MSc defense by Haiyin Ye

On Friday 20th of May at 9am Haiyin will present her M.Sc. thesis “Monitoring Gypsy Moth Outbreak Using Sentinel-1 and Sentinel-2 Data”. From the abstract: “Gypsy moth Lymantria dispar (L.) is a serious pest in deciduous forests, the larvae of which can feed on a large variety of host tree species. This study is focused on two objectives: 1) detect gypsy moth induced defoliation over large area using Sentinel-1 and Sentinel-2 data. 2) explore biotic and abiotic factors that may influence gypsy moth defoliation severity. Three sub-models and the final model were built up in the study. Sub-model 1 (tree species composition) revealed the benefits of Sentinel-2 data in tree species composition and species richness (SR) modeling in oak-dominant forests, with R2 of 0.846 (NMDS1) and 0.72 (SR). Sub-model 2 (stand structure heterogeneity) demonstrated the high performance of Sentinel-1 radar data in stand structure modelling, regarding both vertical structure heterogeneity (PR_CANUND, R2 = 0.731) and horizontal structure heterogeneity (total area of gaps, R2 = 0.576). Sub-model 3 (defoliation severity) quantified defoliation severity in both the intensive sampling area (n = 38) and large area (n = 2800) by as the defoliation index. The final model (factor exploration) showed that the factors considered in this study could explain 46.77% and 17.14% of the defoliation index in intensive sampling area (n = 38) and large area (n = 2800) respectively. The importance of these factors and their correlation with defoliation index were depicted as well. Caterpillar density, egg mass density and temperature showed positive correlation to defoliation index. Tree species with high NMDS1, such as European oak (Quercus robur) and goat willow (Salix caprea), also positively contributed to high defoliation index. Canopies with penetration ratio at a certain level (PR_CANUND around 0.7) might prevent defoliation, while denser canopies (PR_CANUND above 0.8) seemed to be more likely to suffer from defoliation than canopies with less density (PR_CANUND below 0.6) according to the results.”

Supervisors: Dr. Soyeon Bae & Dr. Martin Wegmann

read more news:

Successful Master Thesis Defense by Konstantin Müller

Successful Master Thesis Defense by Konstantin Müller

On January 14th, Konstantin Müller successfully defended his master’s thesis titled "Animal Path Segmentation and Analysis via Generalized Deep Neural Network Regression". Supervised by Jakob Schwalb-Willmann and Dr. Mirjana Bevanda, the presentation was delivered to...

EAGLE Daria did her internship in Bergen

EAGLE Daria did her internship in Bergen

Our EAGLE student Daria recently wrapped up an internship at the University of Bergen in the Remote Sensing research group. With the support of her supervisor, Dr. Benjamin Abreu Robson, she got to work on the Jostedalsbreen glacier using drone and satellite data. Her...

EAGLE alumnus Henrik Fisser presenting polar research

EAGLE alumnus Henrik Fisser presenting polar research

Our EAGLE alumnus Henrik Fisser recently visited us after a research stay in the United States. He is now pursuing his PhD at UiT The Arctic University of Norway, specifically in the Earth Observation Department. UiT is renowned for its cutting-edge research in Earth...