It may seem like a washout, but in fact this is a pretty typical Irish summer, according to meteorologists. And it's likely to continue in the same vein. Blame it on the jetstream, writes CLAIRE O'CONNELL

IF YOU WANTED to cook up a pot of typical Irish summer weather from scratch, what recipe might you follow? Mix together a dash of sunshine and a sprinkle of blue sky, smear on a dollop of cloud, lash in some rain then whisk in a breeze and cook the the whole thing for months at a range of temperatures from annoyingly cool to a happiness-inducing warm.

The weather in Ireland is very variable. While we might yearn for long weeks of glorious sunshine, they rarely happen in our patch of the world. This year, April teased us with settled spells of sunshine, but by May the roller-coaster was back on track and the ensuing weather has been a mixed bag.

It’s just a typical Irish summer, says Peter Lynch, professor of meteorology at University College Dublin. “The main hallmark of the Irish weather is its undependability – it’s normally quite changeable,” he says. “Long spells of beautiful settled weather are just not typical so I don’t think there’s anything out of the ordinary about the weather now – it’s just the usual rather undependable, variable Irish summer weather.”

The jet stream has been a key to our variable summer this year, explains Peter O’Donnell from Irish Weather Online.

“The jet stream is a cylindrical feature in the mid-latitudes (mainly) where stronger winds are found about three to 10 kilomettres above sea level. In this region of the atmosphere, winds are normally quite strong, and travelling lows normally follow this path from west to east,” he says.

“The summer of 2011 has been variable – and looks to become more variable – because the jet stream is not locked into one position but has been meandering around between central France and near Iceland.”

UCD’s Prof Lynch agrees that that wandering waviness in the jet stream can keep our weather mixed up.

“It doesn’t come in a straight line as a solid westerly, it tends to waver in large, wave-like disturbances, going north and south at different longitudes,” he says. “These waves normally move along, so you get a change in the weather; you get a few good days and a few bad days.”

In April, that wave pattern stalled and we got a glorious high pressure sitting over the country, he explains, hence the good weather, but he notes that we can also end up under one of the upper air flow’s wave troughs, with low pressure systems hanging around and rain pouring down for weeks.

“What it would take for us to have a good summer is to have the upper air flow that we had in April occurring around about now – we would have a few weeks of lovely settled weather,” says Prof Lynch.

Our position, which means we bear the brunt of the Atlantic influence, is another factor that can literally cloud our weather, because the incoming winds can pick up moisture, explains Dr Kieran Hickey, a lecturer in physical geography at NUI Galway.

“We are dominated by south-westerly or westerly airflows, which tend to be moisture-laden because they are coming from the oceans,” he says. “And in summer time the oceans are a bit warmer so you tend to get more evaporation and hence more clouds.”

But what about that almost wintery spell earlier this month, which saw many of us staring out our windows in disbelief at the cold sheets of rain? “We were dominated then by northerly, north-westerly winds, which tend to be cooler, coming from the polar regions,” says Dr Hickey.

“When we get weather like that in the summer time it’s generally westerly or southwesterly winds, so they don’t feel as cold, even though they might be as wet.”

Getting the forecast right under such changeable conditions requires some nifty software and plenty of data to feed it. So how does it work in practice? “It all begins with getting observations of the weather,” explains Ray McGrath, who heads the research and applications division at Met Éireann.

He describes how shorter-range forecasts for Ireland are based on weather observations from the Atlantic and Europe: “You gather the observations and you feed them into a computer programme, so you are basically trying to build up a picture of what the atmosphere looks like that is consistent with the laws of physics and the various budgets for energy and mass.” There are gaps in the picture, much like a jigsaw puzzle that’s missing some pieces, and these gaps turn up particularly over the oceans where there tend to be fewer observations, explains McGrath – but you build up the picture on a grid in the computer model, and then move it forward over time.

“You step it forward in a way that’s consistent with the laws of physics, then you stretch out to maybe two or three days ahead,” says McGrath. “The forecaster would then take those forecasts and perhaps adjust them slightly – if there’s a storm coming in they may feel the computer system is bringing it in too quickly, and they would have the very latest information from satellites to see how it’s doing as well. And that essentially is your weather forecast.” But because of the chaotic nature of the atmosphere – that’s chaotic in the mathematical sense – initial flaws in the picture can quickly grow over time, explains McGrath.

“You might start off with something that looks very reasonable and you produce a forecast for an hour ahead that could be extremely good, but by the time you get out to 24 or 48 hours ahead, you may find the errors have begun to affect the patterns and the forecast quality mightn’t be very good at all,” he says.

“One way around this is to go back to your initial picture of the atmosphere and make some changes in areas where you feel the atmosphere is particularly sensitive. You might run 50 forecasts and look to see if there is a consistency: are they all going in the same paths or are they diverging after 24 hours?

Obviously if they are all suggesting we are looking at dry weather for the next two days then you would have good confidence in it. But if some of them were suggesting that a weather system could come in and affect Dublin then you would perhaps have less confidence in it.”

And while increases in computing power mean we can now get accurate forecasts to about a week ahead, McGrath reckons there’s a limit on how far we will be able to stretch it: “Will it rain in Dublin three weeks from now? I don’t think we are ever going to have accurate forecasts at that range because there is a fundamental limit – you have a chaotic system and ultimately that chaos tends to destroy the predictability.”

New model to help capture intense rainshowers that lead to floods

A NEW MODEL looks set to improve weather forecasts over local areas in Ireland.

The new Harmonie model, which has been in use here for a few weeks, is based on the existing model for Irish forecasts, but Harmonie goes into finer detail, according to Ray McGrath of Met Éireann.

"It's focusing on weather developments over a small area," he says. "Before we introduced this new model, the distance between the grid points was about 10km, but this has jumped down to 2.5 km, so you should get a much better forecast over smaller regions. Basically you are trying to pick out details of the weather at much finer scales compared to what we used to previously."

The new model, which is being run in collaboration with the Irish Centre for High-End Computing, could offer new insight into events such as flooding, notes McGrath.

"This is something we would hope to do better in the future with the Harmonie system, to be able to capture these intense rainshowers that can lead to flood in rivers," he says. "We would hope to do better with extreme weather with the new model."

Peter Lynch, professor of meteorology at UCD, agrees that setting the grid points closer together brings an improvement .
"If you think about the Wicklow Mountains and the way they vary, you are going to get a better representation of the surface and of the coastline," he says, " and you are also going to represent the physical processes like large cumulus clouds, cumulonimbus complexes – they are of the order of a few kilometres in scale – so you start to represent these systems more accurately."