MSc thesis subject: Space for time – how do peatland vegetation patterns develop through time?

Boreal peatlands are wet ecosystems containing large amounts of carbon, stored in the form of dead plant remains. When the climate changes, this carbon may be released into the atmosphere, further accelerating climate change. Remote sensing plays a key role in monitoring and predicting how exactly peatland ecosystems will respond to climate change.


The vegetation of boreal peatlands determines the rate of carbon uptake and is therefore an important characteristic to monitor. During the development of peatlands the vegetation (hence carbon uptake) changes. Typically, peatlands show a remarkable systematic patterned structure. How this patterned structure develops over time is important to understand peatland functioning and consequences of climate change. It is, however, very hard (if not impossible) to determine how vegetation patterns change over time, as such patterns are buried by newly formed peat.

A way to understand how peatland patterns changed is by taking an approach where space is substituted for time. After glaciers retreated since the last ice age in Scandinavia, the land surface that was previously lowered due to the weight of glaciers uplifted. As a result, the cover of the Baltic sea reduced and the land cover increased. Increasingly more land became available for peat forming plants to grow, resulting in a gradient of peat initiation, with younger peatlands occurring along the sea shores.

This unique setting provides the opportunity to take a ‘space-for-time’ approach, and analyse how vegetation patterns change in peatlands along the time-sequence.


  • Determine how vegetation patterns change along a time gradient in Sweden/Finland using classified high-res digital aerial images (RGB + IR) and terrain models.
  • Calculate vegetation pattern properties (such as connectivity, patch size, aggregation)
  • Validate vegetation patterns in a field campaign in Sweden or Finland


  • Experience with programming (e.g. R), classification, and working with big spatial datasets is valuable but not necessary.

Theme(s): Sensing & measuring; Integrated Land Monitoring