The Cape Floristic Region (CFR) boasts rich biodiversity but faces threats from invasive species and land-use changes. Fynbos wetland vegetation within the CFR is under-mapped despite its crucial role in supporting biodiversity and maintaining hydrological cycles. This study assessed the potential of UAV VIS-NIR data, gathered during Spring and Summer, to identify the spectral characteristics of eleven Fynbos wetland species in a seep wetland. Spectral distances derived from reflectance data revealed distinct spectral clustering of plant species, highlighting which species could be distinguished from each other. UAV data also captured differences in reflectance across spectral bands for both dates. Spectral statistics indicated that certain species could be more accurately classified in Spring than in Summer, and vice versa. These findings underscore the efficacy of UAV multispectral data in analyzing the reflectance patterns of fynbos wetland species. Additionally, the sensitivity of UAV multispectral data to foliar pigment composition across different seasonal stages was confirmed. Lastly, species classification results demonstrated that a random forest classifier is well suited, with relative producer and user accuracies aligning with the derived spectral distances. The results highlight the potential of UAV imagery for monitoring these endemic species and creating opportunities for scalable mapping of Fynbos seep wetlands.

Despite growing international interest in seasonal effects on wetland vegetation mapping, there is a notable lack of research focused on South Africa’s unique fynbos wetlands, leaving a critical gap in understanding the spatiotemporal dynamics of fynbos ecosystems. This study aimed to assess the ability of Parrot Sequoia and MicaSense RedEdge-M UAV data collected during six seasonal periods between 2018 and 2020 to discriminate between fynbos wetland vegetation species. It also identifies the most suitable time of year for accurate species-level classification. The highest classification accuracy (OA = 98.0%) was achieved in late winter and early summer (OA = 90.1%), while the lowest (OA = 57.2%) occurred in mid-autumn. Most species attained high user and producer accuracies, though Erica serrata and Tetraria thermalis were more inconsistently classified. A Kruskal–Wallis test revealed a significant effect of seasonality on user and producer accuracy as well as kappa (p < 0.05). A Wilcoxon rank-sum test indicated that the accuracy metrics were not significantly different (p > 0.05) when different sensors were used within the same season. The results suggest that conservation agencies and researchers should collect remote sensing data at the end of winter to take advantage of phenological differences between plant species.

The correction of digital elevation models (DEMs) can be achieved using a variety of techniques. Machine learning and statistical methods are broadly applicable to a variety of DEM correction case studies in different landscapes. However, a literature survey did not reveal any research that compared the effectiveness or performance of both methods. In this study, we comparatively evaluate three gradient boosted decision trees (XGBoost, LightGBM and CatBoost) and multiple linear regression for the correction of two publicly available global DEMs: Copernicus GLO-30 and ALOS World 3D (AW3D) in Cape Town, South Africa. The training datasets are comprised of eleven predictor variables including elevation, slope, aspect, surface roughness, topographic position index, terrain ruggedness index, terrain surface texture, vector ruggedness measure, percentage bare ground, urban footprints and percentage forest cover as an indicator of the overland forest distribution. The target variable (elevation error) was derived with respect to highly accurate airborne LiDAR. The results presented in this study represent urban/industrial and grassland/shrubland/dense bush landscapes. Although the accuracy of the original DEMs had been degraded by several anomalies, the corrections improved the vertical accuracy across vast areas of the landscape. In the urban/industrial and grassland/shrubland landscapes, the reduction in the root mean square error (RMSE) of the original AW3D DEM was greater than 70%, after correction. The corrections improved the accuracy of Copernicus DEM, e.g., > 44% RMSE reduction in the urban area and >32% RMSE reduction in the grassland/shrubland landscape. Generally, the gradient boosted decision trees outperformed multiple linear regression in most of the tests.

