Dave Allen

The final meltdown

Retreating glaciers and thinning snow and ice are the future of New Zealand’s mountains. Climate change is predicted to warm the country’s atmosphere by 1–4°C by the end of the century, altering the natural water cycle—how much is frozen as snow, how much falls as rain, and how much flows in rivers. Climate researchers are seeking to predict what will change, and when. What will be the impact on hydroelectric power stations and irrigation schemes? Which areas will be hit hardest by flooding, or increasingly severe drought? The Deep South National Science Challenge is taking a lead role in helping decision-makers plan for the coming century.

Written by       Photographed by Dave Allen and Rob Suisted

The snowline survey is now conducted by NIWA, and its data feeds a Deep South project to predict the future of every glacier in the country. The research is also looking at the future availability of spring and summer meltwater from snow.

It’s early March, and the first crisp hints of autumn are in the Otago air. A red-and-white Cessna takes off from Queenstown Airport and flies south. The plane belongs to Milford Sound Scenic Flights, but this is no ordinary scenic flight. On board are climate scientists, glaciologists and a PhD student who is an expert in creating three-dimensional terrain models from oblique aerial photography.

As the aircraft travels along the spine of the land, it bucks and twists in the mountain updrafts. The passengers barely notice. They are glued to the windows, studying and recording the condition of the snowpack and glaciers of the Southern Alps.

Climate scientist Andrew Lorrey cautiously slides open a door, admitting a blast of cold air into the cabin, and points a high-resolution thermal imaging camera at Tasman Glacier below. From a mile high, he can measure the temperature of a few square metres of ice, water or rock.

This is the annual end-of-summer snowline survey. Over two days, the research team, led by NIWA and including scientists from universities around the country, observe glaciers from Fiordland to Arthur’s Pass to see how the ice has fared over the preceding year. The survey has been conducted since 1977, and has produced an enviably long—and globally significant—dataset on glacier behaviour in a time of climate change.

Another of the climate scientists on the flight, Andrew Mackintosh, is leading a project to predict the future of these glaciers—research which has brought together the country’s leading snow and glacier scientists for the first time, under the banner of the Deep South National Science Challenge.

The snowline survey was started by New Zealand’s doyen of glaciologists, Trevor Chinn, who has personally eyeballed all of this country’s 3153 glaciers. Chinn was responsible for the first glacier inventory in New Zealand, which took him 10 years to complete. He subsequently chose the 50 index glaciers that are tracked annually by the snowline survey. He has lived through times of glacier advance, and now he watches their retreat.

This year has been a bad one, he tells me at his home in Lake Hāwea. A temperature anomaly in the Tasman Sea in late 2017 and into 2018 resulted in a marine heatwave, with sea-surface temperatures averaging two degrees warmer than normal, and on the West Coast spiking up to six degrees warmer than average. The effect on glacial melt over the summer was profound.

“This season is the worst we’ve ever seen,” says Chinn. “Almost half the glaciers have lost all the snow they gained during the previous winter, plus some from the winter before that. There are rocks sticking out everywhere. The meltback is phenomenal.”

When Chinn started the glacier inventory in 1977, the purpose was to update the country’s topographical maps with accurate positions of glaciers and provide estimates of the area and volume of permanent snow and ice. But after a few years of flying and photographing retreating glaciers, he realised something more important was at stake. He wasn’t just mapping, he was monitoring how the mountains were changing, and his photographs were a record of a changing climate.

“Glaciers don’t lie,” says glaciologist Trevor Chinn, who began surveying them from the air in 1977.

There is no more reliable indicator of climate change than a glacier, says Chinn. The annual change of ice volume, known as the glacier’s mass balance, is the sum of all climate processes. “You can’t argue with a retreating glacier.”

Not all years are as bleak for glaciers as 2018, but the bad ones tend to stick in scientists’ memories. Lorrey remembers his first negative mass-balance year, 2011. There had been a strong La Niña, with strong highs over the South Island, and more frequent northerly and easterly winds. The glacier tongues and their associated snowfields were grimy brown. Ablation had been severe over the summer, and the ice looked sick.

Will this be the new visual appearance of alpine New Zealand, I ask him: besmirched snow, grey rock where ice used to be, faded glory?

“With climate change, the dice are definitely loaded,” he says. “In the past, natural multi-decadal patterns of climate  variability sometimes aligned with the warming trend, and sometimes went against it. Now it seems as if the variability is mostly going in the same direction. It’s like putting a concrete block on the accelerator of climate change.”

