recent studies have shown that increasing model resolution from about 1° (typical resolution of current decadal prediction systems) to 0.25° in the ocean has a significant positive impact on climatological biases across different HighResMIP models. This higher resolution allows the model to simulate much smaller-scale features. For example, Fig. 2 shows the respective Sea Surface Temperature bias reduction for a multi-model ensemble along with the associated bias reduction in European blocking, which is shown to be dynamically linked to the former. Such improvements in climatological model biases with increasing horizontal resolution are also expected to occur in decadal prediction systems (Roberts et al., 2020). Moreover, such increases in model resolution have been shown to bring improvements in mid-latitude air-sea interaction with a positive impact on the representation of climate variability (Haarsma et al., 2019; Bellucci et al., 2021; Tsartsali et al., in review) and near-term climate-change signal (Moreno-Chamarro et al., 2021).

Figure 2: Differences in wintertime absolute bias between low-resolution (LR) and high-resolution (HR) multi-model means. (a) SST biases (shading, in K) with the HadISST2 1950–2014 climatology in contours (°C). (b) blocking frequency biases (shading, in % of blocked days) with the ERA-JOINT climatological blocking frequency in contours (% of blocked days). Adopted from Athanasiadis et al. (in review).

Beyond the discussed impacts of increasing the ensemble size (improving the predictive skill) and the horizontal model resolution (improving model realism and variability), there are a number of other recommendations for the next generation of decadal prediction systems. Among these, the realistic initialization of all climatic components (ocean, atmosphere, sea-ice and soil moisture) in a dynamically consistent way – via the use of coupled reanalyses – is expected to reduce initialization shocks in the models. Such developments would help the adoption of a seamless prediction strategy.

References in review:

Athanasiadis, P., Ogawa, F., Omrani, N.-E. et al. (in review) Mitigating climate biases in the mid-latitude North Atlantic by increasing model resolution: SST gradients and their relation to blocking and the jet. Journal of Climate

Tsartsali, E., Haarsma, R., de Vries, H. et al. (in review) Impact of resolution on atmosphere-ocean VMM and PAM coupling along the Gulf stream in global high resolution models. Climate Dynamics

Information about the near-term future climate evolution over Europe is available from a range of sources. In particular, on the EUCP target time scales of the next 40 years, global initialised (decadal predictions) and non-initialised (projections) model simulations, together with regional modelling data, form the basis of our research in Work Package 5 of EUCP. Global decadal predictions are updated every year by making use of the latest oceanic and atmospheric observations for their initialisation and run one decade into the future. In contrast, global climate projections are run once (e.g. a few years ago for the latest Coupled Model Intercomparison Project, CMIP6) and cover all of the 21st century. High-resolution regional model simulations are available for selected time slices of future projections.

For a user interested in the likely climate over a specific geographical region, the plethora of climate data from these different sources of future model simulations can be confusing. The different data sets may not agree on the outcome and there is some ambiguity as to which data to use if there is overlapping availability. Within WP5, we have thus been studying how information from decadal predictions and climate projections can be combined such that the advantages of each can be gathered without losing valuable information. One area of research explored the feasibility of constraining the variability of ensembles of future climate projections using the latest decadal predictions. We were able to demonstrate that, in the case of predicting temperatures over the North Atlantic, the forecasts’ quality improved due to these constraints even after the 10 years for which decadal predictions were available (see here for more). Another novel constraining method developed within WP5 takes into account global patterns of climate variability. It was shown that this approach was able to improve climate projections of the following 20 years. In addition to constraining methodologies, work is also ongoing to explore the strengths, limitations and practicalities of seamlessly merging initialised predictions and non-initialised projections.

Publishing the ETH Zurich climate model weighting method (ClimWIP) as part of ESMValTool

Through its commitment to open science, EUCP also aims to provide a legacy of code for future research and applications. One of the methods used in several studies within EUCP is the Climate model Weighting by Independence and Performance (ClimWIP) method. It provides flexible options to weight multi-model ensembles of global climate models, such as CMIP6, based on each models’ performance in simulating observed climatological fields, for example. The ClimWIP method, as well as so-called recipes to reproduce results from several studies, have recently been included into the Earth System Evaluation Tool (ESMValTool) in version 2.3.0. The results in these studies can serve as a starting point for users to develop their own applications.

The implementation of ClimWIP into the ESMValTool provides a range of benefits for scientists (working with the method or further advancing it) as well as for users:

1. ESMValTool and the implementation of ClimWIP provide extensive documentation of the code and its functionality, including example results. This helps understanding, use, and further development of the method.
2. Using the code to calculate model weights for a specific application is greatly simplified by building on the framework provided by ESMValTool
   • The ESMValTool (and the ClimWIP method with it) can simply be installed via a Python package manager such as conda.
   • A regular release cycle allows easy access to updates and further developments of the method
   • Within EUCP, we have built a Python API for ESMValTool, so that the tool can also be used interactively in a Jupyter environment.
3. ESMValTool comes pre-installed on several ESGF nodes. Users can log in to these nodes directly, without having to download data or install the software themselves. These compute resources are provided through the IS-ENES consortium.
4. The ESMValTool provides input data quality checks as well as extensive options for automated pre-processing of the data (such as selecting time periods or regions, regridding, or averaging) which are then automatically passed to ClimWIP for further usage.
5. The GitHub infrastructure around the ESMValTool provides transparency, as well as ensuring code quality and reliability through an internal review process of the existing code and potential future additions.