The accuracy of digital elevation models (DEMs) in urban areas is influenced by numerous factors including land cover and terrain irregularities. Moreover, building artifacts in global DEMs cause artificial blocking of surface flow pathways. This compromises their quality and adequacy for hydrological and environmental modelling in urban landscapes where precise and accurate terrain information is needed. In this study, the extreme gradient boosting (XGBoost) ensemble algorithm is adopted for enhancing the accuracy of two medium-resolution 30m DEMs over Cape Town, South Africa: Copernicus GLO-30 and ALOS World 3D (AW3D). XGBoost is a scalable, portable and versatile gradient boosting library that can solve many environmental modelling problems. The training datasets are comprised of eleven predictor variables including elevation, urban footprints, slope, aspect, surface roughness, topographic position index, terrain ruggedness index, terrain surface texture, vector roughness measure, forest cover and bare ground cover. The target variable (elevation error) was calculated with respect to highly accurate airborne LiDAR. After training and testing, the model was applied for correcting the DEMs at two implementation sites. The correction achieved significant accuracy gains which are competitive with other proposed methods. The root mean square error (RMSE) of Copernicus DEM improved by 46 to 53% while the RMSE of AW3D DEM improved by 72 to 73%. These results showcase the potential of gradient boosted trees for enhancing the quality of DEMs, and for improved hydrological modelling in urban catchments.

Validation studies of global Digital Elevation Models (DEMs) in the existing literature are limited by the diversity and spread of landscapes, terrain types considered and sparseness of groundtruth. Moreover, there are knowledge gaps on the accuracy variations in rugged and complex landscapes, and previous studies have often not relied on robust internal and external validation measures. Thus, there is still only partial understanding and limited perspective of the reliability and adequacy of global DEMs for several applications. In this study, we utilize a dense spread of LiDAR groundtruth to assess the vertical accuracies of four medium-resolution, readily available, free-access and global coverage 1 arc-second (30 m) DEMs: NASADEM, ASTER GDEM, Copernicus GLO-30, and ALOS World 3D (AW3D). The assessment is carried out at landscapes spread across Cape Town, Southern Africa (urban/industrial, agricultural, mountain, peninsula and grassland/shrubland) and forested national parks in Gabon, Central Africa (low-relief tropical rainforest and high-relief tropical rainforest). The statistical analysis is based on robust accuracy metrics that cater for normal and non-normal elevation error distribution, and error ranking. In Cape Town, Copernicus DEM generally had the least vertical error with an overall Mean Error (ME) of 0.82 m and Root Mean Square Error (RMSE) of 2.34 m while ASTER DEM had the poorest performance. However, ASTER GDEM and NASADEM performed better in the low-relief and high-relief tropical forests of Gabon. Generally, the DEM errors have a moderate to high positive correlation in forests, and a low to moderate positive correlation in mountains and urban areas. Copernicus DEM showed superior vertical accuracy in forests with less than 40% tree cover, while ASTER and NASADEM performed better in denser forests with tree cover greater than 70%. This study is a robust regional assessment of these global DEMs.

Gradient-Boosted Decision Trees (GBDTs), particularly when tuned with Bayesian optimisation, are powerful machine learning techniques known for their effectiveness in handling complex, non-linear data. However, the performance of these models can be significantly influenced by the characteristics of the terrain being analysed. In this study, we assess the performance of three Bayesian-optimised GBDTs (XGBoost, LightGBM and CatBoost) using digital elevation model (DEM) error correction as a case study. The performance of the models is investigated across five landscapes in Cape Town South Africa: urban/industrial, agricultural, mountain, peninsula and grassland/shrubland. The models were trained using a selection of datasets (elevation, terrain parameters and land cover). The comparison entailed an analysis of the model execution times, regression error metrics, and level of improvement in the corrected DEMs. Generally, the optimised models performed considerably well and demonstrated excellent predictive capability. CatBoost emerged with the best results in the level of improvement recorded in the corrected DEMs, while LightGBM was the fastest of all models in the execution time for Bayesian optimisation and model training. These findings offer valuable insights for applying machine learning and hyperparameter tuning in remote sensing.

