̽��ֱ�� of Cambridge - modelling

Giant underwater waves affect the ocean’s ability to store carbon

sc604 — Thu, 16 Mar 2023 15:00:53 +0000

An international team of researchers, led by the ̽��ֱ�� of Cambridge, the ̽��ֱ�� of Oxford, and the ̽��ֱ�� of California San Diego, quantified the effect of these waves and other forms of underwater turbulence in the Atlantic Ocean and found that their importance is not being accurately reflected in the climate models that inform government policy.

Most of the heat and carbon emitted by human activity is absorbed by the ocean, but how much it can absorb is dependent on turbulence in the ocean’s interior, as heat and carbon are either pushed deep into the ocean or pulled toward the surface.

While these underwater waves are already well-known, their importance in heat and carbon transport is not fully understood.

̽��ֱ��results, reported in the journal AGU Advances, show that turbulence in the interior of oceans is more important for the transport of carbon and heat on a global scale than had been previously imagined.

Ocean circulation carries warm waters from the tropics to the North Atlantic, where they cool, sink, and return southwards in the deep ocean, like a giant conveyer belt. ̽��ֱ��Atlantic branch of this circulation pattern, called the Atlantic Meridional Overturning Circulation (AMOC), plays a key role in regulating global heat and carbon budgets. Ocean circulation redistributes heat to the polar regions, where it melts ice, and carbon to the deep ocean, where it can be stored for thousands of years.

“If you were to take a picture of the ocean interior, you would see a lot of complex dynamics at work,” said first author Dr Laura Cimoli from Cambridge’s Department of Applied Mathematics and Theoretical Physics. “Beneath the surface of the water, there are jets, currents, and waves – in the deep ocean, these waves can be up to 500 metres high, but they break just like a wave on a beach.”

“ ̽��ֱ��Atlantic Ocean is special in how it affects the global climate,” said co-author Dr Ali Mashayek from Cambridge’s Department of Earth Sciences. “It has a strong pole-to-pole circulation from its upper reaches to the deep ocean. ̽��ֱ��water also moves faster at the surface than it does in the deep ocean.”

Over the past several decades, researchers have been investigating whether the AMOC may be a factor in why the Arctic has lost so much ice cover, while some Antarctic ice sheets are growing. One possible explanation for this phenomenon is that heat absorbed by the ocean in the North Atlantic takes several hundred years to reach the Antarctic.

Now, using a combination of remote sensing, ship-based measurements and data from autonomous floats, the Cambridge-led researchers have found that heat from the North Atlantic can reach the Antarctic much faster than previously thought. In addition, turbulence within the ocean – in particular large underwater waves – plays an important role in the climate.

Like a giant cake, the ocean is made up of different layers, with colder, denser water at the bottom, and warmer, lighter water at the top. Most heat and carbon transport within the ocean happens within a particular layer, but heat and carbon can also move between density layers, bringing deep waters back to the surface.

̽��ֱ��researchers found that the movement of heat and carbon between layers is facilitated by small-scale turbulence, a phenomenon not fully represented in climate models.

Estimates of mixing from different observational platforms showed evidence of small-scale turbulence in the upper branch of circulation, in agreement with theoretical predictions of oceanic internal waves. ̽��ֱ��different estimates showed that turbulence mostly affects the class of density layers associated with the core of the deep waters moving southward from the North Atlantic to the Southern Ocean. This means that the heat and carbon carried by these water masses have a high chance of being moved across different density levels.

“Climate models do account for turbulence, but mostly in how it affects ocean circulation,” said Cimoli. “But we’ve found that turbulence is vital in its own right, and plays a key role in how much carbon and heat gets absorbed by the ocean, and where it gets stored.”

“Many climate models have an overly simplistic representation of the role of micro-scale turbulence, but we’ve shown it’s significant and should be treated with more care,” said Mashayek. “For example, turbulence and its role in ocean circulation exerts a control over how much anthropogenic heat reaches the Antarctic Ice Sheet, and the timescale on which that happens.”

̽��ֱ��research suggests an urgent need for the instalment of turbulence sensors on global observational arrays and a more accurate representation of small-scale turbulence in climate models, to enable scientists to make more accurate projections of the future effects of climate change.

̽��ֱ��research was supported in part by the Natural Environment Research Council (NERC), part of UK Research and Innovation (UKRI).

