Laven Naidoo

L-band Synthetic Aperture Radar imagery performs better than optical datasets at retrieving woody fractional cover in deciduous, dry savannahs

Abstract Woody canopy cover (CC) is the simplest two dimensional metric for assessing the presence of the woody component in savannahs, but detailed validated maps are not currently available in southern African savannahs. A number of international EO programs (including in savannah landscapes) advocate and use optical LandSAT imagery for regional to country-wide mapping of woody canopy cover. However, previous research has shown that L-band Synthetic Aperture Radar (SAR) provides good performance at retrieving woody canopy cover in southern African savannahs. This study’s objective was to evaluate, compare and use in combination L-band ALOS PALSAR and LandSAT-5 TM, in a Random Forest environment, to assess the benefits of using LandSAT compared to ALOS PALSAR. Additional objectives saw the testing of LandSAT-5 image seasonality, spectral vegetation indices and image textures for improved CC modelling. Results showed that LandSAT-5 imagery acquired in the summer and autumn seasons yielded the highest single season modelling accuracies (R2 between 0.47 and 0.65), depending on the year but the combination of multi-seasonal images yielded higher accuracies (R2 between 0.57 and 0.72). The derivation of spectral vegetation indices and image textures and their combinations with optical reflectance bands provided minimal improvement with no optical-only result exceeding the winter SAR L-band backscatter alone results (R2 of ∼0.8). The integration of seasonally appropriate LandSAT-5 image reflectance and L-band HH and HV backscatter data does provide a significant improvement for CC modelling at the higher end of the model performance (R2 between 0.83 and 0.88), but we conclude that L-band only based CC modelling be recommended for South African regions.

The assessment of data mining algorithms for modelling Savannah Woody cover using multi-frequency (X-, C- and L-band) synthetic aperture radar (SAR) datasets

The woody component in African Savannahs provides essential ecosystem services such as fuel wood and construction timber to large populations of rural communities. Woody canopy cover (i.e. the percentage area occupied by woody canopy or CC) is a key parameter of the woody component. Synthetic Aperture Radar (SAR) is effective at assessing the woody component, because of its capacity to image within-canopy properties of the vegetation while offering an all-weather capacity to map relatively large extents of the woody component. This study compared the modelling accuracies of woody canopy cover (CC), in South African Savannahs, through the assessment of a set of modelling approaches (Linear Regression, Support Vector Machines, REPTree decision tree, Artificial Neural Network and Random Forest) with the use of X-band (TerraSAR-X), C-band (RADARSAT-2) and L-band (ALOS PALSAR) datasets. This study illustrated that the ANN, REPTree and RF non-parametric modelling algorithms were the most ideal with high CC prediction accuracies throughout the different scenarios. Results also illustrated that the acquisition of L-band data be prioritized due to the high accuracies achieved by the L-band dataset alone in comparison to the individual shorter wavelengths. The study provides promising results for developing regional savannah woody cover maps using limited LiDAR training data and SAR images.

Spectral variability within species and its effects on Savanna tree species discrimination

Differences in within-species phenology and structure driven by factors including topography, edaphic properties, and climatic variables present important challenges for species differentiation with remote sensing in the Kruger National Park, South Africa. The objective of this study was to examine probable factors including intraspecies spectral variability and the spectral sample size that could affect remote sensing of Savanna tree species across a land-use gradient in the Kruger National park. Eighteen species were examined: Acacia gerradii, Acacia nigrescens, Combretum apiculatum, Combretum collinum, Combretum hereroense, Combretum imberbe, Combretum zeyheri, Dichrostachys cinerea, Euclea sp (E. divinurum and E. natalensis, Gymnosporia sp (G. buxifolia and G. senegalensis), Lonchocarpus capassa, Peltoforum africanum, Piliostigma thonningii, Pterocarpus rotundifolia, Sclerocarya birrea, Strychnos sp (S. madagascariensis, S. usambarensis), Terminalia sericea and Ziziphus mucronata. Discriminating species using the K-nearest neighbour (K = 1) classifier with spectral angle mapper (SAM) yielded a higher classification accuracy (48% overall accuracy) compared to 16% for the classification involving the mean spectra for each species as the training spectral set. Within-species spectral variability and the training sample size were identified as important factors affecting classification accuracy of the tree species. We recommend a non-parametric classifier such as K-nearest neighbour classifier for classifying and mapping tree species in a highly complex environment such as the savanna system of the Kruger National Park.

