Publications in Remote Sensing of Environment

Field-based monitoring of tree species in forests is often sparse due to logistical constraints. Remote sensing enables repeated, spatially contiguous collection of reflectance data across large areas. Tree species classification accuracy using such data is variable, likely because most studies use observational datasets where species occurrence correlates with environmental variation. We used two sites of a tree biodiversity experiment in Germany (BIOTREE: Kaltenborn and Bechstedt), where different species have been planted with high replication under controlled diversity levels, to assess how well tree species could be classified using reflectance data from airborne imaging spectroscopy and different classification methods (linear discriminant analysis, LDA, and a non-linear support vector machine, SVM). Reflectance data for 589 wavelengths between 400–2400 nm were acquired at 1 m spatial resolution during peak growing season. Reflectance spectra showed large and significant variation between taxonomic classes, orders, and species, and weak, but still significant, interactions between classes or orders and diversity levels. Classification accuracy reached 100% in training datasets, 77%–83% for the four species in Kaltenborn prediction datasets, and 31%–49% for the 16 species in Bechstedt prediction datasets. LDA provided more accurate predictions than SVM, and using similarly-spaced original wavelengths with LDA was as efficient as using principal components derived from the original data. While airborne imaging spectroscopy effectively distinguished up to four tree species in our datasets, classification accuracy was lower when 16 species had to be distinguished. In these cases, the methodology may be more useful for functional diversity monitoring than for tree species classification.

Observing functional diversity continuously in time and space using satellite imagery forms the basis for studying impact, interactions, and feedback of environmental change mechanisms on ecosystems and biodiversity globally. Functional diversity of plant traits links ecosystem functioning and biodiversity. This work presents an approach to map and quantify functional diversity of physiological forest traits derived from 20 m Sentinel-2 data in a temperate forest ecosystem. We used two complementary data sources, namely high-resolution, as well as spatially resampled airborne imaging spectroscopy data and Sentinel-2 data, to ensure our methods support consistently mapping functional diversity from space. We retrieved three physiological traits related to forest health, stress, and potential productivity, namely chlorophyll, carotenoid, and water content, from airborne imaging spectroscopy and Sentinel-2 data using corresponding spectral indices as proxies. We analyzed changes in two functional diversity metrics, namely functional richness and divergence, at different spatial resolutions. Both functional diversity metrics depend on the size and number of pixels to derive functional diversity as a function of distance, leading to different interpretations. When mapping functional diversity using Sentinel-2 data, small-scale patterns <1.1 ha were no longer visible, implying a minimum calculation area with 60 m radius recommended for retrieval of functional diversity metrics. The spectrally convolved and spatially resampled airborne spectroscopy data and the native Sentinel-2 data were correlated with r = 0.747 for functional richness and r = 0.709 for divergence in a 3.1 ha neighborhood. Functional richness was more affected by the differences in trait maps between the acquisitions resulting from effects in illumination and topography compared with functional divergence. Further differences could be explained by varying illumination/observation effects and phenological status of the vegetation at acquisition. Our approach demonstrates the importance of spatial and spectral resolution when scaling diversity assessments from regional to continental scales.