Department of Information Technology

Researcher profile: Gunilla Kreiss

Gunilla Kreiss

Gunilla Kreiss uses so-called numerical simulations to develop models of real-life conditions: "I think it's exciting that we can now move the experiments into the computer instead of just making measurements in a lab". Photo: Mats Kamsten

SHE FINDS NUMEROUS SOLUTIONS TO PHYSICAL PROBLEMS

Gunilla Kreiss is more than happy to zero in on mathematical algorithms, but she also needs to know that her models will come to practical use. Through calculations of physical phenomena, she aims to merge ideas with reality.

“It’s true that the numerical calculations I make on the computer come out as digits, but those digits don’t have anything to do with real life until someone else uses them.”

The professor of numerical analysis becomes animated when she describes how she decided on her research topic. During her civil engineering studies at KTH Royal Institute of Technology, she was so captivated by numerical analysis that she never wanted the courses to end.

“I felt that there was a link between this real problem and its solution – let’s make a model to solve this mathematical problem. I liked that you could actually see that there was a chain,” says Kreiss.

“I was also fascinated at how they moved the experiment into the computer in the form of these numerical simulations instead of doing measurements in a lab.”

Today, computers can be used to quickly and effectively solve various types of calculation problems. Mathematical problems can be transformed into numerical calculation models in a number of different ways. These are tested and combined from countless angles and approaches to find the best possible calculation model. But despite all this efficiency, the challenge still remains to analyse the methods used, says Kreiss. The computer hasn’t quite replaced the blackboard, though.

“Writing on the board is great stuff,” she laughs. “And bouncing ideas off each other, talking about things, that’s the brainstorm we talk about. The calculation itself can then partially be done using analysis. I often use classical mathematical analysis.”

In addition to computer simulations, the methods developed in numerical analysis or scientific computing are applied to problems in many different areas. Gunilla Kreiss conducts research on physical phenomena that can be modelled using partial differential equations, and she is collaborating on one such project with physicists at Uppsala University.

“My role is to develop and analyse simulation tools that are then used to develop small magnetic information carriers for computer data storage. Magnetism is crucial to being able to store and send information. Ultimately, I hope to contribute to the development of even more efficient computer data storage.” says Kriess.

This is one of several projects she leads as part of the research initiative Essence of e-science. In another project, Kreiss is collaborating with geoscientists to develop calculation models using simulations of elastic waves in the earth’s crust. These models can help geologists predict earthquake scenarios in a range of geographical areas.

“Both projects involve a reality that is not so distant – and one in which I can see that my work can make a difference. But the benefits don’t have to appear in the very next stage – they can also come further down the line. And that’s totally fine, as long as I can see that there is a connection.”

But being able to see the logical consequences isn’t the most difficult thing, says Kreiss. It’s formulating the questions.

How does one go about formulating the best questions?

“It’s hard! You have to twist and turn and test, and then decide ‘no, I’m not going to be able to answer this,’ or ‘no, that’ll be too easy”. Because that’s how it is. You might start out with the most difficult question and find that it won’t work – it won’t lead to a solution. Then you have to simplify. At the same time, you have to try to find a balance so that you don’t simplify too much so that you take away everything essential.”

Interview by Anneli Björkman, 2016

FACTS - GUNILLA KREISS

Title: Professor of Numerical Analysis at the Department of Information Technology.

Age: 58 år.

Lives in: Nacka, Stockholm.

Family: Husband and three sons.

Likes: Carpentry and building things.

Greatest research success: An idea that I implemented about fifteen years ago when I was working at KTH. My doctoral student and I were examining how we could make a computer model of what happens to a drop of oil when it gets into a steam of water and is deformed. This sort of thing is important to be able to describe for applications such as small pumping devices where bubbles are sorted according to size. We got an idea which was then implemented by a software company called Comsol. When my doctoral student finished her dissertation, she got a job at Comsol precisely for this computer model we had developed. It was so great. And the company still uses our discovery.”

In her free time: I’m in the Järla orientering club, and do tour skating and cross country skiing in the winter. I like outdoor activities and exercising. If I go downhill, I go on a splitboard, which is a snowboard you can separate into two parts. You attach climbing skins under the parts to go up the slope and then put them back together and snowboard down. We often take skiing trips to Norway, which has amazing mountains.

Updated  2016-12-09 14:38:28 by Peter Waites.