There has been a rapid evolution of tree-based ensemble algorithms which have outperformed deep learning in several studies, thus emerging as a competitive solution for many applications. In this study, ten tree-based ensemble algorithms (random forest, bagging meta-estimator, adaptive boosting (AdaBoost), gradient boosting machine (GBM), extreme gradient boosting (XGBoost), light gradient boosting (LightGBM), histogram-based GBM, categorical boosting (CatBoost), natural gradient boosting (NGBoost), and the regularised greedy forest (RGF)) were comparatively evaluated for the enhancement of Copernicus digital elevation model (DEM) in an agricultural landscape. The enhancement methodology combines elevation and terrain parameters alignment, with feature-level fusion into a DEM enhancement workflow. The training dataset is comprised of eight DEM-derived predictor variables, and the target variable (elevation error). In terms of root mean square error (RMSE) reduction, the best enhancements were achieved by GBM, random forest and the regularised greedy forest at the first, second and third implementation sites respectively. The computational time for training LightGBM was nearly five-hundred times faster than NGBoost, and the speed of LightGBM was closely matched by the histogram-based GBM. Our results provide a knowledge base for other researchers to focus their optimisation strategies on the most promising algorithms.

Several methods have been proposed for correcting the elevation bias in digital elevation models (DEMs) for example, linear regression. Nowadays, supervised machine learning enables the modelling of complex relationships between variables, and has been deployed by researchers in a variety of fields. In the existing literature, several studies have adopted either machine learning or statistical approaches in the task of DEM correction. However, to our knowledge, none of these studies have compared the performance of both approaches, especially with regard to open-access global DEMs. Our previous work has already shown the potential of machine learning approaches, specifically gradient boosted decision trees (GBDTs) for DEM correction. In this study, we share some results from the comparison of three recent implementations of gradient boosted decision trees (XGBoost, LightGBM and CatBoost), versus multiple linear regression (MLR) for enhancing the vertical accuracy of 30 m Copernicus and AW3D global DEMs in Cape Town, South Africa.

Wetlands harbour a wide range of vital ecosystems. Hence, mapping wetlands is essential to conserving the ecosystems that depend on them. However, the physical nature of wetlands makes fieldwork difficult and potentially erroneous. This study used multispectral UAV aerial photography to map ten wetland plant species in the Fynbos Biome in the Steenbras Nature Reserve. We developed a methodology that used K-Nearest Neighbour (KNN), Support Vector Machine (SVM), and Random Forest (RF) machine learning algorithms to classify ten wetland plant species using the preselected bands and spectral indices. The study identified Normalized green red difference index (NGRDI), Red Green (RG) index, Green, Log Red Edge (LogRE), Normalized Difference Red-Edge (NDRE), Chlorophyll Index Red-Edge (CIRE), Green Ratio Vegetation Index (GRVI), Normalized Difference Water Index (NDWI), Green Normalized Difference Vegetation Index ( GNDVI) and Red as pertinent bands and indices for classifying wetland plant species in the Proteaceae, Iridaceae, Restionaceae, Ericaceae, Asteraceae and Cyperaceae families. The classification had an overall accuracy of 87.4% and kappa accuracy of 0.85. Thus, the findings are pertinent to understanding the spectral characteristics of these endemic species. The study demonstrates the potential for UAV-based remote sensing of these endemic species.

We demonstrate the application of geomatics tools (remote sensing and geographic information systems) for spatial data analysis to determine potential locations for wind and solar photovoltaic (PV) energy plants in the Central North region of Namibia. In accordance with sustainable development goal 7 (affordable and clean energy) and goal 13 (climate action), the Namibian government has committed to reducing reliance on fossil fuels. In support of this, suitable locations for renewable energy plants need to be identified. Using multi-criteria decision-making and the analytical hierarchy process, sites were selected considering topographical, economic, climatic, and environmental factors. It was found that the highest potential for solar PV energy plants is in the northwest, southwest, and southern regions of the study area, whereas only the northwest region is highly suitable for wind power plants. These results were substantiated by comparison with global suitability maps, with some differences due to the datasets used. The findings can be used as a guide by governments, commercial investors, and other stakeholders to determine prospective sites for the development of renewable energy in Central North Namibia.