Lorrey says the 2018 survey, with its debris-covered snow, glacier lakes growing in size, and ice retreating upslope, feels like a foretaste of what is to come. But he adds that New Zealand’s climate is variable enough that there will probably be years when snow accumulates and the mass balance of the glaciers tweak into the positive side of the ledger. At those times, the ice will wear a fresh white face.

[Chapter Break]

Trevor Chinn has named only one glacier in New Zealand—compared to nearly 50 in Antarctica—but it remains his favourite. Ivory Glacier, its name a riff on a nearby mountain, The Tusk, provided Chinn with the data he needed to calculate a glacier’s ice volume by measuring its area.

Chinn worked on the Ivory from 1968 until 1975. In those days, he says, the glacier had a marvellous ice-filled cirque and a tiny patch of water at the tip of its tongue. Today, that water is a lake, with blue ducks living on it, and the glacier is a tiny patch, high up on a bare mountainside.

This could be the future for all New Zealand glaciers, but for some it will happen a lot faster than others. Large valley glaciers are covered with a thick layer of rocky surface moraine that acts as a thermal blanket, insulating the underlying ice. For these, retreat will be decades slower than for steep, debris-free glaciers such as the West Coast’s Fox and Franz Josef.

Predicting the fate of such glaciers is by no means straightforward, says Mackintosh, a professor of earth sciences at Victoria University, director of its Antarctic Research Centre, and a lead author of a forthcoming Intergovernmental Panel on Climate Change (IPCC) special report on the links between melting ice and rising oceans. In order to model the behaviour of a debris-covered glacier, you need to know the exact distribution of debris, how thick it is and what its thermal properties are—most of which is unknown, and difficult to measure.

Despite such data gaps, Mackintosh and fellow Victoria glaciologist Brian Anderson have developed some trajectories for New Zealand’s glaciers over the coming century. He shows me two visualisations: one based on the IPCC’s most dire climate pathway, in which greenhouse-gas emissions continue to rise throughout this century, and the second based on the IPCC’s most optimistic alternative, in which emissions peak by 2020 and decline substantially afterwards.

Under the first scenario—“Burn it all and damn the consequences”, as he puts it—most of New Zealand’s ice has disappeared by 2100.

“The tongues of Fox and Franz Josef have gone completely. There’s just a little bit of healthy ice left on top of Aoraki, plus a few debris-covered glaciers like Tasman, where the upper part that used to provide the snow and ice has gone, the lower part has been eaten away by the lake, and you’re left with a slug of debris-covered ice in the middle. That’s one possible future.

“But if we follow the declining-emissions pathway—drawing carbon out of the atmosphere, reducing the carbon budget to zero, limiting warming to two degrees above pre-industrial temperatures—then the glaciers get about 30 per cent smaller, but we end up, at the end of the century, with something not dissimilar to what we have today. Franz and Fox, though greatly diminished, still exist. They’re still recognisable features of the mountain landscape, still icons of the Southern Alps. It’s a fundamental difference: keep the glaciers or lose them. This is what’s at stake.”

[Chapter Break]

Predictions for New Zealand under a warming climate are that the extremes of water supply—either absence or abundance—will become more frequent and more acute. The dry gets drier, the wet gets wetter, and a rising ocean threatens every coastline.

How the country might best confront such challenges is at the heart of the Deep South National Science Challenge, which aims to provide guidance on a range of issues, from flood protection to wastewater infrastructure, drought management to sea-level rise.

Much of the research has to do with modelling the coming climate impacts and accurately assessing their risk. What bedevils this work is the problem of uncertainty. When the earth system was stable, risk could be calculated with relative accuracy, based on the historical frequency of damaging events and the extent of their impact. That stable world has gone.

We have entered the age of uncertainty, in which the past has become an unreliable basis for forecasting the future.

Uncertainty, moreover, is a self-replicating phenomenon. Uncertainty in one parameter spills over into another, creating an uncertainty cascade.

Suppose we want to know what future our society (our children, grandchildren, unborn generations) will face. That future depends on what happens with greenhouse-gas emissions, and the various impacts of those emissions can be predicted by a climate model, which needs to be fine-grained enough to provide forecasts for individual locations. But at every level of this sequence there are uncertainties—of timing, magnitude, location.

