Scientists from the Bjerknes Climate Prediction Unit, affiliated with the Nansen Environmental and Remote Sensing Center, the Bjerknes Centre for Climate Research, and the University of Bergen, contributed to a recent publication in Nature. The results indicate that it is possible to predict how the atmospheric circulation above the North Atlantic will evolve during the next decade. This is crucial for better predicting the winters in Europe and Eastern North America.

Investigating the climate of the past

In order to look forward in time, looking at the past is helpful. This is true in many cases, and the researchers behind this study led by the UK Met Office made use of this principle. They used climate models for investigating how accurately climate can be predicted on a decadal scale over the past sixty years.

Sea level pressure above the North Atlantic influences Norwegian winters 

Rainfall variation over Northern Europe between 1960 and 2005: e) shows observations (black) and modelled predictions (red) with uncertainty range (shaded red) without adjustments, f) shows the improved and adjusted modelled predictions and uncertainty range.

The main pattern of changes in sea level pressure above the North Atlantic, called the North Atlantic Oscillation (NAO), influences the wind and storms over the North Atlantic, which in turn influences the winter weather in Europe and Eastern North America. Two extremes are possible for winters in these regions: stormy, warm, and wet, or calm, cold, and dry. Which extreme the winter weather will tend towards is now shown to be very predictable on a decadal scale, according to the new study.

The researchers investigated the North Atlantic Oscillation and its influence by producing retrospective forecasts of the past climate (called hindcasts) and comparing them to observations made in the past. That way they quantified how accurate the model predictions are.

One of the most important predictions for Europe and especially Norway is the amount of rainfall. The comparison between hindcasts produced by models (Figure f, red line) and the observation (Figure f, black line) shows that the rainfall over Northern Europe can be predicted with high certainty. The model results match the previous observations nicely.

Contribution from the Bjerknes Climate Prediction Unit

Many hindcasts were produced by different research groups worldwide. The different climate models from these groups are part of the Coupled Model Intercomparison Project (CMIP) experiments performed for the last and upcoming Intergovernmental Panel on Climate Change (IPCC) reports. Bergen researchers involved in the study are the following: Noel Keenlyside (UiB/NERSC), François Counillon (NERSC), Ingo Bethke (UiB), and Yiguo Wang (NERSC). The four are part of the Bjerknes Climate Prediction Unit at the Bjerknes Centre for Climate Research. They used their climate model, the Norwegian Climate Prediction Model (NorCPM), which is part of CMIP6, to contribute to this study.

Climate models need to be improved

Apart from the high predictability of the North Atlantic climate indicated by the hindcasts, the study also shows that current climate models are underestimating this exact fact (Figure e). The researchers identified this deficiency and show that climate models need to be and can be adjusted (Figure f) to better predict the behaviour of the pressure above the North Atlantic and in turn the future winter conditions in Europe and Eastern North America.

To sum it up, confidently predicting the winters of the next years for Norway is now a reality, but climate models need to be improved.

Significance of this study: Climate can now be better predicted on short time scalesNoel Keenlyside UiB/Bjerknes Centre for Climate Research/Nansen Center. © Thor Brødreskift, UiB

Noel Keenlyside, leader of the BCPU, commented “This is a major breakthrough for climate research and for the development of climate services in our region. Now we have solid evidence that we can provide to our stakeholders, like BKK and Agder Energi, that we can really say something useful about how the coming winters will be. It will also lead to improved models for providing better long-term projections of climate change.” 

Erik Kolstad (NORCE/Bjerknes Centre), who will lead Climate Futures, a newly established Centre for Research-Based Innovation (SFI) for climate risk research, says “These results show that the climate models now can predict climate on a level that makes it valuable for planning within several sectors, including renewable energies, agriculture, and finance/insurance. With such predictions both businesses and the public sector can be better equipped to handle extreme weather and potentially make use of periods with favourable weather and climate.”

Tarjei Breiteig, a meteorologist at Agder Energi AS, represents one of the stakeholders this study directly impacts: “In this study, the researchers show that there is still a lot of untapped potential in saying something about the possible weather and climate for the next decade. In order to be able to save hydropower in a year with little demand, and to have enough renewable hydropower saved up in the years when demand is high, it is absolutely essential for us to have sufficient information about which fluctuations can be expected in weather and climate in the next decade. The climate research community in Bergen shows here that they take this issue seriously, and that they are far ahead when it comes to interpreting and using climate models in practice.”

Also published in Norwegian at the Bjerknes Centre.

BT wrote about it as well: Ny forsking kan gi gode klimavarsel ti år fram i tid.

Reference:

Smith, D.M., Scaife, A.A., Eade, R. et al. North Atlantic climate far more predictable than models imply. Nature 583, 796–800 (2020). https://doi.org/10.1038/s41586-020-2525-0