April 2014 - The Met Office recently published a paper looking at improved skill in long-range winter forecasting for Europe and North America. Here, Professor Paco Doblas-Reyes, a long-range forecasting expert from the Institut Català de Ciències del Clima (IC3) in Spain, discusses the impact of the paper and the future for the global collaborative effort to improve long-range weather forecasts.
North Atlantic Oscillation (NAO) is the leading mode of atmospheric variability during North Atlantic-European winter (December-February).
It strongly influences the inter-annual variability of European surface temperature, wind and precipitation as it is associated with changes in the westerly flow reaching the continent from the ocean. It also governs changes in extreme weather events.
The quality of long-range, also known as seasonal, predictions for surface winter climate in the Euro-Atlantic sector has been limited so far. For instance, predictions of the winter NAO remained a hurdle for present dynamical prediction systems.
The recent publication by Scaife et al. (2014) of skilful winter NAO predictions using the most recent Met Office operational seasonal forecast system is a significant step forward in breaking this barrier.
The level of skill of the Met Office winter NAO forecasts has been rarely achieved in the past by dynamical forecast systems.
Although the reasons for this clear improvement in long-range prediction ability for Europe are still a matter of speculation, preliminary results suggest that it might be attributed to an improved resolution of both the ocean and atmospheric components of the climate model used, which allows for a more adequate representation of the physical processes at the heart of climate variability.
The adequate initialisation of the module that represents the sea ice and a revamped set of ocean initial conditions might have also played an important role. These hypotheses are supported by colleagues in experiments carried out with an independent climate forecast system, EC-Earth, where an increase in resolution leads to a substantial gain in winter NAO seasonal forecast skill.
Improvements such as this one do not happen easily in climate prediction. They are the result of sustained collaborative work between different communities working on climate modelling.
Climate prediction benefits from the work carried out by the groups that improve the different components that are part of a climate model: atmosphere, ocean, land surface, and sea ice.
Improvements are also assisted by the extraordinary efforts made to integrate the large amount of observations of the climate system into datasets that can be ingested by the climate model to start the prediction in a process known as analysis.
Global prediction centres around the world are continuously investing to integrate the best tools available to make skilful climate predictions, while international collaborative projects funded by the European Commission like SPECS contribute by promoting innovation and fostering the exchange of both tools and knowledge.
The recent increase in winter NAO skill is highly relevant to both the climate-prediction community, particularly those working in Europe, and the ecosystem of users that collaborate with it. An example of this renewed interest took place at the last meeting of the European Wind Energy Association in Barcelona in March 2014.
In the past three decades the development of seasonal predictions has been mainly justified by the proven ability to predict climate variations in the tropical regions, where they have been used to prevent disasters and increase resilience to the impact of extreme climate variability.
With the recent results published by Scaife et al. (2014), the community can, for the first time, seriously consider the possibility to engage with users of long-range climate predictions to co-design climate information and services for Europe months ahead. This is an aspect on which the project funded by the European Commission EUPORIAS is already focusing upon.
Much work lies ahead to, on one side, confirm that these improvements are robust across as many forecast systems as possible and, on the other side, explain the sources of the increased skill and improve it further.
However, these targets look more feasible when the path to achieve them is already lightened by encouraging results and a collaborative spirit.
Last updated: 11 April 2014