Over the past few weeks the media have carried many stories suggesting that the coming winter will be especially severe.
With this being Science Week it's a good time to ask -- is there any science behind this talk, or is it all pure speculation? To answer that question, we have to look at how weather forecasts are made.
The old folklore-based weather predictions were based on the signs seen in nature.
Today's weather forecasts are no different in that everything starts with observations. These days the weather observations come from automated weather stations, from aircraft, from weather balloons, radar and satellite. We can now form a much more accurate picture of what the atmosphere is "doing".
High-speed communications enable us to rapidly gather weather information from every country in Europe, and from across the Atlantic to the Americas. This information is fed into powerful computers that contain a mathematical "model".
Using this "model" we can calculate how all the weather elements (high pressure regions, low pressure regions, cold air, mild air, frontal systems) develop in time, and this forms the basis for the day-to-day weather forecasts.
The more accurate our starting point, the better the forecasts will be.
However, we can only ever know an approximate starting point; the atmosphere is too vast and complex to allow us to form a perfect picture.
As we look further ahead Chaos Theory gradually takes over and the forecast diverges from reality. There is an absolute limit on how far we can forecast ahead with this method, and it's thought to be about 10 days.
In attempting to look further ahead, we rely on our knowledge of what is happening in the oceans.
The atmosphere and oceans are closely linked. The rain falling over Ireland comes from water that evaporates from the oceans; the warmer the surface of the ocean, the more water evaporates and the heavier the rain will be when it eventually falls.
The oceans have warm and cold currents, just like the atmosphere. The difference is that change in the oceans happens much more slowly and, in some parts of the world, in a more predictable fashion.
If we get a good understanding of how the oceans will behave over the coming month and longer, we have some basis for inferring the weather patterns. Not the day-to-day detail of weather but the larger patterns as to whether it will be warm or cold, wet or dry, windy or calm.
This method has been used successfully in some regions, notably some countries bordering the Pacific and Indian oceans.
However, these oceans have large and well-understood water currents, which affect the atmosphere. The Atlantic has no such large phenomena that change regularly and the changes that do occur and not easy to predict. Therefore, inferences we can make about seasonal forecasts are weak and result in forecasts not correct often enough to be much use.
So, scientifically, it is not possible to make any confident forecast of the coming winter.
The "average" winter remains the most likely outcome.
So where do all the predictions of a severe winter come from? From people who do not understand the complexity of the problem, and who make simplistic assumptions.
From people who specialise in speculation, not science.
CAPTION: coming back?: Snow earlier this year