KEY POINTS:
Weather ambassador? Bob McDavitt came up with his own job title in the mid 90s after scouring a thesaurus. Ambassador sounded better than "forecast manager north" - especially considering he didn't have any staff to manage.
McDavitt says he's representing MetService to the people in Auckland - albeit from a tiny office at the back of a bookstore in Westhaven, that's far from ambassadorial.
Doesn't his title suggest he's representing the weather? "Oh well, yes, I act as a go-between for the weather and its users. I can't communicate very well with the weather, but I can communicate with the users of the weather."
Weather users? "The main thing is to make sure people who are using the weather forecasts are aware they are forecasts - that they come from isobar land."
Isobar land? "People live in the real world. We don't. We live in the world of models - so once you have a concept of isobar land you won't have any gripe with forecasters." Right. But why is it so many do? To answer these and other questions, such as what's in store weather-wise this summer, the Herald asked three weather forecasters about how they look into the future. Be warned, isobar land can be a strange and shifting landscape.
What is the outlook for this summer?
Niwa, our National Institute of Water and Atmospheric Research, says above-average temperatures and near normal rainfalls are the most likely outcome across most of the country. The west of the South Island may be a little wetter and there's likely to be less rain than normal in the east. Soils are drier than normal in many eastern regions at the moment and eastern regions could well stay drier than normal over summer.
That only makes sense if you know what normal is - what is it?
Principal Scientist for Climate Variability and Change at Niwa, Dr James Renwick, says "normal conditions" for climate scientists have a very specific meaning - close to the average over the past 30 years of the climate in a particular place.
In the greater Auckland city region, for example, total rainfall averages around 230mm in summer. Rainfalls within about 20 per cent either side of that are in the normal range. Average summer daytime maximum temperatures are around 23C and overnight lows about 15C. Temperature within about 0.3 degrees of the average are in the normal range. Last summer, for example, Auckland's mean temperature of 19.9C was above normal, at plus 0.7C. Whereas average rainfall of 176mm was below normal - 24 per cent below in fact.
In Christchurch summer rainfall of 130mm or so, and temperatures of between 12C and 22C are typical. Last summer Christchurch's mean of 16.9C was plus 0.3C making it just above average. Rainfall of 188mm however, was 144 per cent of normal - well above normal.
So in Auckland, for example, we can expect it to be warm like last summer, but unlike last summer which was quite dry, it's going to be normal rainfall, which means quite wet - right?
"When we say temperatures are likely to be above normal, what that means is likely," says Renwick. "You cannot make a categorical seasonal forecast because of all the chaotic effects. You can say things are tending in a certain direction, but you can never rule out rogue events - things diverging from the average picture predicted by the computer models." Weather ambassador McDavitt is more succinct: "A forecast is a forecast is a forecast - end of argument.
We don't pretend it to be anything else; it is definitely not a reality."
What then, can you predict?
"A fair fraction of the variation in the climate is unpredictable because of chaos effects, so you can't predict the day-to-day sequence of the weather more than a week or ten days in advance," says Renwick.
"So you have to approach the whole thing probablistically. When computer models with all the physics in them are run, they are run many times from slightly different starting conditions. You just tweak the inputs - employing the whole idea of chaos theory, then average to smooth out the chaotic part."
What does chaos theory have to do with it?
Renwick is talking about "the butterfly effect" - the idea that some physical processes such as the weather or the climate system are very sensitive to subtle changes. Blame Edward Lorenz, who, in 1961, had created an early computer program to simulate weather. One day he changed one of a dozen numbers representing atmospheric conditions, from 0.506127 to 0.506. The tiny alteration utterly transformed his long- term forecast, a point Lorenz amplified in his 1972 paper, Predictability: Does the Flap of a Butterfly's Wings in Brazil Set off a Tornado in Texas?
"Tiny little changes in the state of the weather now can lead to huge differences to what the weather does in three or four days," says Renwick.
"The chaos is caused by everyday life - the hubbub of background change," says McDavitt. "Two cars passing each other on the motorway create a little whirlwind of air - a mini vortex. Most of the time it just fades away as noise does. But sometimes the atmosphere is in feedback mode. It will take any small thing and feed on it and create something bigger. You only need to light up a cigarette or close a door and you've jiggled the pattern - you've taken it on a new path."
So you can light a fag and change the weather. Do forecasters really believe this stuff?
Claiming a butterfly's wings can cause a storm, is another way of saying we simply don't know what caused the storm. Nor can we possibly know all the slight connections between small events and big effects. But forecasters do incorporate the notion of chaos effects into their computer modelling.