Reference:
Laura Cimoli et al. ‘Significance of Diapycnal Mixing Within the Atlantic Meridional Overturning Circulation.’ AGU Advances (2023). DOI: 10.1029/2022AV000800

Underwater waves deep below the ocean’s surface – some as tall as 500 metres – play an important role in how the ocean stores heat and carbon, according to new research.

Turbulence plays a key role in how much carbon and heat gets absorbed by the ocean, and where it gets stored

Laura Cimoli

Laura Cimoli/GLODAP

Map of depth-integrated anthropogenic carbon

̽��ֱ��text in this work is licensed under a Creative Commons Attribution 4.0 International License. Images, including our videos, are Copyright © ̽��ֱ�� of Cambridge and licensors/contributors as identified. All rights reserved. We make our image and video content available in a number of ways – as here, on our main website under its Terms and conditions, and on a range of channels including social media that permit your use and sharing of our content under their respective Terms.

Yes

No ‘safest spot’ to minimise risk of COVID-19 transmission on trains

sc604 — Wed, 22 Jun 2022 04:00:00 +0000

̽��ֱ��researchers, from the ̽��ֱ�� of Cambridge and Imperial College London, developed a mathematical model to help predict the risk of disease transmission in a train carriage, and found that in the absence of effective ventilation systems, the risk is the same along the entire length of the carriage.

̽��ֱ��model, which was validated with a controlled experiment in a real train carriage, also shows that masks are more effective than social distancing at reducing transmission, especially in trains that are not ventilated with fresh air.

̽��ֱ��results, reported in the journal Indoor Air, demonstrate how challenging it is for individuals to calculate absolute risk, and how important it is for train operators to improve their ventilation systems in order to help keep passengers safe.

Since COVID-19 is airborne, ventilation is vital in reducing transmission. And although COVID-19 restrictions have been lifted in the UK, the government continues to highlight the importance of good ventilation in reducing the risk of transmission of COVID-19, as well as other respiratory infections such as influenza.

“In order to improve ventilation systems, it’s important to understand how airborne diseases spread in certain scenarios, but most models are very basic and can’t make good predictions,” said first author Rick de Kreij, who completed the research while based at Cambridge’s Department of Applied Mathematics and Theoretical Physics. “Most simple models assume the air is fully mixed, but that’s not how it works in real life.

“There are many different factors which can affect the risk of transmission in a train – whether the people in the train are vaccinated, whether they’re wearing masks, how crowded it is, and so on. Any of these factors can change the risk level, which is why we look at relative risk, not absolute risk – it’s a toolbox that we hope will give people an idea of the types of risk for an airborne disease on public transport.”

̽��ֱ��researchers developed a one-dimensional (1D) mathematical model which illustrates how an airborne disease, such as COVID-19, can spread along the length of a train carriage. ̽��ֱ��model is based on a single train carriage with closing doors at either end, although it can be adapted to fit different types of trains, or different types of transport, such as planes or buses.

̽��ֱ��1D model considers the essential physics for transporting airborne contaminants, while still being computationally inexpensive, especially compared to 3D models.

̽��ֱ��model was validated using measurements of controlled carbon dioxide experiments conducted in a full-scale railway carriage, where CO2 levels from participants were measured at several points. ̽��ֱ��evolution of CO2 showed a high degree of overlap with the modelled concentrations.

̽��ֱ��researchers found that air movement is slowest in the middle part of a train carriage. “If an infectious person is in the middle of the carriage, then they’re more likely to infect people than if they were standing at the end of the carriage,” said de Kreij. “However, in a real scenario, people don’t know where an infectious person is, so infection risk is constant no matter where you are in the carriage.”

Many commuter trains in the UK have been manufactured to be as cheap as possible when it comes to passenger comfort – getting the maximum number of seats per carriage. In addition, most commuter trains recirculate air instead of pulling fresh air in from outside, since fresh air has to be either heated or cooled, which is more expensive.

So, if it’s impossible for passengers to know whether they’re sharing a train carriage with an infectious person, what should they do to keep themselves safe? “Space out as much as you reasonably can – physical distancing isn’t the most effective method, but it does work when capacity levels are below 50 percent,” said de Kreij. “And wear a high-quality mask, which will not only protect you from COVID-19, but other common respiratory illnesses.”

