Past and current theses projects at FORM

Bachelor student: Maria Enarson
Supervisor: Dr. Mirela Beloiu Schwenke (mirela.beloiu(at)usys.ethz.ch); Prof. Dr. Christof Bigler (christof.bigler(at)env.ethz.ch).

Aim of the bachelor thesis: Maria will examine the effectiveness of protected areas, as categorized by the International Union for Conservation of Nature (IUCN), in maintaining structural complexity and thereby protecting valuable habitats and biodiversity. Structurally complex forests are of particular value, as they enhance biodiversity as well as functional diversity, making them more resilient against disturbances and possible consequences of climate change. As human activities have been shown to have a major impact on forest structural complexity, it is vital to understand how its conservation differs between IUCN protection categories.

Master student: Elias Berger
Supervisor: Dr. Mirela Beloiu Schwenke (mirela.beloiu(at)usys.ethz.ch); Prof. Dr. Arthur Gessler (arthur.gessler(at)usys.ethz.ch)

Aim of the master thesis: Elias aims to train deep learning algorithms for the identification and geolocation of healthy, stressed, and dead trees within diverse temperate forests using publicly accessible aerial orthophotos. Currently, monitoring forest health relies heavily on labor-intensive and costly field inventories. Although new deep learning methodologies show promising outcomes in identifying tree species, automatically monitoring health of various species within forests is still challenging.

The development of a fully automated assessment of individual tree vitality and mortality using low-cost sensors, would thus represent a cost-effective and efficient way to enhance forest health monitoring. Such a system could help identify problem areas with biotic or abiotic stressors and detect deadwood, benefitting both research and various practical applications.

Master student: Tomoki Loeillot
Supervisor: Mirela Beloiu (mirela.beloiu(at)usys.ethz.ch); Fanny Petibon (fanny.petibon(at)usys.ethz.ch)

Aim of the master thesis: Tomoki investigates recent changes in the elevational distribution of forest fires in Southern and Central Europe using geospatial data on fire occurrence and burned area. As a result of rising temperatures, longer drought periods, tree species range shifts, higher elevations are becoming increasingly susceptible to forest fires. Yet, few studies explored the evolution of the fire regime along an elevation gradient over the past few decades in Europe. Mountain forests ecosystems provide a wide range of services including supporting biodiversity, regulating water quality or stabilizing slopes. The aim of this work is to identify potential trends and their main drivers. This information is crucial for determining high risk regions that can be targeted for fire management practices.

Master student: Cécile Reichmuth
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Mirela Beloiu (mirela.beloiu(at)usys.ethz.ch)

Aim of the master thesis: Cécile applies a deep learning approach to identify different tree species in Swiss forests using high resolution orthophotos. Detection and identification of tree species is important in forest practice for improving forest management and assessing biodiversity. In recent years, deep learning approaches such as Convolutional Neural networks (CNNs) have evolved substantially and provide a promising but challenging method with high predictive accuracy. Moreover, remote sensing data is becoming cheaper through data collection with e. g. drones and is openly accessible in Switzerland. Therefore, the application of CNNs on remote sensing data has the potential to become a cost-effective and accurate method for researchers and forest practitioners to identify tree species in heterogeneous forests.

Bachelor student: Lina Muntwyler
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Arnadi Murtiyoso (arnadi.murtiyoso(at)usys.ethz.ch)

Aim of the bachelor thesis: Lina is working with low-cost 360 cameras to map forest environments. As 3D techniques become more and more ubiquitous, a trend towards their use in forest mapping also increases. However, most solutions available today are either costly (lidar, mobile laser scanning), time-consuming (TLS) or do not give a good enough quality (smartphones). The aim of Lina’s research is to see whether by using 360 videos we may develop a fast and low-cost solution that is precise enough for forestry needs, especially with an eventual National Forest Inventory (NFI) application in the future.

Master student: Lucas Flores Gutierrez
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Noëmi Brüggemann (noemi.brueggemann(at)usys.ethz.ch)

Aim of the master thesis: Lucas’ thesis focuses on the management of special forest reserves, in which specific species and habitats are promoted through targeted interventions. Although oriented towards nature conservation, such reserves do not exclude other forest functions such as timber production and protection against natural hazards making long-term planning in them particularly challenging.
Building on geospatial forest data, forest growth models, economic indicators and optimization methods he works towards identifying management alternatives on the strategic level that account for multiple objectives.
The primary goal is the creation of a basis for decision-making for the canton of Zug who is interested in converting their forests into special forest reserves.

Master student: Flavian Stocker
Supervisor: Mirela Beloiu (mirela.beloiu(at)usys.ethz.ch); Lars Waser (lars.waser(at)wsl.ch)

Aim of the master thesis: Flavian is focusing on the detection of standing dead trees using remote sensing technologies. In his Master’s thesis, he will work with already existing Swiss orthophotos (high resolution remotely sensed data) and deep learning approaches, such as Convolutional Neural Networks. To date, deep learning algorithms often focus on detecting living tree species and do not consider dead or low-vitality trees. However, this information is crucial for detecting deadwood, monitoring forest health, or getting information about biotic and abiotic stressors, and therefore for supporting forest management decisions.

Bachelor student: Adriana Niggeli
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch)

DownloadView thesis (PDF, 1.9 MB)vertical_align_bottom.

Bacehlor student: Paula Meri
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Anna Krucher

DownloadView thesis (PDF, 21.2 MB)vertical_align_bottom

Master student: Lucca Heinzmann
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Mirela Beloiu (mirela.beloiu(at)usys.ethz.ch)

Aim of the master thesis: The identification and mapping of trees using remote sensing data is an important task for practitioners and forest research. In recent years, research has also focused on using satellite and aircraft data to identify tree species. However, even though there have been steps in this direction, identifying trees in mixed forests remains a challenge. Tree species maps are essential for better forest management, forest cover, biomass, and biodiversity assessment. Here, we aim to develop a methodological approach that allows us to identify tree species in mixed forests.

Master student: Gioele Madonna
Supervisor: Verena Griess (verena.griess(at)usys.ethz.ch); Mirelo Beloiu (mirela.beloiu(at)usys.ethz.ch)

Aim of the master thesis: Drought- and heat-induced tree decline has been observed in various forest ecosystems around the world. It is usually assumed that vegetation recovers immediately and fully from drought, but this is inconsistent with basic plant physiology. Here, we aim to assess tree growth recovery after severe drought periods to determine drought legacy effects among tree species.  

Bearbeiter/in: Lioba Rath
Betreuer/in: Verena Griess (verena.griess(at)usys.ethz.ch); Leo Bont (leo.bont(at)wsl.ch); Janine Schweier (janine.schweier(at)wsl.ch)

Beschreibung Masterarbeit

Bearbeiter/in: Thomas Mutsaers
Betreuer/in: Verena Griess (verena.griess(at)usys.ethz.ch); Monika Niederhuber (monika.niederhuber(at)usys.ethz.ch)