How do we make predictions?

Arguably, one of the most important tasks of a glaciologist in the modern world is to try and predict what might happen to glacier and ice sheets in the future. This is an important consideration because of our concerns about climate change, and the implications of this for the planet’s ice bodies.

In order for us to be able to make predictions about the future, we of course need to gain as full an understanding as possible of current and past glacier behaviour; in other words, it is only through developing as full an understanding as possible of how glaciers have responded and changed in the past, as well as how they interact with their environment and are impacted by climate, can we hope to make any sort of predictions.

Once we have a good understanding of glacier/climate/environment interactions, then we can tentatively try to work out what might happen to them in the years to come, as a result of our changing climate.

Predictions and modelling

We make these predictions using models. Scientific models are simplifications of reality, so in the case of a glacier or ice sheet, we might try to come up with a series of mathematical equations that accurately describe the key relationships that operate.

So, for example, through our understanding of how much glaciers have retreated and melted in the past, we might be able to derive equations that explains the relationship between the amount of incoming solar radiation and how much melt is generated. From long-term observations of glacier motion, we might also devise equations that describe the speed with which glaciers move across their underlying beds, taking into account how rough the terrain they are moving over is, and what happens to sliding velocities when water is added to the bed.

We might also be able to derive equations to describe what happens at the boundary between ice and large bodies of water such as lakes and even the ocean – this is particularly important for some glaciers, particularly many of the outlet glaciers and ice streams that are part of the Greenland and Antarctic ice sheets, since there are significant concerns about how ocean temperature change is impacting on glacier melt and behaviour.

Finally, we might also derive relationships to explain the complex internal behaviour of ice itself – i.e. how it deforms and responds to different stresses.

If we can use all this knowledge to build a series of equations describing these various relationships, then we are able to build a model of glacier behaviour. If that model (and its constituent equations) does a good job of representing reality, but it is simplified to the extent that we can describe our glacier in a clear and understandable way, then it becomes very powerful to us. In modelling, we are striving to find a balance between complexity (such that our model successfully mimics reality) and simplicity (such that it is not so complex that it become unwieldy and difficult for us to use). If we can do this, then we have a very powerful tool.

Why we use modelling

A glacier or ice sheet model might be used for a number of different reasons and to do a number of different things. In constructing a model, it forces us to think about the processes involved, it can help us to fill in data gaps, to simulate the past, and perhaps most importantly, to predict the future. Modelling has the potential to provide us with unparalleled insights into the future of glaciers and ice sheets, specifically relating to exploring how glaciers might change in the future in response to a warming climate, and what their contribution to sea level rise might be.

What are the pros and cons of predicting using modelling? Please leave your ideas and comments below.