Figure 3: Change in the weighted multi-model mean surface air temperature over land (2081-2100 minus 1995-2014)

MUF: Should we average sub-daily precipitation extremes? Are the indicators still reliable when they’re averaged over basins? Is there a criterion for the size of the basin? The capacity to manage large amounts of data is limited. How could this be technically managed?

EUCP: Having a tool that would allow extracting only the area of interest would be helpful. The users need to let us know clearly what they need so we could find the most appropriate approach. It has to be dealt with on a case by case basis.

EUCP: What timescales are of most interest for your work?

MUF: Seasonal to decadal scale for drought analysis.

EUCP: How should quantification of uncertainty be presented?

MUF (Water management): We need ranges of projections in the future. Then we choose from there and adapt strategies as we go on. We cannot wait though, because reservoirs take up to 15 years to construct so we have to start now.

MUF: How reliable is the decadal prediction of precipitation on a decadal time scale?

EUCP: There is skill if you carefully calibrate your forecast. So it is reliable, but it still depends on what exactly you need. For more information on this then research conducted by Verfaillie et al. (2020) should prove useful.

MUF: We need guidance to select good approaches for selecting information from climate projections.

EUCP: We work mainly on temperature and precipitation, but in theory we could work on something else. We would like to have input from users to provide more tailored information. We should be careful to avoid overly easy solutions, as each should be tailored depending on the user’s needs.

EUCP: Do you have access to ESGF and/or C3S?

MUF: Yes and we prefer C3S, because it is easier to find the data you need.

EUCP: What (other) way of accessing EUCP data or using EUCP methods would be convenient for you?

MUF: Jupyter Notebooks

●      Skillful decadal prediction of unforced Southern European summer temperature variations, Borchert et al., Aug 2021, Environmental Research Letters

Climate prediction on a decadal scale is very important for climate policy and planning, but chaotic variability within the climate system often reduces the skill of these predictions. This study assesses the decadal skill of the latest CMIP6 climate models at predicting the chaotic variability portion of European summer temperatures, as well as whether a link to North Atlantic temperatures, described here for the first time, can improve the predictions. The team found these models’ predictions are an improvement on those of older models, but they still rely on predicting external temperature forcing, with less skill for internal variability. The more-predictable variation in North Atlantic Ocean temperature can, however, be used to improve the European temperature predictions beyond the forcing. Improving our decadal climate predictions is a key part of informing effective climate policy, helping limit the impact of future climate change.

●     How does the CMIP6 ensemble change the picture for European climate projections?, Palmer et al., Aug 2021, Environmental Research Letters

Understanding the latest climate projections is important for keeping our strategies for addressing climate change in line with the latest science. This study compares projections from the latest generation of climate models, dubbed CMIP6, with older models. The team found that the newer models project greater increases in summer temperature and reductions in summer rainfall over Europe in the future than their CMIP5 predecessors. Greater global sensitivity and, in some regions, increased regional sensitivity are behind some of these changes. This may be due to improvements in the understanding of physical processes. These results are useful for updating our strategies for mitigating and adapting to climate change, ensuring that its future impact is limited.

●      Subtle influence of the Atlantic Meridional Overturning Circulation (AMOC) on seasonal sea surface temperature (SST) hindcast skill in the North Atlantic, Carvalho-Oliveira et al., Aug 2021, Weather and Climate Dynamics

The temperature of the sea surface is known to have an influence on weather and climate, including tropical cyclones. Predicting sea temperature variation in advance therefore has many benefits, however the mechanisms behind this variation are unclear. This study analyses how ocean circulation can be used to help predict sea temperatures up to 6 months in advance. The team found that circulation has the greatest discernible effect on sea temperatures 2-4 months later and has a stronger effect in more recent years. Accounting for circulation strength does improve sea surface temperature simulations, but only in the summer months. Knowing this, strategic use of circulation data could help us improve our seasonal simulations of sea surface temperature, potentially improving our predictions of some extreme weather events.

On February 8th 2021, we hosted our first workshop with the members of our Multi-User Forum (MUF), as previewed in our last Newsletter. This forum brings together representatives from public bodies and authorities, civil society organisations, businesses, financial organisations and academic institutions interested in climate information. Its aim is to enhance collaboration between these users of climate information and scientists to influence the design of the climate prediction system that EUCP is developing, and ensure it is as useful as possible.

The goal of this workshop was to identify areas where we can further improve the information we produce. The event was successful in bringing 65 participants together. Their feedback was extremely useful in highlighting areas that would make climate information more useable.

The workshop offered users different paths for further collaboration to progress these areas, and EUCP partners are looking forward to developing these engagements over the coming months.

The next EUCP user engagement workshop

At the last user engagement workshop in February 2021 (described above), several outcomes were discussed, including the project's legacy. Since then, several users have engaged with different members of the project to adapt the EUCP products, which is fundamental to producing seamless, authoritative climate information. On November 24th 2021, the next EUCP user engagement workshop will take place, which will be in line with the last event. The main goal of the workshop is to evaluate the progress of user engagement within the project. To do so, the workshop will showcase several new  examples of products produced in EUCP which we hope will fuel discussions on several items. These include usability within the decision-making process, research needs and research gaps.

We are looking forward to engaging with our users from the Multi-User Forum. Look for future editions of the newsletter for updates on the outcome of the workshop.

This event will be followed soon after, on November 26th, by a half-day mini-assembly, bringing together more members of the EUCP team to work on achieving the desired project outcomes.