In recent years, remote sensing researchers have investigated the use of different modalities (or combinations of modalities) for classification tasks. Such modalities can be extracted via a diverse range of sensors and images. Currently, there are no (or only a few) studies that have been done to increase the land cover classification accuracy via unmanned aerial vehicle (UAV)–digital surface model (DSM) fused datasets. Therefore, this study looks at improving the accuracy of these datasets by exploiting convolutional neural networks (CNNs). In this work, we focus on the fusion of DSM and UAV images for land use/land cover mapping via classification into seven classes: bare land, buildings, dense vegetation/trees, grassland, paved roads, shadows, and water bodies. Specifically, we investigated the effectiveness of the two datasets with the aim of inspecting whether the fused DSM yields remarkable outcomes for land cover classification. The datasets were: (i) only orthomosaic image data (Red, Green and Blue channel data), and (ii) a fusion of the orthomosaic image and DSM data, where the final classification was performed using a CNN. CNN, as a classification method, is promising due to hierarchical learning structure, regulating and weight sharing with respect to training data, generalization, optimization and parameters reduction, automatic feature extraction and robust discrimination ability with high performance. The experimental results show that a CNN trained on the fused dataset obtains better results with Kappa index of ~0.98, an average accuracy of 0.97 and final overall accuracy of 0.98. Comparing accuracies between the CNN with DSM result and the CNN without DSM result for the overall accuracy, average accuracy and Kappa index revealed an improvement of 1.2%, 1.8% and 1.5%, respectively. Accordingly, adding the heights of features such as buildings and trees improved the differentiation between vegetation specifically where plants were dense.

Predicting landslide occurrences can be difficult. However, failure to do so can be catastrophic, causing unwanted tragedies such as property damage, community displacement, and human casualties. Research into landslide susceptibility mapping (LSM) attempts to alleviate such catastrophes through the identification of landslide prone areas. Computational modelling techniques have been successful in related disaster scenarios, which motivate this work to explore such modelling for LSM. In this research, the potential of supervised machine learning and ensemble learning is investigated. Firstly, the Flexible Discriminant Analysis (FDA) supervised learning algorithm is trained for LSM and compared against other algorithms that have been widely used for the same purpose, namely Generalized Logistic Models (GLM), Boosted Regression Trees (BRT or GBM), and Random Forest (RF). Next, an ensemble model consisting of all four algorithms is implemented to examine possible performance improvements. The dataset used to train and test all the algorithms consists of a landslide inventory map of 227 landslide locations. From these sources, 13 conditioning factors are extracted to be used in the models. Experimental evaluations are made based on True Skill Statistic (TSS), the Receiver Operation characteristic (ROC) curve and kappa index. The results show that the best TSS (0.6986), ROC (0.904) and kappa (0.6915) were obtained by the ensemble model. FDA on its own seems effective at modelling landslide susceptibility from multiple data sources, with performance comparable to GLM. However, it slightly underperforms when compared to GBM (BRT) and RF. RF seems most capable compared to GBM, GLM, and FDA, when dealing with all conditioning factors.

Forest conservation is crucial for the maintenance of a healthy and thriving ecosystem. The field of remote sensing (RS) has been integral with the wide adoption of computer vision and sensor technologies for forest land observation. One critical area of interest is the detection of active forest fires. A forest fire, which occurs naturally or manually induced, can quickly sweep through vast amounts of land, leaving behind unfathomable damage and loss of lives. Automatic detection of active forest fires (and burning biomass) is hence an important area to pursue to avoid unwanted catastrophes. Early fire detection can also be useful for decision makers to plan mitigation strategies as well as extinguishing efforts. In this paper, we present a deep learning framework called Fire-Net, that is trained on Landsat-8 imagery for the detection of active fires and burning biomass. Specifically, we fuse the optical (Red, Green, and Blue) and thermal modalities from the images for a more effective representation. In addition, our network leverages the residual convolution and separable convolution blocks, enabling deeper features from coarse datasets to be extracted. Experimental results show an overall accuracy of 97.35%, while also being able to robustly detect small active fires. The imagery for this study is taken from Australian and North American forests regions, the Amazon rainforest, Central Africa and Chernobyl (Ukraine), where forest fires are actively reported.