̽��ֱ��researchers are now looking to extend their 1D-model into a slightly more complex, yet still energy-efficient, zonal model, where cross-sectional flow is characterised in different zones. ̽��ֱ��model could also be extended to include thermal stratification, which would offer a better understanding of the spread of an airborne contaminant.

̽��ֱ��research was funded in part by the Engineering and Physical Sciences Research Council (EPSRC), part of UK Research and Innovation (UKRI).

Reference:
Rick JB de Kreij et al. ‘Modelling disease transmission in a train carriage using a simple 1D-model.’ Indoor Air (2022). DOI: 10.1111/ina.13066

Researchers have demonstrated how airborne diseases such as COVID-19 spread along the length of a train carriage and found that there is no ‘safest spot’ for passengers to minimise the risk of transmission.

We hope this research will give people an idea of the types of risk for an airborne disease on public transport

Rick de Kreij

Seksan Mongkhonkhamsao via Getty Images

Woman wearing a mask on public transport

Yes

Experiment evaluates the effect of human decisions on climate reconstructions

sc604 — Mon, 07 Jun 2021 09:21:35 +0000

̽��ֱ��experiment, designed and run by researchers from the ̽��ֱ�� of Cambridge, had multiple research groups from around the world use the same raw tree-ring data to reconstruct temperature changes over the past 2,000 years.

While each of the reconstructions clearly showed that recent warming due to anthropogenic climate change is unprecedented in the past two thousand years, there were notable differences in variance, amplitude and sensitivity, which can be attributed to decisions made by the researchers who built the individual reconstructions.

Professor Ulf Büntgen from the ̽��ֱ�� of Cambridge, who led the research, said that the results are “important for transparency and truth – we believe in our data, and we’re being open about the decisions that any climate scientist has to make when building a reconstruction or model.”

To improve the reliability of climate reconstructions, the researchers suggest that teams make multiple reconstructions at once so that they can be seen as an ensemble. ̽��ֱ��results are reported in the journal Nature Communications.

Information from tree rings is the main way that researchers reconstruct past climate conditions at annual resolutions: as distinctive as a fingerprint, the rings formed in trees outside the tropics are annually precise growth layers. Each ring can tell us something about what conditions were like in a particular growing season, and by combining data from many trees of different ages, scientists are able to reconstruct past climate conditions going back hundreds and even thousands of years.

Reconstructions of past climate conditions are useful as they can place current climate conditions or future projections in the context of past natural variability. ̽��ֱ��challenge with a climate reconstruction is that – absent a time machine – there is no way to confirm it is correct.

“While the information contained in tree rings remains constant, humans are the variables: they may use different techniques or choose a different subset of data to build their reconstruction,” said Büntgen, who is based at Cambridge’s Department of Geography, and is also affiliated with the CzechGlobe Centre in Brno, Czech Republic. “With any reconstruction, there’s a question of uncertainty ranges: how certain you are about a certain result. A lot of work has gone into trying to quantify uncertainties in a statistical way, but what hasn’t been studied is the role of decision-making.

“It’s not the case that there is one single truth – every decision we make is subjective to a greater or lesser extent. Scientists aren’t robots, and we don’t want them to be, but it’s important to learn where the decisions are made and how they affect the outcome.”

Büntgen and his colleagues devised an experiment to test how decision-making affects climate reconstructions. They sent raw tree ring data to 15 research groups around the world and asked them to use it to develop the best possible large-scale climate reconstruction for summer temperatures in the Northern hemisphere over past 2000 years.

“Everything else was up to them – it may sound trivial, but this sort of experiment had never been done before,” said Büntgen.

Each of the groups came up with a different reconstruction, based on the decisions they made along the way: the data they chose or the techniques they used. For example, one group may have used instrumental target data from June, July and August, while another may have only used the mean of July and August only.

̽��ֱ��main differences in the reconstructions were those of amplitude in the data: exactly how warm was the Medieval warming period, or how much cooler a particular summer was after a large volcanic eruption.

Büntgen stresses that each of the reconstructions showed the same overall trends: there were periods of warming in the 3^rd century, as well as between the 10^th and 12^th century; they all showed abrupt summer cooling following clusters of large volcanic eruptions in the 6^th, 15^th and 19^th century; and they all showed that the recent warming since the 20^th and 21^st century is unprecedented in the past 2000 years.

“You think if you have the start with the same data, you will end up with the same result, but climate reconstruction doesn’t work like that,” said Büntgen. “All the reconstructions point in the same direction, and none of the results oppose one another, but there are differences, which must be attributed to decision-making.”