Hyper-Temporal C-Band SAR for Baseline Woody Structural Assessments in Deciduous Savannas

Savanna ecosystems and their woody vegetation provide valuable resources and ecosystem services. Locally calibrated and cost effective estimates of these resources are required in order to satisfy commitments to monitor and manage change within them. Baseline maps of woody resources are important for analyzing change over time. Freely available, and highly repetitive, C-band data has the potential to be a viable alternative to high-resolution commercial SAR imagery (e.g., RADARSAT-2, ALOS2) in generating large-scale woody resources maps. Using airborne LiDAR as calibration, we investigated the relationships between hyper-temporal C-band ASAR data and woody structural parameters, namely total canopy cover (TCC) and total canopy volume (TCV), in a deciduous savanna environment. Results showed that: the temporal filter reduced image variance; the random forest model out-performed the linear model; while the TCV metric consistently showed marginally higher accuracies than the TCC metric. Combinations of between 6 and 10 images could produce results comparable to high resolution commercial (C- & L-band) SAR imagery. The approach showed promise for producing a regional scale, locally calibrated, baseline maps for the management of deciduous savanna resources, and lay a foundation for monitoring using time series of data from newer C-band SAR sensors (e.g., Sentinel1).

Reducing Leaf-Level Hyperspectral Data to 22 Components of Biochemical and Biophysical Bands Optimizes Tree Species Discrimination

The high dimensionality of hyperspectral data constitutes a challenge for species classification. This study assessed 1) whether tree species classification can be optimized with the selection of bands which relate to known plant properties and 2) whether a partial least square (PLS) transformation improve species classification above principal component analysis (PCA). Leaf spectra between 400 and 2500 nm were measured for six evergreen tree species in the spring of 2011, in the KwaZulu-Natal Province of South Africa. Twenty-two bands which relate to pigment, foliage biomass, nutrients, and leaf structural components were selected from the hyperspectral data set. The 2100 bands of 1 nm were resampled to 421 bands at 5 nm spectral resolution, ensuring the number of variables are less than the number of samples. The random forest (RF) classification algorithm was used to assess the accuracy for both PCA and PLS transformations on the 421 and 22 bands. The accuracy of individual species classes was calculated as the average of ten iterations, for each data reduction option. The three 22-band models resulted in comparable accuracies to the 421-band classifications (OA of 84 ± 4.9% for untransformed, 78 ± 5% for PCA, and 84 ± 4% for PLS) and no significant differences between the 421 and 22-band models (p > 0.4). The optimized PLS model (22 bands, 8 components) showed a 6% (p <; 0.01) increase in accuracy above the optimized PCA model (22 bands, 3 components). Reducing hyperspectral data to bands which relate to plant properties, and the use of PLS for data transformation, optimizes species classification.

Woody cover assessments in a Southern African savanna, using hyper-temporal C-band ASAR-WS data

Southern African savanna ecosystems and their woody resources are under pressure. Governments in the region need locally calibrated, cost effective, and regularly updated information on these resources in order to satisfy both national and international commitments to manage them. Using LiDAR data as a calibration dataset, this paper sets out to investigate the potential of hypertemporal C-band ASAR SAR data in mapping woody structural related parameters in a savanna environment. Images spanning three years where grouped by years (2007-2009), season (Wet or Dry) and polarization (HH or VV), and relationships were sought for the woody parameter total canopy cover (TCC). Results show that: Dry season combinations of images outperformed wet season images; HH co-polarised images outperformed VV images; temporally filtered images showed marked improvement on unfiltered images. While non-parametric random forest models achieved better validation accuracies than other models did. The single best result was achieved by combining all the temporally filtered images, from all of the various scenarios (R2=0.74; RMSE=8.52; SEP=35.27). The results show promise in delivering regional scale, locally calibrated, baseline products for the management of Southern Africas woody resources.

Use of Radarsat-2 polarimetric SAR images for fuel moisture mapping in the Kruger National Park, South Africa

Fully polarimetric Radarsat-2 imagery from wet and dry conditions over the South African Lowveld is compared to assess its value for fuel moisture mapping. Imagery was acquired at two different dates, in May (end of summer, wet) and in August (mid of winter, dry). Sample plots were classified into two broad Lowveld site types (herbaceous-dominated and shrub and tree-dominated). Linear and circular polarized backscatters, polarimetric discriminators and polarimetric decomposition parameters were computed to find suitable parameters for fuel moisture estimation. The results show a significant distinction between wet and dry conditions for C-HH, C-HV, C-RR, and C-LL, all Freeman-Durden and van Zyl decomposition parameters and some polarimetric discriminators (dmin, Prmax, Prmin, Smax, Smin). In almost all cases the normalized difference between wet and dry condition is lower for the shrub and tree-dominated sites. The Freeman-Durden double bounce scattering decomposition parameter performs best in both site types.