Gradient boosted decision trees (GBDTs) have repeatedly outperformed several machine learning and deep learning algorithms in competitive data science. However, the explainability of GBDT predictions especially with earth observation data is still an open issue requiring more focus by researchers. In this study, we investigate the explainability of Bayesian-optimised GBDT algorithms for modelling and prediction of the vertical error in Copernicus GLO-30 digital elevation model (DEM). Three GBDT algorithms are investigated (extreme gradient boosting - XGBoost, light boosting machine – LightGBM, and categorical boosting – CatBoost), and SHapley Additive exPlanations (SHAP) are adopted for the explainability analysis. The assessment sites are selected from urban/industrial and mountainous landscapes in Cape Town, South Africa. Training datasets are comprised of eleven predictor variables which are known influencers of elevation error: elevation, slope, aspect, surface roughness, topographic position index, terrain ruggedness index, terrain surface texture, vector roughness measure, forest cover, bare ground cover, and urban footprints. The target variable (elevation error) was calculated with respect to accurate airborne LiDAR. After model training and testing, the GBDTs were applied for predicting the elevation error at model implementation sites. The SHAP plots showed varying levels of emphasis on the parameters depending on the land cover and terrain. For example, in the urban area, the influence of vector ruggedness measure surpassed that of first-order derivatives such as slope and aspect. Thus, it is recommended that machine learning modelling procedures and workflows incorporate model explainability to ensure robust interpretation and understanding of model predictions by both technical and non-technical users.

The remote sensing community has identified data fusion as one of the key challenging topics of the 21st century. The subject of image fusion in two-dimensional (2D) space has been covered in several published reviews. However, the special case of 2.5D/3D Digital Elevation Model (DEM) fusion has not been addressed till date. DEM fusion is a key application of data fusion in remote sensing. It takes advantage of the complementary characteristics of multi-source DEMs to deliver a more complete, accurate, and reliable elevation dataset. Although several methods for fusing DEMs have been developed, the absence of a well-rounded review has limited their proliferation among researchers and end-users. Combining knowledge from multiple studies is often required to inform a holistic perspective and guide further research. In response, this paper provides a systematic review of DEM fusion: the pre-processing workflow, methods and applications, enhanced with a meta-analysis. Through the discussion and comparative analysis, unresolved challenges and open issues are identified, and future directions for research are proposed. This review is a timely solution and an invaluable source of information for researchers within the fields of remote sensing and spatial information science, and the data fusion community at large.

Shack extraction is a growing application area for solving contemporary geographical complexities in developing countries. Traditional ground-based surveys that provided shack counts are being overtaken by the opportunities around rapid advances in spatial data collection. The ability to automate and report high accuracies against ground truth remains key to the complex nature of the digital shack extraction matrix. This study uses 8 cm aerial footage and laser scanning data to map informal settlement areas within built environments. The research design includes a set of experiments in defining the footprint of shacks obtained through profiling sampled shack/non-shack areas. Average shack heights, behaviour of unclassified points, number of non-ground points, scan texture and intensity variations were investigated. A five-parameter signature was applied to a pit-free normalized digital surface model (NDSM) and as expected on both test and validation, proved suitable for distinct shacks identification. This approach reported a 95% large settlement detection within a 1-metre-wide boundary position tolerance, coupled with a few false positives in isolated cases of multi-storey, sheds and also complex shacks. The results are proposed for inclusion in city-wide urban planning frameworks that can also be tweaked to include social and environmental parameters.