So, how will we know whether to trust a particular climate reconstruction in future? In a time where experts are routinely challenged, or dismissed entirely, how can we be sure of what is true? One answer may be to note each point where a decision is made, consider the various options, and produce multiple reconstructions. This would of course mean more work for climate scientists, but it could be a valuable check to acknowledge how decisions affect outcomes.

Another way to make climate reconstructions more robust is for groups to collaborate and view all their reconstructions together, as an ensemble. “In almost any scientific field, you can point to a single study or result that tells you what to hear,” he said. “But when you look at the body of scientific evidence, with all its nuances and uncertainties, you get a clearer overall picture.”

Reference:
Ulf Büntgen et al. ' ̽��ֱ��influence of decision-making in tree ring-based climate reconstructions.’ Nature Communications (2021). DOI: 10.1038/s41467-021-23627-6

̽��ֱ��first double-blind experiment analysing the role of human decision-making in climate reconstructions has found that it can lead to substantially different results.

Scientists aren’t robots, and we don’t want them to be, but it’s important to learn where the decisions are made and how they affect the outcome

Ulf Büntgen

Hrafn Óskarsson

Subfossil trees preserved in Iceland

Yes

̽��ֱ��uncertain unicycle that taught itself and how it’s helping AI make good decisions

lw355 — Wed, 14 Feb 2018 13:51:19 +0000

In the centre of the screen is a tiny unicycle. ̽��ֱ��animation starts, the unicycle lurches forward and falls. This is trial #1. It’s now trial #11 and there’s a change – an almost imperceptible delay in the fall, perhaps an attempt to right itself before the inevitable crash. “It’s learning from experience,” nods Professor Carl Edward Rasmussen.

After a minute, the unicycle is gently rocking back and forth as it circles on the spot. It’s figured out how this extremely unstable system works and has mastered its goal. “ ̽��ֱ��unicycle starts with knowing nothing about what’s going on – it’s only been told that its goal is to stay in the centre in an upright fashion. As it starts falling forwards and backwards, it starts to learn,” explains Rasmussen, who leads the Computational and Biological Learning Lab in the Department of Engineering. “We had a real unicycle robot but it was actually quite dangerous – it was strong – and so now we use data from the real one to run simulations, and we have a mini version.”

Rasmussen uses the self-taught unicycle to demonstrate how a machine can start with very little data and learn dynamically, improving its knowledge every time it receives new information from its environment. ̽��ֱ��consequences of adjusting its motorised momentum and balance help the unicycle to learn which moves were important in helping it to stay upright in the centre.

“This is just like a human would learn,” explains Professor Zoubin Ghahramani, who leads the Machine Learning Group in the Department of Engineering. “We don’t start knowing everything. We learn things incrementally, from only a few examples, and we know when we are not yet confident in our understanding.”

Ghahramani’s team is pioneering a branch of AI called continual machine learning. He explains that many of the current forms of machine learning are based on neural networks and deep learning models that use complex algorithms to find patterns in vast datasets. Common applications include translating phrases into different languages, recognising people and objects in images, and detecting unusual spending on credit cards.

“These systems need to be trained on millions of labelled examples, which takes time and a lot of computer memory,” he explains. “And they have flaws. When you test them outside of the data they were trained on they tend to perform poorly. Driverless cars, for instance, may be trained on a huge dataset of images but they might not be able to generalise to foggy conditions.

“Worse than that, the current deep learning systems can sometimes give us confidently wrong answers, and provide limited insight into why they have come to particular decisions. This is what bothers me. It’s okay to be wrong but it’s not okay to be confidently wrong.”

̽��ֱ��key is how you deal with uncertainty – the uncertainty of messy and missing data, and the uncertainty of predicting what might happen next. “Uncertainty is not a good thing – it’s something you fight, but you can’t fight it by ignoring it,” says Rasmussen. “We are interested in representing the uncertainty.”

It turns out that there’s a mathematical theory that tells you what to do. It was first described by 18th-century English statistician Thomas Bayes. Ghahramani’s group was one of the earliest adopters in AI of Bayesian probability theory, which describes how the probability of an event occurring (such as staying upright in the centre) is updated as more evidence (such as the decision the unicycle last took before falling over) becomes available.

Dr Richard Turner explains how Bayes’ rule handles continual learning: “the system takes its prior knowledge, weights it by how accurate it thinks that knowledge is, then combines it with new evidence that is also weighted by its accuracy.

