by Ashley Hay
* * * * *
An increasingly sophisticated line-up of science, technology and talent at the BOM has been monitoring, analysing, predicting and publicising the changing climate and the accompanying dramatic weather events. Underpinning all of these functions – and lying at the heart of all meteorological work – are observations.
This applies particularly to forecasting. To predict future weather, you need to know what it’s doing now. The more detailed the observations, the more precise the forecast. ‘We observe the ground surface, the atmosphere, the state of the ocean and the atmosphere above the ocean,’ says Sue Barrell, the bureau’s assistant director of observations. ‘At ground level it’s things like the temperature on the ground, soil moisture and the characteristics below the ground. It’s the profile of temperature up through the atmosphere, right up to the stratosphere. It’s the particles in the ozone, the composition of the atmosphere, things like CO2 and ozone.’
Observations don’t stop at the stratosphere; they go on out into space, providing information about the Earth’s magnetic field, solar activity and solar radiation. To amass this data, the bureau deploys a formidable array of hardware. This includes high-tech Doppler radar as well as conventional radar; automatic and staffed weather stations; radiosonde balloons; marine buoys; and solar and terrestrial radiation monitoring observatories. There are also volunteer weather-watching ships, automatic weather sensors on commercial aircraft and robotic under-water gliders.
Above all, there are satellites. The US, Japan, Korea and China are some of the countries operating geostationary and polar-orbiting satellites, and the data are shared internationally. ‘Our bread-and-butter satellites in Australia are the Japanese geostationary satellites,’ says Barrell. ‘From them we get imagery hourly or half-hourly that covers the whole hemisphere.’
Observations of current weather give what meteorologists call the initial conditions. From the initial conditions, a computer-based model – mathematical equations that simulate the behaviour of the atmosphere – calculates future conditions. A humungous computer does the number crunching.
For the purpose of modelling, the globe’s surface and the atmosphere above it are divided into a three-dimensional grid or mesh. The numerical model checks all the weather elements at each point on this grid – some 20 million of them. From the initial conditions, it creates a forecast for a short time ahead. This then becomes the new initial state for another short forecast. The process is repeated over and over to the limit of the required forecast.
‘You move in what are called time steps,’ says Kamal Puri, leader of the bureau’s Earth system modelling research program. ‘Typically, depending on how fine your mesh is, these time steps can be about 10 minutes long. So you keep moving forward 10 minutes, then 10 minutes again, and 10 more minutes, until you get to 10 days ahead.’
This process would give perfect predictions but for one hitch: the initial conditions can never be exact; they always contain uncertainties and flaws. With each time step, these become magnified.
‘The errors in the initial conditions can grow very rapidly, and after about 10 days or two weeks you lose most of the predictability,’ Puri says. ‘So the shorter the range of the forecast, the more accurate it is, but this short range is now moving further and further out. In the old days it was just one day. Now you’re pushing to four or five days and beyond because of all the improvements.’
* * * * *
‘Three big things have happened in the past 20 years that have affected what we are capable of doing now and how good we are,’ says Rob Vertessy, director of meteorology. ‘First is the proliferation of data, chiefly from satellites. The satellite era has flourished and we get a phenomenal amount of data at very high resolution now. It has been a fundamental game changer. It’s been aided by the internet: big, fat pipes of data coming in from all over the world.’ Thirty-three geostationary and polar-orbiting satellites – 23 more than in 1993 – feed a stream of images to a dozen ground stations around Australia and its Antarctic bases. Not only are there more satellites today, they’re also better. But what has really changed profoundly is the bureau’s ability to exploit the data they provide.
Looking at the bureau’s website, you might think that all satellites provide are pictures of land, sea and clouds. But there’s more to it than that. From satellite images, meteorologists now extract information on solar radiation, sea surface temperature, vegetation, volcanic ash, temperature, water vapour and wind.
‘A lot of science is required to extract that material,’ says Sue Barrell. ‘We put a lot of effort into developing the algorithms and enhancing the products that come from the satellite data.’
Radar is also contributing to the data stream. Since 1992 the bureau has boosted its radar numbers from 38 to 66. A dozen are Doppler machines that not only show precipitation intensity, but also wind flow at different heights. This allows forecasters to see the structure of individual thunderstorms and gauge their ferocity. The second big leap has been in modelling and computing. The bigger data flow could not be processed without better models and more computing muscle. Once these were in place, forecasting accuracy shot up.
‘Twenty years ago we were doing forecasts probably out to three or four days, if we were lucky,’ says Alasdair Hainsworth, assistant director of weather services. ‘Now we’re doing it out to seven days. The statistics show that the seven-day forecasts are about as accurate as the four-day forecasts were 20 years ago.’
With greater skill has come fine detail or, to use the technical term, greater resolution. This means being able to tailor forecasts for ever-smaller areas. Kamal Puri explains it in terms of the mesh size of the global grid.
‘Ten years ago we would have had a mesh size typically of 100–200 kilometres,’ he says. ‘Now we are getting close to a global model running at about 10–15 kilometres. That is a dramatic improvement.’
