For many years, photographs of emaciated, bedraggled polar bears clinging desperately to melting ice sheets have been circulating the internet, as newspapers wield headlines proclaiming record-breaking high temperatures, collapsing ice sheets, and rising sea levels. Dr Anne Le Brocq of the University of Exeter is taking a different approach to climate change awareness; the creation of a computer game. Anne is a Senior Lecturer of Physical Geography at Exeter University, specialising in ice sheet modelling and glaciology. She has spent time working in collaboration with the Natural Environment Research Council (NERC), the Climate Research Group (CRG), and has contributed to an IPCC Assessment on Climate Science. Recently however, a NERC grant has allowed Anne to explore a new area; game-based learning.
Launched about a month ago, Anne’s game Ice Flows has been attracting a lot of publicity. On the surface, the game is very a simple representation of the conditions in the Antarctic. An ice sheet is formed on land as falling snow is compacted. The ice sheet then moves out to sea, thinning as it goes, and eventually melting away. By modifying precipitation and temperature, the player can control the thickness of the ice sheet as it moves from the land out to sea. An increase in precipitation means an increase in the initial thickness of the ice sheet, whilst an increase in temperature means that the sheet melts faster. The aim of the game is to manipulate these two variables such that a penguin walking on top of the ice sheet is initially at the correct height to high-five a seagull (because as we all know, penguins and seagulls are best of friends), whilst also ensuring that that the ice melts a certain distance away from the land, so that the penguin can dive into the water at the right time to catch a shoal of fish. Whilst a child playing the game will be focussing on catching as many fish as possible, what they do not realise is that they are gaining knowledge of Antarctic ice sheets. This knowledge comes from two key areas:
Firstly, the game teaches children about the dynamic nature of ice sheets in the Antarctic; many people imagine ice sheets to be stationary objects, seeing any collapsing ice shelves as a natural disaster and clear sign of global warming. However, this is simply not true. In actual fact, as suggested by the game title Ice Flows, ice sheets are constantly moving from the land to the sea, melting as they go, and it is natural for the shelves to break off during this process. The question to ask, for those concerned about climate change, is about the dynamic equilibrium of the ice sheet; is the ice at sea melting faster than it is being formed on-land? If the answer is yes, then there will be an overall reduction in the mass of ice, and this can be attributed to climate change. The other key point is less related to the Antarctic itself, and is more about how the user controls the game. The two variables controlled by the user (precipitation and temperature) affect the ice sheet in different ways, and so each combination of the two produces a different outcome. By adjusting the two controls children can see how the ice depth changes. Modelling an outcome given a combination of input variables like this is essentially the basis computer modelling.
“Modelling the weather involves collecting data from a seemingly endless list of variables…”
A computer model uses equations to simulate real-life events, given a certain set of inputs. Although we often lament its inaccuracies, one of the most common everyday uses of computer modelling is for our morning weather reports. It often seems that in this age of technology, predicting whether or not it will rain tomorrow ought to be a breeze. However, this could not be further from the truth. Modelling the weather involves collecting data from a seemingly endless list of variables, such as air pressure, temperature, prevailing winds, and ocean currents, and then using these values as initial conditions for a forecast model. To run a standard 5-day forecast requires an incredible amount of computing power, and so the Cray supercomputer crunches data at a rate of over 23,000 trillion calculations per second in order to produce these results. The input values used for these forecasts come from real life data, and so it is unavoidable that they will contain inaccuracies. These inaccuracies are amplified the longer the model is run, and so predicting the weather for anything over two weeks is very difficult.
“Climate prediction relies on an understanding of the processes which occur within the atmosphere and oceans…”
Whilst weather forecasting relies on an accurate knowledge of the current state of the atmosphere, climate prediction uses a different type of model that is based on boundary conditions. Climate prediction relies on an understanding of the processes which occur within the atmosphere and oceans, i.e. how much sunlight reaches the earth, how reflective the earth’s surface is, how the oceans move, and how the presence of greenhouse gases affect the radiation escaping from the surface of the planet. So long as these boundary conditions are correct, the climate can be simulated for hundreds of years, regardless of the accuracy of the initial conditions. Even if the initial conditions are strange, only the first few years will be affected before the program settles down into pattern. To solve this problem, scientists often use ‘spin ups’; they run the simulation for an arbitrary amount of time (say 30 years), and then use the conditions shown after 30 years of the simulation as the starting conditions for the actual model.
To complicate climate modelling further, scientists have to consider the fact that many of the variables used are interconnected, often leading to feedback loops. There are two types of feedback loops; positive and negative. An example of a positive feedback loop is the ice/albedo feedback effect. The albedo of a material tells you how much light it reflects; the higher the albedo, the more light is reflected. Light coloured materials tend to have very high albedos, and ice, being white, is no exception, reflecting up to 70% of light shone onto it. However, with temperatures increasing and sea ice melting, more of the ocean is becoming visible. The ocean, with an albedo of only around 6%, absorbs far more light than ice, and so the overall albedo of the Antarctic drops as the ice melts.
This decrease in albedo means that more light is absorbed, causing temperatures to increase further. This in turn causes even more ice to melt, and so the cycle continues. These positive feedback loops are very dangerous, as they can have a runaway effect, and once they have started, they are very hard to stop. There are also negative feedback loops that slow down the rate of climate change, for example as the fact that the hotter the oceans get, the more CO2 they are able to hold (currently, about a third of CO2 emitted is absorbed by the oceans), but according to current models, these are just not enough to slow the rate of climate change.
“Hopefully this game, and others like it, will be a stepping stone to help understanding about climate change amongst the younger generation.”
There are many different types of computer models currently in use to predict the melting of ice sheets in Antarctica. They vary depending on the physics used to represent the ice flow, how the elements of equations are broken down, and the specific purpose of the code. Although these codes are all unique, it is commonly predicted that over the next few years, an increase in precipitation will cause ice sheets to be thicker on land. However, this will be offset by an increase in temperature, leading to the melting of the ice sheets, and the rising sea level that the newspapers so often mention. Although all of these models are incredibly complex, they are derived from the same basic ideas as Ice Flows; inputting variables and analysing the outcome. Hopefully this game, and others like it, will be a stepping stone to help understanding about climate change amongst the younger generation.bookmark me