Funded by BMBF

Project Leader at TUM
Prof. Dr.-Ing. Xiaoxiang Zhu

Project Scientists at TUM
Veronika Buchauer

Cooperation Partners
Helmholtz Munich

Rifcon GmbH

Munich Science Communication Lab (LMU)

Runtime
2024 – 2027

The GEBIKI project develops a comprehensive, non-invasive approach for biodiversity monitoring by combining genomic sequencing from the Helmholtz Center with advanced Earth Observation (EO) and AI methods at TUM. The goal is to track biodiversity in urban environments across Germany over multiple years, combining molecular data with remote sensing data to analyze biodiversity changes and anthropogenic influences. TUM’s contribution focuses on the development of a time series, multimodal self-supervised learning (SSL) model to detect and analyze anthropogenic impact. This includes creating large-scale EO datacubes from Sentinel-1, -2, and -5p missions, and designing a foundation model tailored to biodiversity-relevant downstream tasks such as Local Climate Zone (LCZ) classification. The resulting EO-based anthropogenic impact will be validated against genomic data, enabling robust, scalable, and interpretable biodiversity monitoring for urban planning.