(1) We monitored post-fire shrubland recovery responses to changes in rainfall seasonality using a multi-year field experiment in the Cape Floristic Region (CFR) of South Africa. A primary objective was to test the utility of UAVs for monitoring ultra-fine-scale vegetation changes in the early post-fire context. (2) By comparison with detailed ground-based measurements, we showed that UAVs improved detection of integrated community growth responses, given that the appropriate relative radiometric normalisation techniques were applied to repeated imagery data. UAVs supported ground-based findings and, moreover, helped to identify previously undetected growth form responses. However, due to the limitations in detecting species-specific demographic changes, UAVs could not completely replace ground-based measurements. (3) Our combined UAV-based and ground-based monitoring approaches indicated strong coupling between post-fire shrubland recovery and seasonal rainfall patterns in the CFR but also demonstrated that sensitivity to rainfall seasonality could differ between neighbouring shrubland communities occurring on different soil types. (4) The careful integration of UAV-based and ground-based monitoring approaches provided the fullest understanding of early post-fire shrubland recovery patterns.

Non-forest ecosystems, dominated by shrubs, grasses and herbaceous plants, provide ecosystem services including carbon sequestration and forage for grazing, and are highly sensitive to climatic changes. Yet these ecosystems are poorly represented in remotely sensed biomass products and are undersampled by in situ monitoring. Current global change threats emphasize the need for new tools to capture biomass change in non-forest ecosystems at appropriate scales. Here we developed and deployed a new protocol for photogrammetric height using unoccupied aerial vehicle (UAV) images to test its capability for delivering standardized measurements of biomass across a globally distributed field experiment. We assessed whether canopy height inferred from UAV photogrammetry allows the prediction of aboveground biomass (AGB) across low-stature plant species by conducting 38 photogrammetric surveys over 741 harvested plots to sample 50 species. We found mean canopy height was strongly predictive of AGB across species, with a median adjusted R-2 of 0.87 (ranging from 0.46 to 0.99) and median prediction error from leave-one-out cross-validation of 3.9%. Biomass per-unit-of-height was similar within but different among, plant functional types. We found that photogrammetric reconstructions of canopy height were sensitive to wind speed but not sun elevation during surveys. We demonstrated that our photogrammetric approach produced generalizable measurements across growth forms and environmental settings and yielded accuracies as good as those obtained from in situ approaches. We demonstrate that using a standardized approach for UAV photogrammetry can deliver accurate AGB estimates across a wide range of dynamic and heterogeneous ecosystems. Many academic and land management institutions have the technical capacity to deploy these approaches over extents of 1-10 ha(-1). Photogrammetric approaches could provide much-needed information required to calibrate and validate the vegetation models and satellite-derived biomass products that are essential to understand vulnerable and understudied non-forested ecosystems around the globe.

SUMMARY This study explores the geomorphological expression and geological context of a normal fault scarp in a stable continental region (SCR) which we interpret as having failed in large (Mw >7) earthquakes. Records of such large normal faulting events in an SCR (or even in more rapidly deforming regions) are extremely rare, and so understanding this feature is of international interest. The scarp is exceptionally well-preserved due to the extensive calcrete/silcrete cementation. In areas where this cementation is reduced or absent the scarp is more diffuse, as expected for a feature formed by one or more paleoearthquakes. The exceptional preservation aids comparison with data sets based on scarps which have formed more recently. Our analysis is based on a high-resolution digital elevation model of the Hebron Fault scarp in southern Namibia using pan-sharpened Worldview-3 satellite stereophotos (0.31 m resolution). We make scarp height measurements at 160 locations providing improved estimates of the average displacement (5.9 m), maximum displacement (10.1 m), and the minimum fault length (45 km). No consistent evidence of lateral displacements in water courses or alluvial fan margins were found implying predominantly normal displacement. A newly described section in the northwest has en-echelon scarps consistent with a component of strike-slip motion that may be explained by its difference in strike from the central section. Most channels crossing the fault show a single knick-point. The displacement varies smoothly as it crosses a number of different generations of alluvial fan surfaces. No evidence of a multiscarp or a composite scarp were observed. We have therefore found no evidence for a mutiple-event origin for the scarp, although, this lack of evidence does not conclusively demonstrate a single-event origin. Published regressions, based on the limited data available for SCRs, suggest that the mean expected average displacement ($\bar{D}_{\rm av}$) for a faults of this length is 1.2–3.1 m implying that the scarp is likely to have formed in 2–5 events with an expected Mw = ∼7.1 though displacements in individual events may exceed these average values. Comparison with the regional geology and aeromagnetic data sets suggests that the fault reactivates a Mesoproterozoic ductile structure, the Nam Shear Zone, and that the location, orientation and segmentation of the scarp is controlled by the alignment of pre-existing structurally weak zones with the present-day stress regime. The fault has undergone repeated brittle reactivation, accumulating ∼110 m of vertical offset since the deposition of the Ediacaran-to-Cambrian Nama Group. This is less than expected from global compilations of total displacement and fault length data, suggesting that the fault rapidly attained its current length by recruiting an existing weak zone and is expected to accumulate displacement at a relatively constant length in the future.