“This is much more data-efficient than the way a standard neural network works,” he adds. “New information can cause a neural network to forget everything it learned previously – called catastrophic forgetting – meaning it needs to look at all of its labelled examples all over again, like relearning the rules and glossary of a language every time you learn a new word.

“Our system doesn’t need to revisit all the data it’s seen before – just like humans don’t remember all past experiences; instead we learn a summary and we update it as things go on.” Ghahramani adds: “ ̽��ֱ��great thing about Bayesian machine learning is the system makes decisions based on evidence – it’s sometimes thought of as ‘automating the scientific method’ – and because it’s based on probability, it can tell us when it’s outside its comfort zone.”

Ghahramani is also Chief Scientist at Uber. He sees a future where machines are continually learning not just individually but as part of a group. “Whether it’s companies like Uber optimising supply and demand, or autonomous vehicles alerting each other to what’s ahead on the road, or robots working together to lift a heavy load – cooperation, and sometimes competition, in AI will help solve problems across a huge range of industries.”

One of the really exciting frontiers is being able to model probable outcomes in the future, as Turner describes. “ ̽��ֱ��role of uncertainty becomes very clear when we start to talk about forecasting future problems such as climate change.”

Turner is working with climate scientists Dr Emily Shuckburgh and Dr Scott Hosking at the British Antarctic Survey to ask whether machine learning techniques can improve understanding of climate change risks in the future.

“We need to quantify the future risk and impacts of extreme weather at a local scale to inform policy responses to climate change,” explains Shuckburgh. “ ̽��ֱ��traditional computer simulations of the climate give us a good understanding of the average climate conditions. What we are aiming to do with this work is to combine that knowledge with observational data from satellites and other sources to get a better handle on, for example, the risk of low-probability but high-impact weather events.”

“It’s actually a fascinating machine learning challenge,” says Turner, who is helping to identify which area of climate modelling is most amenable to using Bayesian probability. “ ̽��ֱ��data are extremely complex, and sometimes missing and unlabelled. ̽��ֱ��uncertainties are rife.” One significant element of uncertainty is the fact that the predictions are based on our future reduction of emissions, the extent of which is as yet unknown.

“An interesting part of this for policy makers, aside from the forecasting value, is that you can imagine having a machine that continually learns from the consequences of mitigation strategies such as reducing emissions – or the lack of them – and adjusts its predictions accordingly,” adds Turner.

What he is describing is a machine that – like the unicycle – feeds on uncertainty, learns continuously from the real world, and assesses and then reassesses all possible outcomes. When it comes to climate, however, it’s also a machine of all possible futures.

Inset image: read more about our AI research in the ̽��ֱ��'s research magazine; download a pdf; view on Issuu.

Cambridge researchers are pioneering a form of machine learning that starts with only a little prior knowledge and continually learns from the world around it.

This is just like a human would learn. We don’t start knowing everything. We learn things incrementally, from only a few examples, and we know when we are not yet confident in our understanding

Zoubin Ghahramani

̽��ֱ��District

̽��ֱ��text in this work is licensed under a Creative Commons Attribution 4.0 International License. For image use please see separate credits above.

Yes

New analysis of 'swine flu' pandemic conflicts with accepted views on how diseases spread

sc604 — Tue, 01 Jul 2014 08:12:49 +0000

̽��ֱ��most detailed analysis to date of the spread of the H1N1 2009 pandemic influenza virus, known informally as ‘swine flu’, has found that short-range travel was likely the primary driver for the 2009 pandemic in the United States, in contrast with popularly accepted views on the way diseases spread.

̽��ֱ��study, based on data gathered from health insurance claims made throughout 2009, found that international air travel, which was previously thought to be important in the pandemic, played only a minor role in its spread within the US.

A team of researchers from ̽��ֱ�� of Cambridge and the US, including Princeton ̽��ֱ�� and National Institutes of Health, analysed data from 271 American cities and their surrounding suburban areas, covering 90% of the population of the 48 contiguous states.

̽��ֱ��data were used to test mathematical models to pinpoint the role and importance of factors associated with H1N1’s arrival and spread, including demographics, school opening dates, humidity levels and immunity from previous outbreaks.