The third big leap has been in understanding the physics of the Earth’s climate system. Twenty years of climate research has found its way into forecasting models, which now more faithfully reflect the system’s complexity. This has refined not only short-term forecasts but also longer-term services such as the bureau’s three-month seasonal outlook.
For its short-term (seven-day) forecasts, the bureau uses a series of models collectively called the Australian Community Climate and Earth-System Simulator (ACCESS), which Puri helped develop. ACCESS is what’s known as a coupled model, since it takes into consideration both atmospheric and ocean conditions. After it began operating in 2010, the error rate in short-term forecasts dropped sharply.
For long-term seasonal forecasts, the bureau uses another coupled model, the Predictive Ocean Atmosphere Model for Australia (POAMA). The first version of this model went live in 2002, producing routine forecasts of El Niño conditions, and is now being used routinely as part of the bureau’s service.
The BOM’s burgeoning skill set has sparked an explosion of services, many unthinkable 20 years ago. The statistics are mind-boggling. Every year the bureau delivers more than 330 000 weather forecasts, 36 000 tidal predictions and 350 000 aviation forecasts and warnings.
Its website, which went live in 1996, is its biggest outlet for this information. As well as myriad services for general users – forecasts, warnings, charts, satellite and radar images – it offers specialist programs for farmers, the defence forces, aviation, shipping, mariners, miners, water managers and commercial meteorological agencies. You can get information (with warnings, where relevant) on climate, tides, wave height, tsunamis, river flows, cyclones, bushfires, coral bleaching, space weather and volcanic ash. Some of this material comes packaged in two relatively new programs, one providing water data (on rivers, storages and aquifers) and the other environmental data.
The site gets more than 33 billion hits a year, with numbers spiking during tropical cyclones and floods. It’s Australia’s top site for weather information. And it continues to evolve. Currently the
bureau is rolling out NexGen, its revamped modelling system.
‘NexGen encompasses a new forecasting system, new visualisations, a complete makeover of the forecasting services on our website to make it more user-friendly, adding the ability to point and click anywhere on the map and generate a forecast,’ says project manager Howard Jacobs, an exuberant tech whiz with an infectious zeal for what he does. Through its Forecast Explorer graphical viewer tool, NextGen will eventually offer seven-day forecasts for some 650 locations nationwide. And it will zoom in on 6 × 6 kilometre areas, giving severe-weather warnings where necessary.
It doesn’t stop there: just over the horizon are yet more amazing gizmos. Jacobs worked on one of them, named MetEye, in 2012–13, and an ‘experimental’ version of its service has now gone live. He calls it Google Maps for weather. It will replace Forecast Explorer and allow users to customise their weather maps, combining features such as forecasts, radar images, cloud, temperature, rainfall figures, wind, waves and tides for their chosen location.
* * * * *
All the skills the BOM can now mobilise to peer into the distant future are revealing an unsettling scenario, one where climbing temperatures continue to cook our weather and climate into an increasingly volatile stew.
‘Climate change is driving changes in sea levels, ocean chemistry, marine, terrestrial and aquatic ecosystems, agricultural production, water security and the frequency and magnitude of severe weather events,’ the bureau told the 2013 Senate Inquiry. ‘Further change is now locked in for centuries, whether or not emissions are reduced or even halted in the near term.’
Neil Plummer says future weather may look much like it does today, but the frequencies and intensity of some events will be different. ‘There will be a lot more warmer weather … fewer occurrences of cooler weather, and the chances of warm-event records being broken will increase as time goes on,’ he says. ‘The models also suggest we should get increased frequency of intense rainfall events in many areas.’
So the long-term projection for Australia is more and longer heatwaves and droughts alternating with periods of heavier rain. This is likely to mean more bushfire weather and more flooding. Overall, though, wet years will become fewer and dry years more frequent, particularly in southern Australia. As for individual weather phenomena, scientists say only that there may be fewer tropical cyclones, though they are likely to be more intense.
Recent weather events have convinced many Australians of the reality of global warming. One who has needed no convincing is Mike O’Neill. Like many in the storm-chasing fraternity, he’s sure weather events will become more extreme, and with that will come more potent lightning. ‘I doubt I will see these events in my lifetime, but if I do I will certainly be ready for them with camera in hand!’
We’re 70 kilometres south of Darwin and it’s 2.15 p.m. The sky looks less blue now. Puffy cumulus clouds are everywhere, some boiling up promisingly into cumulonimbus storm clouds. There’s a big one ahead.
Half an hour later, on a hill just outside the town of Adelaide River, we get a grandstand view of the storm. It’s building up at astonishing speed, 10 kilometres to the east of us. The radar image shows it blotched with orange and yellow, indicating moderate to heavy rain. Mike O’Neill and photographer Dave Hancock, who has been following us in his LandCruiser, set up tripods and cameras fitted with lightning triggers – devices that set off the shutters when they detect flashes.