The trick, as economist Anita Wreford explains to me in her office at Lincoln University, is not to become paralysed by uncertainty, but to find ways to make robust decisions despite uncertainty.

Local climate variability led to lower air temperatures and caused 58 glaciers to advance between 1983 and 2008. (Overall, New Zealand glaciers lost mass during this period.)

The specific decision-making arena Wreford and her colleagues are modelling is water storage for agriculture in Canterbury. As the climate warms, Canterbury is projected to become drier, and the availability of water becomes less certain. Not only may there be less precipitation, but the volume of snow-melt in the Southern Alps will decline, and many of Canterbury’s eastward-flowing rivers will dwindle and even dry up. Flow in other rivers may increase due to more rainfall in their catchments. Competition for scarce water will intensify in some areas, but it isn’t clear where those areas will be.

Storage is a way for farmers to adapt to a changing waterscape and safeguard a reliable supply. But water storage is a costly investment, and its benefits are uncertain. Suppose inflows are insufficient to keep the reservoir replenished (as has happened at times at Opuha dam, in the Mackenzie country). Suppose other factors (shifting markets, new diseases, loss of social licence) cause farming to become less viable. Over-investment could become a crippling financial burden on those in the storage scheme. On the other hand, under-investment could result in lost production. A robust decision needs to take into account the full gamut of possibilities.

The model that Wreford’s team is working on simulates 24 different hydrological scenarios over the course of this century. These are used to estimate the amount of storage required to meet agricultural needs for each of those potential futures, along with the associated costs and benefits. The model can’t produce certainty of outcome—that state no longer exists—but offers a more inclusive way of assessing the options and preserving flexibility in the face of change.

[Chapter Break]

Models and predictions are only as robust as the observations they draw on—but when it comes to alpine climate, the data are less comprehensive than one might expect.

NIWA, the country’s chief collector of long-term environmental data, has a network of 600 weather stations which feed the National Climate Database. But only 12 of them routinely monitor snow accumulation,  and only three are able to provide separate data on precipitation that falls as snow and that which falls as rain.

Future variability in New Zealand’s climate is a special problem, because the climate is highly variable to begin with. As a result, separating the statistical signal of a changing climate from the background noise of natural variability can be difficult.

Since 1977, the country’s 3000-plus glaciers have lost 30 per cent of their ice volume due to rising air temperatures. Tasman Glacier’s terminal lake formed in 1973, and the glacier has been retreating since, the rate governed by the calving of icebergs as the lake eats away at the glacier snout.

NIWA hydrologist Christian Zammit, who leads a Deep South project researching climate impacts on the national water cycle, says in New Zealand the temperature signal is clear—temperature is rising—but the signal regarding changes in precipitation isn’t yet clear.

“We know the signal is there,” he says, “but it’s not strong enough that you can identify it.”

At the moment, he says, depending on the characteristics being looked at, the impact of climate change on hydrological process will start to be statistically different from current natural variability between 2040 and 2060. The conundrum is that if we wait to act until the signal is unequivocal, it will be too late to do anything about it: “You’ve run out of time to mitigate or adapt,” he says.

Zammit’s project aims to create hydrological simulations for 66,000 catchments, exploring the impacts of climate change on agriculture, hydroelectric generation and flooding. One of the goals is to learn how climate components such as precipitation interact with particular landscapes—for instance, how snow falls in a particular catchment.

Since 2017, the University of Otago has operated a weather station in the Pisa Range, and students from the New Zealand School of Surveying have been using a sophisticated drone to photographically map snow accumulation in the catchment of the Leopold River over several winters.

I visit the site with climatologist Jono Conway, who works for atmospheric research company Bodeker Scientific in Alexandra. We set out from the Snow Farm near Cardrona, where cross-country skiers wearing all the colours of the Lycra rainbow are limbering up for an international championship race. Conway and I strap on somewhat less sportif snowshoes and trudge downhill in the direction of the weather station. We rock-hop across the upper reaches of the Roaring Meg—the Murmuring Meg at this time of year—and make our way along a west-facing slope to the station.

Weather stations aren’t much to look at: a wind-speed vane on top of a vertical pole, a perpendicular arm with small upward- and downward-pointing instruments to measure incoming solar radiation and the reflectivity of the snow, a sonic range-finding device for measuring snow height, a data logger.

