Modelling Climate Change
(from Neil Adams NAACE)
Climate change is very much in the minds of many young people, but how do we know that the dire predictions are true? It’s all about the application of silicon chips rather than crystal balls and gives ICT teachers a route into discussing models, at least with brighter students or as a summary of the principles that have already been taught.
Models are based on variables and rules designed to represent a real world situation, whether of the travel of a roller coaster, the flight of a moon rocket or changes to the environment. The variables represent known quantities like the global temperature and levels of carbon dioxide, whereas rules are mathematical statements of the assumed linkages between the parts of the system. A simple rule may state that if the level of carbon dioxide doubles the global temperature will rise by 2C. Of course it is not so simple, since a further doubling of carbon dioxide would be expected to raise the temperature rather more.
The suggested rise of 2C with doubling of carbon dioxide that scientists from the Intergovernmental Panel on Climate Change (IPCC) have given is not actually a rule but an output in this particular model. The change in temperature due to carbon dioxide levels is the particular variable that is of interest to scientists, so it cannot be used as an underlying assumption in the construction of the model itself.
Some of the variables in climate models could include:
Levels of carbon dioxide
Air temperatures
Sea temperatures
Cloud cover
Levels of carbon dioxide emissions (from cars, factories etc.)
Vegetation cover (as plants use up carbon dioxide)
Simple rules could be stated as follows:
Infrared heat radiated from the earth’s surface tends to be absorbed and heat carbon dioxide molecules more than oxygen or other gases. (This is the “greenhouse effect”.)
As the amount of carbon dioxide in the atmosphere increases, the global air temperature will rise.
As plants grow levels of carbon dioxide are reduced, as it is used in the process of photosynthesis.
More daytime cloud will reflect sunlight and cool the planet.
More night time cloud will trap heat and warm the planet.
An increase in soot and other “aerosols” will cool the planet by reducing sunlight that reaches the surface.
Those rules may not sound very simple, but each one needs to be quantified on the basis of known science; if they are not expressed as a mathematical formula they cannot be represented in the computer model. It is important, for example, to know how much carbon dioxide is taken up by a field of wheat and to what extent this differs from a pond full of water lilies or an area of rain forest. On a global scale there are few water lily ponds, so these can probably be ignored, but this demonstrates the need to select the important variables (areas of rainforest and land under agricultural production) from these that are less important.
The rule that links a hectare of wheat to the amount of carbon dioxide taken up is affected by a number of variables including the fertility of the ground, the amount of sunlight available, rainfall and - oh dear - temperature. I say that because the change in temperature is precisely what we are trying to establish, but it seems to be part of the model itself. This illustrates feedback: if plants are too cold they will die and eliminate no carbon dioxide; if they are “just right”, they will reduce the amount of carbon dioxide in the atmosphere, which in turn will lower the rate at which sunlight can heat the atmosphere; if plants are too hot they will shrivel up and not help remove carbon dioxide, so temperatures will keep on rising, more plants will die and the effect could get larger and larger.
It all gets much more complicated when you realise that the hot air and warm water just will not stay where they are. There are local convection currents in the atmosphere causing thunderstorms; movements of whole weather systems causing both drizzle and hurricanes, and ocean currents such as the Gulf Stream. These work on a vertical level as well as horizontally, so climate models must include ways to calculate movement of air, formation of clouds and other processes in three dimensions.
The atmospheric model currently used by the Met Office uses columns that measure 135×135km over land, divided vertically into 38 “cells”. Compared to the 1990s, this is twice the “resolution” in each dimension, giving eight times as many cells through which to calculate the movement of air. This means that we have had better forecasts (honest!) but needs much greater computer power - an increase of 256-fold since the 1970s. By the standards of the day (whichever decade you pick), these models have needed extremely powerful super computers to run them to produce both short-term (weather) and long-term (climate) predictions.
A scientist from Reading University has said that we really need computers 1000 times more powerful than those now available to produce a good model of global climate based on what we know at present. However, there are many things we still don’t know in enough detail, such as the degree to which clouds will act as reflectors of heat or insulators and what happens to carbon dioxide locked up in the soil as temperatures rise.
How might you use this in the classroom? Students could be asked to read an article (such as those quoted from the BBC below) and identify:
Variables
Rules (written in common English)
Feedback mechanisms
Limitations of the model
Implications of incorrect predictions for policy-makers
One final point (from a geographer!): climate has always been changing. We know very little about past climates, even (in some significant variables like cloud distribution) as recently as ten years ago. This makes it very difficult to check models against known data and be certain whether the effects predicted are reasonable extrapolations or media-enabled scare-mongering. It seems clear that human factors are producing long term changes in climate, but these is sill a vocal minority of scientists that remain sceptical.
Most of the factual information used above has been drawn from two BBC News online articles (the first of which contains brief video commentaries and animations):
Models ‘key to climate forecasts’ http://news.bbc.co.uk/1/hi/sci/tech/6320515.stm
Climate prediction: No model for success http://news.bbc.co.uk/1/hi/sci/tech/7381250.stm
A helpful animation about the greenhouse effect is at
http://news.bbc.co.uk/1/shared/spl/hi/sci_nat/04/climate_change/html/greenhouse.stm
