Increasing woody cover and overgrazing in semi-arid ecosystems are known to be the major factors driving land degradation. This study focuses on mapping the distribution of the slangbos shrub (Seriphium plumosum) in a test region in the Free State Province of South Africa. The goal of this study is to monitor the slangbos encroachment on cultivated land by synergistically combining Synthetic Aperture Radar (SAR) (Sentinel-1) and optical (Sentinel-2) Earth observation information. Both optical and radar satellite data are sensitive to different vegetation properties and surface scattering or reflection mechanisms caused by the specific sensor characteristics. We used a supervised random forest classification to predict slangbos encroachment for each individual crop year between 2015 and 2020. Training data were derived based on expert knowledge and in situ information from the Department of Agriculture, Land Reform and Rural Development (DALRRD). We found that the Sentinel-1 VH (cross-polarization) and Sentinel-2 SAVI (Soil Adjusted Vegetation Index) time series information have the highest importance for the random forest classifier among all input parameters. The modelling results confirm the in situ observations that pastures are most affected by slangbos encroachment. The estimation of the model accuracy was accomplished via spatial cross-validation (SpCV) and resulted in a classification precision of around 80% for the slangbos class within each time step.

Understanding forest decline under drought pressure is receiving research attention due to the increasing frequency of large-scale heat waves and massive tree mortality events. However, since assessing mortality on the ground is challenging and costly, this study explores the capability of satellite-borne Copernicus Sentinel-1 (S-1) C-band radar data for monitoring drought-induced tree canopy damage. As droughts cause water deficits in trees and eventually lead to early foliage loss, the S-1 radiometric signal and polarimetric indices are tested regarding their sensitivities to these effects, exemplified in a deciduous broadleaf forest. Due to the scattered nature of mortality in the study site, we employed a temporal-only time series filtering scheme that provides very high spatial resolution (10 m ×10 m) for measuring at the scale of single trees. Finally, the anomaly between heavily damaged and non-damaged tree canopy samples (n = 146 per class) was used to quantify the level of damage. With a maximum anomaly of −0.50 dB ± 1.38 for S-1 Span (VV+VH), a significant decline in hydrostructural scattering (moisture and geometry of scatterers as seen by SAR) was found in the second year after drought onset. By contrast, S-1 polarimetric indices (cross-ratio, RVI, Hα) showed limited capability in detecting drought effects. From our time series evaluation, we infer that damaged canopies exhibit both lower leaf-on and leaf-off backscatters compared to unaffected canopies. We further introduce an NDVI/Span hysteresis showing a lagged signal anomaly of Span behind NDVI (by ca. one year). This time-lagged correlation implies that SAR is able to add complementary information to optical remote sensing data for detecting drought damage due to its sensitivity to physiological and hydraulic tree canopy damage. Our study lays out the promising potential of SAR remote sensing information for drought impact assessment in deciduous broadleaf forests.

<p>Describing forest structure is fundamental to understanding forest ecology and calculating biomass estimations. To enable its characterization with large spatial coverage, we investigate data recorded by airborne LiDAR and three different radar frequencies over a deciduous broadleaf forest at the Hainich National Park in central Germany. This study aims at distilling the microwave frequencies and polarisations that most closely relate to structural metrics extracted from the LiDAR point clouds, and are therefore most promising for extending spatial or temporal coverage.</p> <p>The LiDAR point clouds, which are provided openly by the Thuringian State Office for Land Management and Geoinformation, were processed to five structural metrics at 25 m x 25 m pixel size. These metrics comprise an estimation of fractional cover based on vegetation return numbers,  an intensity-based fractional cover approach (Hopkinson & Chasmer 2009), the skewness and standard deviation of the height distribution, as well as the the vertical complexity index as defined by van Ewijk (2011). These metrics were compared to terrain-corrected backscatter of phenologically matching scenes from three different sensor frequencies: an X Band scene from DLR TerraSAR-X, C Band from Copernicus Sentinel-1, and L Band from JAXA ALOS-2. </p> <p>The scenes represent leaf-off conditions. To reduce misleading factors, the analysis was limited to areas with moderate slope angles below 10 degrees. Subsequently, regression models between the lidar metrics and backscatter intensities were built.<br />First results from bivariate correlations indicate the best match between ALOS-2 HV and fractional cover (r²=0.41) as well as standard deviation (r²= 0.43). Among the metrics, fractional cover is associated most closely with backscatter in all frequencies: the highest correlation coefficients amount to 0.37 for X Band (VV), 0.22 for C Band (VH), and 0.41 for L Band (HV), respectively. In general, C Band exhibits the lowest pairwise correlations with most density metrics, compared to L- and X Band. <br />The poster will show the results of multivariate regression models and discuss which combination of frequencies and polarizations is best suited for the derivation of specific forest structure parameters at larger scales.</p> <p>---</p> <p>Hopkinson, C., & Chasmer, L. (2009). Testing LiDAR models of fractional cover across multiple forest ecozones. Remote Sensing of Environment, 113(1), 275–288. DOI:10.1016/j.rse.2008.09.012</p> <p>van Ewijk, K. Y., Treitz, P. M., & Scott, N. A. (2011). Characterizing Forest Succession in Central Ontario using Lidar-derived Indices. Photogrammetric Engineering & Remote Sensing, 77(3), 261–269. DOI: 10.14358/PERS.77.3.261</p>

<p>Increasing woody cover and overgrazing in semi-arid ecosystems are known to be major factors driving land degradation. During the last decades woody cover encroachment has increased over large areas in southern Africa inducing environmental, land cover as well as land use changes. </p><p>The goal of this study is to synergistically combine SAR (Sentinel-1) and optical (Sentinel-2) earth observation information to monitor the slangbos encroachment on arable land in the Free State province, South Africa, between 2015 and 2020. Both, optical and radar satellite data are sensitive to different land surface and vegetation properties caused by sensor specific scattering or reflection mechanisms they rely on. </p><p>This study focuses on mapping the slangbos aka bankrupt bush (Seriphium plumosum) encroachment in a selected test region in the Free State province of South Africa. Though being indigenous to South Africa, the slangbos has been documented to be the main encroacher on the grassvelds (South African grassland biomes) and thrive in poorly maintained cultivated lands. The shrub reaches a height and diameter of up to 0.6 m and the root system reaches a depth of up to 1.8 m. Slangbos has small light green leaves unpalatable to grazers due to their high oil content and is better adapted to long dry periods compared to grass communities.</p><p>We used the random forest approach to predict slangbos encroachment for each individual crop year between 2015 and 2020. Training data were based on expert knowledge and field information from the Department of Agriculture, Forestry and Fisheries (DAFF). Several input variables have been tested according to their model performance, e.g. backscatter, backscatter ratio, interferometric coherence as well as optical indices (e.g. NDVI (Normalized Difference Vegetation Index), SAVI (Soil Adjusted Vegetation Index), EVI (Enhanced Vegetation Index), etc.). We found that the Sentinel-1 VH backscatter (vertical–horizontal/cross-polarization) and the Sentinel-2 SAVI time series information have the highest importance for the random forest classifier among all input parameters. The estimation of the model accuracy was accomplished via spatial-cross validation and resulted in an overall accuracy of above 80 % for each time step, with the slangbos class being close to or above 90 %. </p><p>Currently we are developing a prototype application to be tested in cooperation with local stakeholders to bring this approach to the farmers level. Once field work in southern Africa is possible again, further ground truthing and interaction with farmers will be carried out.</p>