A number of forecasting centres around the world run dynamical climate models - essentially a big computer crunching measures of circulation of the atmosphere and the oceans, winds, temperature and rainfall - collected in real time. Using laws of physics, the models are run forward again and again with tiny changes to the starting conditions - such as changing a temperature input by a quarter of a degree. Little by little a consensus about likely outcomes is built.
The other approach is to use statistical relationships and correlations between what the oceans and atmosphere are now and and how they have behaved in other years. While it sounds as though the dynamic models - using all the correct physics - should be more accurate, Renwick says sometimes for seasonal outlooks, statistical forecasts do just as well. Why? Because of chaos theory and the inherent unpredictability of weather.
Which is why, I guess, that forecasters get it wrong sometimes?
"By far the majority of the time the MetService gets it right," says Public Weather Services manager Peter Kreft. "Certainly often enough to make a valuable contribution to the safety of life and property in New Zealand landmass."
To get the weather forecast right, forecasters first have to predict the future state of the atmosphere - where the highs and lows will be tomorrow or the next day, their shape and intensity, and how fast, and in what direction, they will move. "Weather modelling is quite good, but not perfect, at answering these sorts of questions," says Kreft.
But having the atmosphere sorted still doesn't guarantee that the forecast will be accurate. As Kreft points out, people don't really care about the future state of the atmosphere, they want to know what the weather is going to be like at their place.
Which is where it gets tricky - because there are so many variables including New Zealand's varied geography. In certain air streams, for example, a 5-10 degree change in direction may be enough to turn showers/cloud/etc on or off. Or a front may move a little faster than predicted, so that rain forecast for tomorrow morning may fall in the night hours before dawn. Or there may be a situation where all the ingredients for thunderstorms are present, but there is no "push" to get them started. Without a trigger nothing, not even a teeny bit of cloud, may develop.
What happens when severe weather warnings don't happen?MetService measures its success in severe weather warnings with two ratios - Probability of Detection (POD) and False Alarm Ratio (FAR). Such warnings are issued for events expected to occur over an area of 1000 square kilometres or more - which at about 32km by 32km square isn't that big.
The POD for heavy rain in 2008 was 95 per cent. Translation: of all the heavy rainfall events of 1000 square kilometres or more which occurred, 95 per cent of them were correctly forecast and the remaining 5 per cent were not forecast. The FAR for heavy rain in 2008 was 27 per cent. Translation: of all the heavy rainfall events of 1000 square kilometres or more which were forecast, 27 per cent of those forecasts were incorrect.
But as Kreft points out, such "cry wolf" measures are based on strict criteria. "The commonest cause of false alarms for heavy rainfall, for example, is insufficient quantity (maybe 75mm instead of 150mm), not no rain."
But severe weather warnings do sometimes cause problems. In October Far North Mayor Wayne Brown demanded an apology from MetService for a weather forecast he claims drove Labour Weekend visitors away and cost the region dearly. Kreft says the Thursday evening before Labour Weekend MetService put out a minimal severe weather warning for eastern parts of northland for Friday only. "It didn't rain as much as we expected and the whole
thing was over on the Friday - there was never at any time a warning that was issued that was applicable to Labour Weekend."
What are the oceans and atmosphere telling us right now?
"It's neutral," says Renwick. "The oceans aren't doing anything in the tropical Pacific. Sea temperatures are close to their long-term average." These are measured all over the world by satellites and other sensors including Argo, a global array of more than 3000 free-drifting profiling floats. "The ocean holds hugely more heat than the atmosphere and evolves more slowly - that's where most of the memory of the climate system resides."
But though the oceans are remembering to be normal, the atmosphere is not. The Southern Oscillation Index (SOI), calculated from the monthly or seasonal fluctuations in the air pressure difference between Tahiti and Darwin, is reading +1.
That means strong Pacific trade winds - which is what happens in La Nina conditions when there is extensive cooling of the central and eastern Pacific Ocean. The opposite, El Nino, is when there's extensive warming of the central and eastern Pacific and a negative SOI.
Both these weather patterns, which start in the tropics and ripple out to the higher latitudes, have a profound effect on what weather we get in New Zealand. But this year, we have neither La Nina nor El Nino. We have an ocean that isn't warming up or cooling down and an atmosphere above +1. At the moment, things are out of synch.
"The atmosphere is still thinking there is a La Nina going on - the trade winds are still stronger than normal," says Renwick. "It's quite unusual for there to be something happening in the atmosphere and not something happening in the ocean - we're not in a textbook situation."
This is because El Nino and La Nina cycles are what are known as coupled oscillations - the ocean and atmosphere talk to each other. "The reason the sea temperatures change is because the trade winds change, and the trade winds change because the sea temperatures change - it's a chicken and egg thing."