Monitoring fine-scale vegetation processes over wider spatial extents is often logistically challenging. We demonstrate a practical approach for monitoring high-resolution (i.e sub-meter) vegetation cover changes using UAV-mounted multispectral cameras. We use the post-fire shrublands of the Cape Floristic Region, South Africa as a case study. Repeated NDVI images of post-fire sites were generated biannually over three years using UAV-mounted multispectral cameras and commercial, image-processing software. We applied a procedure for identifying and correcting temporal radiometric noise in repeated NDVI images using pseudoinvariant features (PIFs). We then extracted vegetation cover data using NDVI thresholding. The quality of UAV-based vegetation cover data was assessed using detailed ground-truthing. Relative radiometric image normalization reduced radiometric noise in NDVI data and resulted in stronger correlations between UAV-based and ground-based area cover measurements over time. UAV-mounted multispectral cameras are effective high-resolution vegetation monitoring tools which could help ecologists to investigate fine-scale vegetation processes across wider spatial extents.

The implementation of Unmanned Aerial Vehicles (UAVs) and Structure-from-Motion (SfM) photogrammetry in assessing forest structures for forest inventory and biomass estimations has shown great promise in reducing costs and labour intensity while providing relative accuracy. Tree Height (TH) and Diameter at Breast Height (DBH) are two major variables in biomass assessment. UAV-based TH estimations depend on reliable Digital Terrain Models (DTMs), while UAV-based DBH estimations depend on reliable dense photogrammetric point cloud. The main aim of this study was to evaluate the performance of multi-rotor UAV photogrammetric point cloud in estimating homogeneous and heterogeneous forest structures, and their comparison to more accurate LiDAR data obtained from Aerial Laser Scanners (ALS), Terrestrial Laser Scanners (TLS), and more conventional means like manual field measurements. TH was assessed using UAV(SfM) and LiDAR point cloud derived DTMs, while DBH was assessed by comparing UAV(SfM )photogrammetric point cloud to LiDAR point cloud, as well as to manual measurements. The results obtained in the study indicated that there was a high correlation between UAV(SfM) TH and ALS(LiDAR) TH (R-2 = 0.9258) for homogeneous forest structures, while a lower correlation between UAV(SfM) TH and TLSLiDAR TH (R-2 = 0.8614) and UAV(SfM) TH and ALS(LiDAR) TH (R-2 = 0.8850) was achieved for heterogeneous forest structures. A moderate correlation was obtained between UAV(SfM) DBH and field measurements (R-2 = 0.5955) for homogenous forest structures, as well as between UAV(SfM) DBH and TLSLiDAR DBH (R-2 = 0.5237), but a low correlation between UAV(SfM) DBH and UAV(LiDAR) DBH (R-2 = 0.1114). The study demonstrated that UAV acquired imagery can be used to accurately estimate TH in both forest types, but has challenges estimating DBH. The research does not suggest that UAV(SfM) serves as a replacement for more high-cost and accurate LiDAR data, but rather as a cheaper adequate alternative in forestry management depending on accuracy requirements.