Institute for Biodiversity and Ecosystem Dynamics

Focus on research: Earth scientist Harry Seijmonsbergen

Harry Seijmonsbergen

Photo: Bob Bronshoff

Earth scientist Harry Seijmonsbergen of the Computational Geo-Ecology research group has a genuine love of landscapes. Both his research and lectures successfully combine classic fieldwork with the latest computer technologies, such as GIS and Remote Sensing.

An important part of Seijmonsbergens' research is devoted to assessment of natural hazards. ‘This type of information is crucial to landscape management. For example, we need the ability to predict landslides', he explains. ‘I try to find clear-cut scientific criteria that will allow us to determine and prevent natural hazards. We can then apply these criteria to provide solid guidelines on preserving geomorphologic diversity.'

GIS

Seijmonsbergen describes himself as an earth scientist from the old school: going out into the field, taking soil samples, getting your hands dirty. He specialises in geomorphology: the science of describing and explaining landforms and the processes that shape them. ‘In the past, research mainly consisted of describing geomorphologic processes. Over the last 15 years, that process has become increasingly automated, with researchers using GIS, which stands for Geographic Information Systems.'

GIS is a collection of hardware and software designed to collect, analyse, present and visualise spatial data. With a few clicks of a mouse, geomorphologic information can now be linked to ecological, geographic and biological data, and visualised on a computer screen in various map layers. A lot of this information is stored in databases that can be linked to maps. ‘For example, GIS allows you to calculate the correlation between the distribution of plant species and abiotic characteristics, such as the geological subsurface, the distribution of materials or places likely to experience landslides. We can also add variables such as water levels, fauna, ground levels and human influences. The use of Digital Terrain Models (DTMs) is also invaluable, as it provides a great deal of quantitative insight into landscape development. GIS allows scientists to link up all the data they need to solve a specific problem.'

Field work

Today's earth scientists have access to a large number of technological tools that make their job easier. ‘We didn't have GPS in the old days. You had to use an altimeter, which wasn't necessarily all that accurate. Altimeters use air pressure, which meant they had to be constantly recalibrated to certain points on the elevation map.' Although today's fieldwork is still largely similar to the methods used in the past, Seijmonsbergen does see significant differences. ‘I often went out on my own to carry out measurements. Nowadays, students aren't allowed to do that anymore, they have to work in teams of at least two. The reason: if you're standing in a river bed in the mountains and it starts to rain, you need to get out in a hurry. One time in Switzerland, lightning struck about 15 metres from where I was standing. That left me deaf for a few minutes.'

Seijmonsbergen enjoys going abroad as part of his work. Each year, he gives a course for Master's students on Tenerife and takes second-year students on a field trip to Luxembourg, where he teaches them to survey with both traditional methods and GIS. His personal favourite, though, is the province of Vorarlberg in the far west of Austria.

UvA scientists have been doing research in this part of Alps for 30 years now. ‘It's really become my neck of the woods', Seijmonsbergen admits. ‘Every two years, I spend a few weeks there with a group of students or a doctoral candidate.' The area has long served as a site for research on climate reconstruction and slope instability, which offers a crucial advantage: the opportunity to combine old and new surveying data. ‘We can assess the validity of old ideas using the very latest technologies. Landscapes aren't static, they're constantly evolving. Rivers take a different course, landslides affect water levels and local flora and fauna.' Armed with a mobile version of GIS with built-in GPS, the students are sent out to survey the area. The use of new high-resolution elevation models based on laser technology have become an integral part of the process. Automatic surveying based on this latest data seems to be the future. ‘It used to take a great deal of time to update maps. These days, GIS allows us to do that with far less effort.'

Remote sensing

Last year, Seijmonsbergen taught an optional course in Peru, which he visited to survey difficult terrain. He had students study the vulnerability of paramo grasslands, located high in the Andes. They surveyed a small, inaccessible and relatively difficult terrain, assessing factors such as degradation, the effects of open-pit mining and the potential for CO2 storage. The resulting data will be combined with satellite images in order to extrapolate the results to a larger area. ‘The greatest threat to this area comes from changing land use and climate change, which causes fragmentation and degradation of the landscape. We keep track of these developments by means of remote sensing. From the 1970s onwards, we've been using infrared satellite sensors to carry out landscape measurements. Infrared light can measure certain data that is invisible to the naked eye. This allows us to discern differences in terms of vegetation, such as the distinction between deciduous and coniferous trees. Amongst other applications, the analysis and regular comparison of the images and data we gather at these locations allows us to monitor illegal foresting.'

Sharing knowledge

Despite the fact that academic staff are assessed in terms of the quality of their research rather than their lectures, Seijmonsbergen is highly committed to teaching. He was elected Lecturer of the Year by his own students on several occasions. ‘I was actually surprised to find I enjoy teaching as much as I do. Over the years, I've fully mastered the subject matter and gained so much experience that I now enjoy sharing my knowledge with others. I also try to innovate by using tools such as Google Earth and incorporating the latest research results in my lectures and practical training.' Seijmonsbergen hasn't let this logical progression go to his head. ‘I may be lucky to teach subjects such as landscape analysis, GIS, remote sensing and various practical fieldwork courses. Most Earth Science students tends to love those subjects.'

Seijmonsbergen's colleagues and students appreciate his ability to combine fieldwork with modern computer technology, the majority of which takes place in the new GIS studio in the Science Park. ‘You can do an a great deal from behind a computer. You create a digital environment in which you can combine images, rotate them in 3D, zoom in: the possibilities are limitless. Those technologies are incredibly helpful tools in interpreting landscape processes from a distance. Although we're making more and more use of computer models, the link to the real world remains essential. In order to understand the virtual world, you must always combine data with actual field measurements. You can do a lot using automated, remote systems, but the human brain can tell more about an image in one look than a complex computer programme. With all the powerful new computers, software and satellite data at our disposal, I regard digital landscape surveying as a new challenge to which I can apply my extensive experience in the field.'

Published by  Faculty of Science

20 August 2012