So what's it going to be - neutral or weak La Nina conditions?"Officially everybody is saying we're in a neutral state, because the ocean is neutral or near its long-term average," says Renwick. "But we are aware the atmosphere isn't quite playing the game. The indication is the situation in the atmosphere will ease back to neutral over the summer."
Because the situation is unusual Renwick says there is more reliance on dynamic rather than statistical models. "There is a chance the atmospheric circulation difference from normal will actually convince the oceans to play along and the ocean will actually start to cool in the Eastern Pacific." But only a handful of the models are suggesting the oceans might cool. Most say the wannabe La Nina atmosphere is going to fizzle out to neutral.
What sort of summer do we get with neutral conditions?
Bob McDavitt suggests we're in for a summer of jazz. It's a complicated musical analogy. When La Nina is to the fore in the orchestra of weather, it's all slower tunes and stringed instruments, apparently the sound of the hot dry anticyclones of last summer. When El Nino is in charge, expect strident marching sounds from the brass section - just like the noise that cold outbreak in mid-November made when it brought snow and hail to the country. But with neither La Nina nor El Nino you get, well..., "weather jazz" - with each instrument in succession playing a solo piece. In other words, a mixed bag. It sounds frustrating. "It's not - jazz is quite good. It's very satisfying - everyone gets their own due," says McDavitt. "If you take 10 days off for your holiday I can pretty well guarantee 50 per cent of it will be affected by easterly type conditions and 50 per cent by westerlies." With not as many hot, dry days, farmers should be happier, not to mention the rural fire authority.
Neutral conditions also provide more latitude for the path anticyclones track across New Zealand - a zone moulded by the subtropic ridge which divides the trade winds to the north from the Roaring 40s to the south. Watch out, says McDavitt, for any weather map that shows an anticyclone over the Chatham Islands and a low pressure system to the north - a recipe in northeastern districts for a period of vigorous easterly winds with driving rain. "It's the sort of thing we will issue heavy rain warnings for, but it's not something you need to cancel your holiday plans on, or even change your holiday plans, because it is only going to last 24-48 hours."
If short bursts of heavy rain are on the menu - the kind of rain that will cause temporary evacuation of some camping grounds - what else can weather users do to see it coming?
"Radar allows you to see things you cannot see any other way, and therefore allows you to forecast things you previously couldn't forecast," says Kreft. "The ideal would be to have all of New Zealand within about 180km of weather radar."
Radar is the secret weapon in the arsenal of the Meteorological Service. There are now five around the country - the newest at New Plymouth - and four more planned over the next few years. What radar sees is gathering precipitation on the edge of the horizon. And, from the echo it receives, whether it's rain, hail or snow, its intensity and how fast it's moving. "It equips us so much better to observe, and then forecast small-scale weather phenomena, particularly those that are severe," says Kreft.
The only problem with radar is that the warnings it can give are only a few hours in advance. At the moment the MetService provides a severe weather warning service for broad-scale weather that covers all of New Zealand. By the middle of next year it plans to have severe thunderstorm warnings - sent out to emergency managers by cellphone or pager - for those areas which are within radar coverage. Meanwhile, weather users can check Severe Thunderstorm Outlook and Watch online (www.metservice.com). And, if they're in range, for the latest sweep of the radar.
What about tropical cyclones - more or fewer in neutral conditions?
"Our records going back 40 years show the El Nino/La Nina cycle does affect whereabouts tropical cyclones form and how many you tend to have in a given season," says Renwick.
"To affect New Zealand, one has to come out of the tropics in the right place and go through an ex-tropical transition - so it has to become like a regular storm that would cross the country. It's quite a rare event and relatively unpredictable."
According to the statistics, in the long term average, in four out of five summers, one tropical cyclone comes out of the tropics and gets close enough to New Zealand to be an issue.
McDavitt notes that 12 years out of the past 40 years of records were neutral years. And comparing the three types of conditions - El Nino, La Nina and Neutral - the greatest likelihood of cyclone coming towards New Zealand is in the neutral bracket. "The probability of a cyclone moving out of the tropics and affecting New Zealand this coming summer is slightly above normal."
McDavitt makes the observation, too, that it was in previous neutral years that we got the big ones - cyclone Giselle which coincided with the Wahine disaster in April 1968; and cyclone Bola which struck Hawke's Bay and Gisborne/East Cape in March 1988.
Coincidence? Maybe.
But perhaps, with cyclone season starting in January, and a butterfly flapping its wings in say Fiji, our jazzy summer could strike up a very strident tune.