At one point a vivid bolt zigzags out of the side of the storm and hits the ground well clear of it. The cameras go ape. ‘Wow, that was huge!’ exclaims O’Neill. ‘It would have travelled at least 10 kilometres outside the storm. So we’re within range.’ Having had several lightning encounters that were too close for comfort, he has nothing but respect for thunderstorms. Because of the danger, the bureau in no way encourages chasing, but it welcomes information that chasers provide.
By 5 p.m. a curtain of rain has dropped beneath the storm. The downdraft of cool air that it brings is spreading out and spawning new storm cells. Sure enough, the radar image has several more big cells lining up to the south-east. We hit the road again.
Come 6.30 p.m. we’re parked beside the highway, 30 kilometres to the south. There are storms all around us. Lightning is stabbing the ground at all compass points and the tripod-mounted cameras are clicking away. We’re right in the middle of a stupendous lightning show. O’Neill’s mental forecast was spot on.
Later, as we head back to Darwin in the car, he says: ‘That was a whole month’s lightning in one night. I give it 20 out of 10.’
Firefront
Antarctic ice: Going, going …
Firefront
Ian Gibbins
The proposition: a firefront, climbing the hillface, approaching lines of grey box,
an edge, a vibration, ragged, the juxtaposition of above and what lies below.
You must decide upon a frame of reference, a coordinate system, within which
local events, diary entries, arrivals and departures can be securely placed.
Option one: (as usual) the sky. Some common descriptors: oppressive, leaden,
foreboding. Alternatively, overcast, cloud-streaked, ambivalent. And yet,
notwithstanding prior predictions, there is absolutely nothing to see: (as usual)
the air, through all its troughs and ridges, typical for the season, remains clear.
Option two: the earth. Once again, far too familiar. You already know what
it means: bedrock solid, unable to move without the application of heavy
machinery, set fast, interlocked to tectonic plates, a foundation stone, like
a mother’s mother, off-white, like salt, or milk, or thoroughly unexpected snow.
Option three: an ocean. How does it go? Roiling? Tumultuous? Surging with swell
and storm and eddy? Fathomless? Uncharted? The boundary we cannot extend?
A source of endless lies, stories that intrigue, inveigle, insist on continued disbelief.
Shallows tempting? Rising to cover your curling toes, your reefscarred shins.
Option four: the fire itself. This you also know. The things that can burn: lava flows,
molten glass, cast iron, magnesium. Your throat, raw as it is. A blue-lined notebook,
school-yard friendship, fingertips, letters dreamt at midnight, music ringing from
plaster walls, a road you barely recognise. Objects singed and ashen and burst apart.
A final reminder. To make a list. The items we must not forget. Ingredients we do not
grow here: cinnamon, clove, cardamon, Indian tea, black currant, berries, blueberries.
Materials we must find time to mine: cobalt, nickel, molybdenum, opal, fully
oxidised zinc, diamond, tourmaline, malachite, crystalline quartz, pure and simple.
The direction of the wind. A return address. The passwords we require. The encryption
keys that preserve our integrity, hold our neighbours to account, plot a pathway out.
To repeat: the direction of the wind. Disentangle arms from safety blankets, scarlet
across our backs. What else? Count the numbers that name exploding supernovae.
Reached by committee, nineteen eighty-three
Liner notes, Voyager Golden Record
Antarctic ice: Going, going …
Nerilie Abram
It was January 2008 and I was on the back deck of HMS Endurance, wearing a full-body survival suit and eager for the short helicopter ride that would take me onto Antarctic ice for the first time. The ship was travelling through the channel that divides James Ross Island from the Antarctic Peninsula – a trip that would have been impossible not so long ago.
Since the 1990s, a series of ice shelves along the Antarctic Peninsula have collapsed, including the ice shelf that had once permanently connected James Ross Island to the rest of the continent. Most famously, the collapse of the nearby Larsen B ice shelf had been captured by satellite photographs.
These images have been held up as an example of climate change happening before our eyes. But are they? This was what I was here to find out.
* * * * *
The Antarctic Peninsula is warming quickly. Over the last 50 years the climate here has warmed three times faster than the global average. The problem is that temperature measurements in this remote region don’t go much further back than that. So how can we put the current warming into perspective? The answer lies locked within Antarctica’s ice. The ice blanketing most of the Antarctic continent is made of snow that has fallen and been buried. Scientists use these ancient ice layers as a window into Earth’s past climate.
The deepest parts of Antarctica’s great ice sheets might hold a climate record that goes back more than a million years. In the 2013–14 summer, scientists from the Australian Antarctic Division led an ice-drilling expedition to Aurora Basin, high on the East Antarctic plateau. This was part of a coordinated international effort towards the most ambitious and technically challenging piece of ice-core research ever attempted: the quest for Antarctica’s ‘oldest ice’.
For the much smaller – and earlier – James Ross Island ice-drilling project, our team of seven scientists and engineers lived and worked in tents on the ice for almost two months. The top 283 metres of this ice cap consist of snow that’s built up over the past thousand years. We know the age of the snow layers by counting the yearly summer-winter cycles of chemical impurities, such as sea salt, in the ice, and by the fixed time markers left in the snow by ash from volcanic eruptions.