But Conway says this sort of instrumentation is revolutionising knowledge about the seasonal snowpack. Data from stations such as this help refine and validate a national snow model that can predict how much precipitation falls as snow, where it falls, when it melts, and where the meltwater ends up. Then, as part of Conway’s Deep South project, this model will be incorporated into NIWA’s existing national hydrological model to provide a better picture of how snow and ice feed river flows and become stored in lakes and groundwater. Later, the model may be able to predict water threats, especially extreme ‘rain-on-snow’ events which can result in catastrophic floods, and forecast changes in water availability, such as rivers running dry.

In the short term, Conway tells me, an increase in glacier melt will keep hydro lakes replenished during dry summers, and warmer winters will provide meltwater at the time when electricity demand is highest. But at some point in the future, as glaciers lose mass, the meltwater contribution to rivers will start to run out. The term hydrologists use for the point of maximum runoff is ‘peak water’—analogous to the fossil-fuel industry’s ‘peak oil’.

Climate scientists are unsure when that moment will arrive. In the European Alps it already has, says Mackintosh: “The glaciers there are already too small. Peak water still lies ahead for the Himalayas, and for New Zealand.”

If models are only as good as the data they crunch, it is also true that they are only as useful as society’s willingness to accept and apply them. And just as accelerated climate change has destabilised the earth’s physical systems, it is leading to volatility in political systems as well.

Rapid change is now the defining feature of both nature and society. Many scientists think you can’t analyse one without the other. A new research discipline called sociohydrology has emerged for just this reason. The future of water and the future of humanity are inseparable.

Deep South researcher Belinda Storey works at this interface. She has degrees in political science, disaster risk and finance, and serves as a director on the boards of Landcorp Farming and 350 Aotearoa.

For low-rainfall agricultural areas such as the Canterbury Plains, water storage (such as the Rangitata South Irrigation Scheme) is seen as a crucial adaptation to ensure reliability of supply. But with uncertainty over future precipitation and river flows, such investments carry a high financial risk. Deep South researchers are creating improved analytical models for making robust decisions in a changing climate.

One of her areas of study is the psychology of climate risk, especially the notion of ‘future proofing’—the attempt to neutralise the negative impacts of future events. She believes the very phrase implies a level of certainty about the future that doesn’t exist.

“When people talk about ‘future proofing’ they often haven’t taken account of how much things are going to change,” she says. “For example, many water consents for irrigation are for 35 years, but we don’t have modelling that will tell us what the water supply will be in even five years’ time. If there’s one thing people need to understand it’s this: the future will look nothing like the past.”

As the climate warms, weather impacts are predicted to come from both ends of the pluvial spectrum: droughts and floods. Both require expensive infrastructure—storage reservoirs and flood-protection schemes—to mitigate their unwanted effects. And their frequency and intensity upend traditional ways of thinking about extreme events.

The changing distribution of extreme events is one of Storey’s areas of research. “It’s really hard to get your head around,” she admits. “I was at a conference recently and someone said there had been four one-in-100-year floods in the previous year. What’s actually happening is the distribution of extreme rain events is shifting, and those aren’t one-in-100-year events any more.”

She draws me a graph, a classic bell curve with some intervals along the tails of the curve—standard deviations—to indicate extremes in the distribution.

“The thing to realise is that change is not linear. For instance, if a rare event used to have an occurrence of one in 740 years—average plus three standard variations—a reduction in one standard deviation changes the probability to one in 44 years.

“Suppose a council has to repair a stopbank, or build some other flood defence. They may decide that the flooding risk is changing by, say, five per cent, and therefore they ought to be conservative and add a 10 per cent safety margin. In fact, they should be increasing the margin several hundredfold.”

Another problem with artificial defences, whether embankments to protect homes or storage reservoirs to provide irrigation, is psychological: they engender a false sense of security, a phenomenon known as the ‘levee effect’.

“If you build a stopbank, people feel safe and invest in assets such as homes and farms behind it,” says Storey. “If you build a storage reservoir, farmers invest in certain types of agriculture that rely on irrigation. Both groups of people increase their exposure without an accurate assessment of risk. When failure happens, the damage is compounded.”

Underlying such psychological blind spots, believes Storey, is the Western presumption of dominance over nature and expectation of control. As a way of challenging this thinking, she started exploring the idea of a ‘climate lease’ for property—a way of shifting the mindset from permanent ownership to temporary rights. She initially applied the concept to coastal property, but says it could apply to any situation where there are escalating natural hazards.

