In some years the score reached 90 of cases

Meteorology can one day predict time to more than a week The question may seem provocative to you this sometimes annoying science for hikers, winemakers or barbecue enthusiasts. Its progress but were constant over the past decades with a forecast day won every ten years. The newsletter within a week today marks the boundary of that know how to do the meteorologists. Beyond that, the chaotic nature of time degrades the probability of modelling. Researchers turn simulations to more than ten days, but scores have no great interest: they are little better than the heads or tails.

Christophe Cassou, climatologist with the European Centre for research and training in scientific computing (Cerfacs), has yet to push the horizon of this science. In a recent study published in the journal "Nature", he has explored this fuzzy area between climatology, predicting long term science and meteorology. "European models have so far neglected tropical fluctuations." It influences our time in longer time scales, it cannot be taken into account by the statistics. However we have now thirty years since the appearance of satellite data. "These data sets are starting to become interesting to use," says Christophe Cassou.

Comfortable in winter models

The weather in our region varies generally four scenarios. The first two are played by a well known duo of all viewers: the depression of Iceland and the anticyclone of the Azores. The NAO , the most common occurs where the depression of Iceland Northern Europe for low pressures, strengthening its compère of high pressure to the South. In winter, this plan diverts storms to the North. The Ch'tis, their neighbours and Scandinavia suffer humidity and winds but advantage of milder temperatures. Southern Europe and the Mediterranean basin live dry days. In the periods NAO, the duo momentum, storms are less violent but they confuse to the South of Europe.

Third plan, blockade involves a high on the North Sea that rejects the Siberian cold to Western Europe. The France rule then the radiators thoroughly as in NAO plans. In regime of Ridge, the Azores anticyclone ventured very North of habitual residence and covers Europe of a wave of drought and of weather.

Statistical link

Researchers have successfully for decades in many model each of these plans, with the average life time of a week. The models are very uncomfortable in winter when one of these plans moved comfortably to Europe, they know well to calculate regional implications. The problem is the transition between two regimes. Meteorologists knew not far if the transition between these four scenarios was randomly or if external causes pulled the strings.

Christophe Cassou just found a statistical link between these oscillations and variations of the time in the tropics. More precisely, he discovered a probabilistic relationship between European regimes and the oscillation of Madden-Julian (OMJ) along the equator. The OMJ is a disturbance that moves from East to West around the Earth and the Equator at a rate of thirty to seventy days. In two points of his career, West of the Pacific and West of the Atlantic, she met the 8x12m. Christophe Cassou discovered that this global current can be used as a conveyor belt to the European sky. From the Western Atlantic, the OMJ can enable to weeks of interval type NAO European storms. NAO plans are activated in the Eastern Pacific.

The researcher is amused to assess the average prediction score on thirty years of a newsletter to twelve or thirteen days. In 75 of cases, he knows how to predict the sign that will be the NAO. In some years, the score reached 90 of cases. And this is the problem that still separates its search for operational use: "this approach works very well in the months where the OMJ is very active, but at certain periods, the oscillation is soft and little influence plans." "Gold is not determine in advance the OMJ force", he said.

This is why Christophe Cassou encourages meteorologists to accentuate their research on this still poorly understood oscillation. An issue shared with climatologists, whose models simulate poorly tropical phenomena such as the OMJ. "With its periods of thirty to seventy days, it is located in full in the blurred area between weather and climate science.". "The two communities have understood the interest to work together, their reconciliation is currently fast enough," concludes, optimist, Christophe Cassou.