Am Großen Bruch, HySpex VNIR ATCOR, 2018-05-07, 2019-04-23 & TERENO boundaries GPKG

3 different data sets:

(1) Hyperspectral data (atmosphere corrected with ATCOR4) of the TERENO site "Am Großen Bruch" (DE-GsB) acquired on 2018-05-07 and 2019-04-23. Only contains data from the HySpex VNIR sensor (400-1000nm, Pixelsize XY 0.16m x 0.32m, Frameperiod 18000µs, Integrationtime 10000).

(2) Vector data with bounding boxes of the Bode TERENO sites as a geopackage.

(3) Evaluation data set (table) containing subsampled variables derived from the hyperspectral data (vegetation indices), a digital elevation model (DEM) and geophysical variables (electromagnetic induction & gamma ray data) for the research site DE-GsB.

Method:
The data can be used to reproduce results from a vegetation stress detection analysis. Simplex volume maximisation (SiVM), an unsupervised learning method, was used to detect vegetation stress patterns from the high-dimensional hyperspectral data. The resulting classification was evaluated using established vegetation indices, a DEM and different geophysical measurements.

Results:
Statistical evaluation of the resulting stress metric showed a strong correspondence of stress-related vegetation indices and the SiVM-based stress metric. DEM and geophysical variables were irrelevant, in comparison.

Algorithms used for data preprocessing, analysis and visualization can be found at https://git.ufz.de/hermanns/rs_sivm or https://github.com/flossi/rs_sivm.