Computers calculate how the glaciers move
In order to understand how glaciers move it is important to understand how they are affected by global warming. Today's technology allows researchers to both access and use much more data than before. However, this demands computer power to calculate and run models.
Increased global temperature means not only that the earth's glaciers are melting faster. It also means that those that run out into estuaries will slide out faster. This in turn affects melting as the underside of the glaciers comes into contact with water, which is relatively warm.
The research team that professor Veijo Pohjola is active in is investigating the process when glaciers slide over bedrock. His research area is Svalbard.
‘As global warming continues, it becomes increasingly important to take the temperature of glaciers that are changing.´
The work is performed, among others, under Svali, a Nordic research centre studying the climate environment and energy issues.
The development of technology has changed the basis of glacier research.
‘In the beginning of your career you travelled around and took measurements on a glacier with an instrument and performed an analysis. Now we gather a great deal more information and conduct advance data processing that requires smart algorithms. Today you need to pair climate models, an ice flow model, balance models and databases of elevation models.´
The researchers model how a glacier moves in order to predict how it will move in the future, run the model and compare with the data they collect in the field. Input data is the gravitational field from the glacier, which gives a picture of the mass distribution, how the glacier surface moves and changes, the energy flow in the glacier and precipitation.
It is running the models that demand substantial computing power. This applies in particularly to the ice flow measurements. They must be repeated many times with slightly different parameters. Frequently this is done in several steps, where researchers use the results from previous runs to improve the input of the next. This creates large amounts of data that must be analysed.
‘The fact that we can perform more accurate analysis today, is obviously positive. We get better results. But it also becomes extremely complicated, so it’s not always so easy to understand the results,’ says Veijo Pohjola and laughs.
Veijo Pohjola and his research group collaborate with many Swedish and international teams. Data exchange is an integral part of the research.
‘The sphere we work in is quite small and most people know each other, which means that you trust that you will be invited as an author if the work leads to a publishable result. But if you do not know the people, it may advisable to enter into an agreement about how data can be used.’
Kim Bergström