According to this new analysis, school-age children accelerated the spread of the pandemic, which was transmitted over short distances, in contrast with widespread reports at the time linking the pandemic to international air travel and population density. ̽��ֱ��results are published in the journal PLOS Computational Biology.

̽��ֱ��H1N1 influenza virus spread rapidly around the globe in 2009 after it was first identified in Mexico. ̽��ֱ��US Centers for Disease Control (CDC) estimates that the global death toll from the 2009 pandemic was more than 284,000.

̽��ֱ��virus hit the United States in two distinct waves during 2009: one in the spring, which was limited mainly to the North Eastern part of the country; and a second wave in the autumn, which started in the South Eastern US and gradually spread across the whole country over a period of three months.

Previous research on the pandemic found that environmental factors, timing of the end school holidays, population sizes and air travel contributed to its spread, but this new research has found that transmission occurred primarily over short distances and that school age children may have catalysed the spread.

“There is so little detailed analysis of the way in which pandemics spread – so much of the current thinking is based on opinion and presumption,” said Dr Julia Gog of Cambridge’s Department of Applied Mathematics and Theoretical Physics, who led the research. “We have a view of how diseases spread in the medieval times, and it is often said that the modern world is completely different as we have long distance high volume air travel. However what we find here is that although air travel must have caused the initial ‘sparks,’ the bulk of the pandemic wave was very slow, travelling at about 22 kilometres per day. This really challenges our views of modern pandemics.”

In collaboration with researchers from a number of American universities, Dr Gog analysed weekly numbers of patients who reported influenza-like illnesses to their physician, organised by zip code. ̽��ֱ��data enable the researchers to explore the pandemic’s spread in more detail than had previously been possible, as well as identify any gaps in the data and validate existing models.

“It’s remarkable that in an era of widespread air travel and regional ground transport that this pandemic spread so slowly,” said Dr Gog. While international air travel did play a role in the initial seeding of the outbreaks in spring 2009, it played a relatively minor role in the later spread. ̽��ֱ��researchers hypothesise that H1N1’s relatively low rate of transmission meant that in many cases, the virus failed to ‘take’ after being introduced to a community through air travel. In contrast, repeated transmission over short distances started chains of infection which then contributed to the overall spatial spread of the pandemic.

̽��ֱ��travel patterns of children also appear to have played an important role. While one might assume that the movements of children would primarily be between home and school, there is little available information to confirm this. “This is just not studied,” said Dr Gog. “Especially in densely populated areas, any parent will tell you that their child travels over relatively short distances each day, but we simply don’t have the data to back that up.”

In addition to broadening understanding the dynamics of a pandemic, the researchers hope that the potential role of children in influenza spread might lead to more information on this age group’s role in infection transmission.

New analysis of the 2009 H1N1 influenza pandemic in the US shows that the pandemic wave was surprisingly slow, and that its spread was likely accelerated by school-age children.

There is so little detailed analysis of the way in which pandemics spread – so much of the current thinking is based on opinion and presumption

Julia Gog

Mapping the spread of H1N1 influenza

Julia Gog

Map of the H1N1 influenza pandemic in 2009. ̽��ֱ��size of the dots is relative to city size, and the colours relate to the number of influenza cases, with green the lowest and purple the highest.

This work is licensed under a Creative Commons Licence. If you use this content on your site please link back to this page.

Yes

Modelling impacts of a warming world

lw355 — Wed, 03 Oct 2012 13:37:36 +0000

How different will the world be if it’s 2°C, 3°C or 4°C warmer? Ask this question of the multitude of different climate change impact models – each built by researchers interested in different aspects of global warming – and the likelihood is that you will get a multitude of answers. Modelling the global impact of climate change is an extremely complex process, and yet it’s absolutely essential if policy makers are to understand the consequences tomorrow of emissions policies adopted today.

Earlier this year, an international group of researchers initiated a joint project to attempt the first systematic quantification of some of the uncertainties surrounding climate change impacts to agriculture, health, biomes and water. Uncertainties such as: to what extent will the world’s vegetation change? Which regions will succumb to drought or flood? What will be the impact on global food crops? And how will the spread of human diseases be affected?

̽��ֱ��Inter-Sectoral Impact Model Intercomparison Project (ISI-MIP), coordinated by the Potsdam Institute for Climate Impact Research in Germany and the International Institute for Applied Systems Analysis in Austria, involves two- dozen research groups from eight countries.