“Most people already get the idea that the coastline has an intrinsic time limit on it now,” she says.

“Those who own coastal property are deluding themselves if they think they have permanent rights to that land. In reality, they’re leaseholders. The owner of the lease isn’t the government or an organisation or a person, it’s nature. And you can’t bargain with nature. You can’t ask for an extension.”

Storey argues that the change from freehold to leasehold thinking needs to happen, not just with property owners, but with councils and utilities, to reflect the fact that property has a time limit built into it.

Such a profound shift will not happen easily, but Storey believes insurance is a likely lever. According to the Insurance Council of New Zealand, 2017 was the most expensive year for weather-related claims since the Wahine storm of 1968. Payouts were in the order of $240 million. This year’s claims are expected to surpass that figure.


Insurance is coupled to actual risk, not an out-of-date model based on knowledge of the past. In addition, insurance contracts are for a year at a time, so insurers can respond quickly to change, with significant financial implications for their customers.

“You might have a mortgage for 30 years, but your insurance policy is annual,” explains Storey. “An insurer can respond to changing risk almost overnight, either by hiking premiums or even choosing to exit an insurance market. If an insurer decides to withdraw, and your mortgage is conditional on insurance, you go into technical default. At that point the people who bear the consequences are property owners and banks, but the first link in the chain are insurers. They’re the ones with the greatest sensitivity to changing risk.”

A climate lease normalises the reality of increased risk, says Storey, conditioning the property owner to the expectation of diminishing value over time.

“People delude themselves into thinking, ‘This won’t happen in my lifetime’, or, ‘I’ll sell before there’s a problem’,” she says. “If you accept a time limit built into a lease, you can’t fudge the issue or deceive yourself. You see the value declining as the sea is rising.”

[Chapter Break]

Still, I wonder if the economic rationality of tools such as climate leases is a sufficient force to overcome public and political inertia on climate change. Is something else needed: a sense of stewardship and aroha for the landscapes that have shaped us?

I felt that tug of connection when I visited Haupapa/Tasman Glacier, New Zealand’s longest glacier. I hadn’t taken the road to Mount Cook Village for years, but on a glittering winter afternoon I did, driving beside Lake Pukaki, my eyes darting constantly, hungrily to that pool of inestimable colour, a sky-blue topaz among sky-piercing peaks.

The snow on the northern faces of the looming mountains had the sheen of licked ice cream, and strong winds at altitude were making the summits smoke. I walked the track to the lookout over Tasman Lake and scanned the drifts of glacier-scoured gravel along the lake sides, the rubble-covered snout, the perfect reflection of Novara Peak in the cement-coloured water, the clutch of icebergs melting in curious symmetries like turned chess pieces.

This is part of me, I thought. I belong.

New Zealand’s hydroelectric generation, much of it powered by meltwater from glaciers, may benefit in the short term from a warming climate. But once ‘peak water’ is reached, some time in mid-century, flows will decline as glaciers melt away, and hydro stations will struggle.

Can something as enigmatic as emotional attachment to landscapes mobilise people to work for their preservation, when that involves as vast a challenge as climate change?

It happened with water quality. A cherished mythology of pure rivers and pristine lakes crashed into the reality of choking sediment, invasive algae, toxic pollutants and native aquatic life headed for extinction. Collective dismay prompted restorative action.

Can it happen with water quantity? Can emotions such as these achieve what rationality hasn’t? Love of place. Sense of guardianship. Concern for future generations. Desire to preserve a country’s heritage—not least its rivers of ice.

In the Southern Alps, the scientists on the snowline survey ponder similar questions.

“When we see these changes happening in the environments we go into, we’re asking not just, ‘What does this mean for the world, as part of the global climate shift?’ but ‘What does it mean for people as well?’” says Andrew Lorrey. “Most of us who go on that flight have children, and our kids are young, and we’re wondering what their world is going to be like. I think of Trevor Chinn, handing off the baton to my generation of scientists. What will the glaciers be like when we’re his age? Will we be saying, ‘Do you remember when there was ice around this place?’ Or will it be a different story with a different ending, one where we say, ‘We knew we had a problem, we did something about it, and look at the glaciers coming back’?

“That would be a hope—a slim hope, perhaps, but a hope that could lead us to act.”

More by

More by Dave Allen and Rob Suisted