Database launched

deadtrees.earth – towards revealing global tree mortality dynamics.

deadtrees.earth is an initiative addressing the global issue of excessive tree mortality. We harness crowd-sourced drone data, Earth observation satellites and machine learning to map and analyze tree mortality. Our initiative overcomes the lack of comprehensive georeferenced data by developing a dynamic database where users can upload and analyze aerial imagery to detect deadwood. The database was now described in detail by Clemens Mosig and more than 140 co-authors.

Publication in Nature Communications Earth & Environment

Temporal dynamics in vertical leaf angles can confound vegetation indices widely used in Earth observations

Since the first NASA Earth observation missions, biophysical vegetation properties such as biomass and chlorophyll content have been monitored using satellites – however, often with significant uncertainties. Our research shows that part of these uncertainties is caused by leaves moving over time.

Seminar

Online-Seminar of deadtrees.earth hosted by the International Tree Mortality Network (ITMN)

Teja Kattenborn and Clemens Mosing gave an overview of the deadtrees.earth platform. The seminar will in about 60 minutes give and overview on the central idea, the current state and future plans of deadtrees.earth. We were very honoured that we could represent deadtrees.earth at the ITMN seminar series, which is an international network of leading scientistis in the field of tree mortality monitoring.

New partnership

geosense is partner of NFDI4Earth, the national research data infrastructure for the Earth System Sciences.

geosense will support NFDI4Earth in the context of data product generation, standardization and analytical tools. This partnership will be kicked-off with the pilot project GeoLabel, which will deliver a web-based labelling tool for geospatial data. GeoLabel will particularly enable collaborative and AI-supported labeling of various phenomena.

New DFG funded research project

LeafH2O – A UAV-mounted dual-wavelength LiDAR for leaf water content retrieval

In November, the new research project LeafH2O will kick off! LeafH2O is a collaboration project of the Department for Sensor-based Geoinformatics and INATECH and will be supported by Fraunhofer IPM. The aim of LeafH2O is to develop a drone-based dual-wavelength LiDAR tailored to retrieve leaf water content in 3D of entire plant canopies. This technology will eventually enable to track leaf water content for a range of applications, such as precision agriculture or monitoring drought stress in ecosystems. We are very much looking forward to working in this fantastic consortium!

Publication in Biogeosciences

From simple labels to semantic image segmentation: leveraging citizen science plant photographs for tree species mapping in drone imagery

Salim Soltani et al. demonstrate how crowd-sourced plant photographs from smartphones can be used with deep learning to map plant species in drone images. The presented framework could ultimately provide a framework for biodiveristy monitoring, where thousands of species could be mapped without having actual reference data for model training.