Dr Andrew Friend from Cambridge’s Department of Geography is coordinating the analysis of results concerning changes to the world’s biomes – the communities of plants, animals and soil organisms that are characterised by a similar structure and climatic requirement.

It’s a fast-track programme. All of the teams are working to a tight deadline and, by January 2013, they hope to be ready to publish their findings on the likely impacts of climate change predictions. ̽��ֱ��collective results will contribute towards the next report of the Intergovernmental Panel on Climate Change (IPCC), the leading international body that provides a clear scientific view on the current state of knowledge in climate change and its potential environmental and socioeconomic impacts.

Each group is using their own model, together with the very latest climate predictions produced by leading climate modelling groups around the world, to run comparable simulations for four different warming scenarios – conservative, drastic and two in between – from 1950 to 2099.

For Friend, this means Hybrid, the model he first developed 15 years ago. At its heart is a set of equations that dynamically model global vegetation: “It works by simulating the dynamics of individual trees and an underlying herbaceous layer. You assume the plants compete within patches, and then scale these up to 50-km grid boxes. We use data to work out the mathematics of how the plant grows – how it photosynthesises, takes-up carbon and nitrogen, competes with other plants, and is affected by soil nutrients and water – and we do this for different vegetation types. Effectively, the whole of the land surface is understood in 2,500 km² portions. We then input real climate data up to the present and look at what might happen every 30 minutes to 2099.”

For the most extreme scenario of climate change being modelled, Friend expects to see significant impact: “this scenario could show the whole of the Amazon rainforest disappearing this century, depending on the climate model. ̽��ֱ��circumpolar Boreal forest, which began to emerge at the end of the last ice age, could migrate into the tundra and perish at its southern limit. By contrast, Russia may benefit from an increased ability to grow crops in regions that were previously too cold, and this greater productivity would help absorb atmospheric carbon.”

Modelling impacts is complex, as Friend explained: “the increase in CO₂ in the atmosphere from the burning of fossil fuels creates global warming. CO₂ can act on vegetation, increasing their rate of photosynthesis and therefore productivity. However, in heatwaves, ecosystems can emit more CO₂ than they absorb from the atmosphere. We saw this in the 2003 European heatwave when temperatures rose 6°C above mean temperatures and the amount of CO₂ produced was sufficient to reverse the effect of four years of net ecosystem carbon sequestration.”

One of the greatest uncertainties in climate change is the feedback from changes in terrestrial carbon cycling. “Many scientists think that if soil warms it will release carbon because of the increased breakdown of dead organic matter by bacteria and fungi,” added Friend. “But there’s a lot of debate over whether this stimulation will be sustained over a long time – if it is, then you end up releasing enormous amounts of CO₂ into the atmosphere, causing further global warming.”

Working with PhD student Rozenn Keribin, Friend is using Darwin, the ̽��ֱ��’s High Performance Computing Service’s supercomputer, to run the simulations; what takes a month to perform on a PC can be easily accomplished overnight on Darwin.

As the results of each group’s simulations become available over the coming months, the data will be assembled and compared. Friend fully expects that this process will reveal differences: “each equivalent model has its own strengths and weaknesses. That’s why the comparison process is so valuable – no single model is sufficient but together we can reduce the uncertainty.”

Why is this so important? “To make policy you need to understand the impact of decisions,” said Friend. “There hasn’t been a coordinated impacts project for IPCC across sectors before, and now this is covering four key sectors across four climate change scenarios from multiple climate models. ̽��ֱ��idea is to understand at what point the increase in global temperature starts to have serious effects across all the sectors, so that policy makers can weigh up the probable impacts of allowing emissions to go above a certain level, and what mitigation strategies are necessary to avoid significant risk of dangerous climate change.”

For more information, please contact Louise Walsh (louise.walsh@admin.cam.ac.uk) at the ̽��ֱ�� of Cambridge Office of External Affairs and Communications.

A community-driven modelling effort aims to quantify one of the gravest of global uncertainties: the impact of global warming on the world’s food, health, vegetation and water.

Each model has its own strengths and weaknesses. That’s why the comparison process is so valuable – no single model is sufficient but together we can reduce the uncertainty.

Dr Andrew Friend

Andrew Friend

Initial ISI-MIP simulation showing the effects on vegetation productivity at the highest emissions scenario (reduction: red to yellow; increase: green to blue)

This work is licensed under a Creative Commons Licence. If you use this content on your site please link back to this page.

Yes