Assessment of the most appropriate groundwater conditioning factors (GCFs) is essential when performing analyses for groundwater potential mapping. For this reason, in this work, we look at three statistical factor analysis methods—Variance Inflation Factor (VIF), Chi-Square Factor Optimization, and Gini Importance—to measure the significance of GCFs. From a total of 15 frequently used GCFs, 11 most effective ones (i.e., altitude, slope angle, plan curvature, profile curvature, topographic wetness index, distance from river, distance from fault, river density, fault density, land use, and lithology) were finally selected. In addition, 917 spring locations were identified and used to train and test three machine learning algorithms, namely Mixture Discriminant Analysis (MDA), Linear Discriminant Analysis (LDA) and Random Forest (RF). The resultant trained models were then applied for groundwater potential prediction and mapping in the Haraz basin of Mazandaran province, Iran. MDA has been successfully applied for soil erosion and landslide mapping, but has not yet been fully explored for groundwater potential mapping (GPM). Although other discriminant methods, such as LDA, exist, MDA is worth exploring due to its capability to model multivariate nonlinear relationships between variables; it also undertakes a mixture of unobserved subclasses with regularization of non-linear decision boundaries, which could potentially provide more accurate classification. For the validation, areas under Receiver Operating Characteristics (ROC) curves (AUC) were calculated for the three algorithms. RF performed better with AUC value of 84.4%, while MDA and LDA yielded 75.2% and 74.9%, respectively. Although MDA performance is lower than RF, the result is satisfactory, because it is within the acceptable standard of environmental modeling. The outcome of factor analysis and groundwater maps emphasizes on optimization of multicolinearity factors for faster spatial modeling and provides valuable information for government agencies and private sectors to effectively manage groundwater in the region.

Large damages and losses resulting from floods are widely reported across the globe. Thus, the identification of the flood-prone zones on a flood susceptibility map is very essential. To do so, 13 conditioning factors influencing the flood occurrence in Brisbane river catchment in Australia (i.e., topographic, water-related, geological, and land use factors) were acquired for further processing and modeling. In this study, artificial neural networks (ANN), deep learning neural networks (DLNN), and optimized DLNN using particle swarm optimization (PSO) were exploited to predict and estimate the susceptible areas to the future floods. The significance of the conditioning factors analysis for the region highlighted that altitude, distance from river, sediment transport index (STI), and slope played the most important roles, whereas stream power index (SPI) did not contribute to the hazardous situation. The performance of the models was evaluated against the statistical tests such as sensitivity, specificity, the area under curve (AUC), and true skill statistic (TSS). DLNN and PSO-DLNN models obtained the highest values of sensitivity (0.99) for the training stage to compare with ANN. Moreover, the validations of specificity and TSS for PSO-DLNN recorded the highest values of 0.98 and 0.90, respectively, compared with those obtained by ANN and DLNN. The best accuracies by AUC were evaluated in PSO-DLNN (0.99 in training and 0.98 in testing datasets), followed by DLNN and ANN. Therefore, the optimized PSO-DLNN proved its robustness to compare with other methods.

The estimation and mapping of forest stand characteristics are vital because this information is necessary for sustainable forest management. The present study considers the use of a Bayesian additive regression trees (BART) algorithm as a non-parametric classifier using Sentinel-2A data and topographic variables to estimate the forest stand characteristics, namely the basal area (m²/ha), stem volume (m³/ha), and stem density (number/ha). These results were compared with those of three other popular machine learning (ML) algorithms, such as generalised linear model (GLM), K-nearest neighbours (KNN), and support vector machine (SVM). A feature selection was done on 28 variables including the multi-spectral bands on Sentinel-2 satellite, related vegetation indices, and ancillary data (elevation, slope, and topographic solar-radiation index derived from digital elevation model (DEM)) and then the most insignificant variables were removed from the datasets by recursive feature elimination (RFE). The study area was a mountainous forest with high biodiversity and an elevation gradient from 26 to 1636 m. An inventory dataset of 1200 sample plots was provided for training and testing the algorithms, and the predictors were fed into the ML models to compute and predict the forest stand characteristics. The accuracies and certainties of the ML models were assessed by their root mean square error (RMSE), mean absolute error (MAE), and R-squared (R²) values. The results demonstrated that BART generated the best basal area and stem volume predictions, followed by GLM, SVM, and KNN. The best RMSE values for both basal area (8.12 m²/ha) and stem volume (29.28 m³/ha) estimation were obtained by BART. Thus, the ability of the BART model for forestry application was established. On the other hand, KNN exhibited the highest RMSE values for all stand variable predictions, thereby exhibiting the least accuracy for this specific application. Moreover, the effectiveness of the narrow Sentinel-2 bands around the red edge and elevation was highlighted for predicting the forest stand characteristics. Therefore, we concluded that the combination of the Sentinel-2 products and topographic variables derived from the PALSAR data used in this study improved the estimation of the forest attributes in temperate forests.