Machine Learning-Based Gully Erosion Susceptibility Mapping: A Case Study of Eastern India.

Gully erosion is a form of natural disaster and one of the land loss mechanisms causing severe problems worldwide. This study aims to delineate the areas with the most severe gully erosion susceptibility (GES) using the machine learning techniques Random Forest (RF), Gradient Boosted Regression Tree (GBRT), Naïve Bayes Tree (NBT), and Tree Ensemble (TE). The gully inventory map (GIM) consists of 120 gullies. Of the 120 gullies, 84 gullies (70%) were used for training and 36 gullies (30%) were used to validate the models. Fourteen gully conditioning factors (GCFs) were used for GES modeling and the relationships between the GCFs and gully erosion was assessed using the weight-of-evidence (WofE) model. The GES maps were prepared using RF, GBRT, NBT, and TE and were validated using area under the receiver operating characteristic(AUROC) curve, the seed cell area index (SCAI) and five statistical measures including precision (PPV), false discovery rate (FDR), accuracy, mean absolute error (MAE), and root mean squared error (RMSE). Nearly 7% of the basin has high to very high susceptibility for gully erosion. Validation results proved the excellent ability of these models to predict the GES. Of the analyzed models, the RF (AUROC = 0.96, PPV = 1.00, FDR = 0.00, accuracy = 0.87, MAE = 0.11, RMSE = 0.19 for validation dataset) is accurate enough for modeling and better suited for GES modeling than the other models. Therefore, the RF model can be used to model the GES areas not only in this river basin but also in other areas with the same geo-environmental conditions.

Measuring landslide vulnerability status of Chukha, Bhutan using deep learning algorithms.

Landslides are major natural hazards that have a wide impact on human life, property, and natural environment. This study is intended to provide an improved framework for the assessment of landslide vulnerability mapping (LVM) in Chukha Dzongkhags (district) of Bhutan. Both physical (22 nos.) and social (9 nos.) conditioning factors were considered to model vulnerability using deep learning neural network (DLNN), artificial neural network (ANN) and convolution neural network (CNN) approaches. Selection of the factors was conceded by the collinearity test and information gain ratio. Using Google Earth images, official data, and field inquiry a total of 350 (present and historical) landslides were recorded and training and validation sets were prepared following the 70:30 ratio. Nine LVMs were produced i.e. a landslide susceptibility (LS), one social vulnerability (SV) and a relative vulnerability (RLV) map for each model. The performance of the models was evaluated by area under curve (AUC) of receiver operating characteristics (ROC), relative landslide density index (R-index) and different statistical measures. The combined vulnerability map of social and physical factors using CNN (CNN-RLV) had the highest goodness-of-fit and excellent performance (AUC = 0.921, 0.928) followed by DLNN and ANN models. This approach of combined physical and social factors create an appropriate and more accurate LVM that may-support landslide prediction and management.

A novel ensemble computational intelligence approach for the spatial prediction of land subsidence susceptibility.

Land subsidence (LS) is a significant problem that can cause loss of life, damage property, and disrupt local economies. The Semnan Plain is an important part of Iran, where LS is a major problem for sustainable development and management. The plain represents the changes occurring in 40% of the country. We introduce a novel-ensemble intelligence approach (called ANN-bagging) that uses bagging as a meta- or ensemble-classifier of an artificial neural network (ANN) to predict LS spatially on the Semnan Plain in Semnan Province, Iran. The ensemble model's goodness-of-fit (to training data) and prediction accuracy (of the validation data) are compared to benchmarks set by ANN-bagging. A total of 96 locations of LS and 12 LS conditioning factors (LSCFs) were collected. Each feature in the LS inventory map (LSIM) was randomly assigned to one of four groups or folds, each comprising 25% of cases. The novel ensemble model was trained using 75% (3 folds) and validated with the remaining 25% (1 fold) in a four-fold cross-validation (CV) system, which is used to control for the effects of the random selection of the training and validation datasets. LSCFs for LS prediction were selected using the information-gain ratio and multi-collinearity test methods. Factor significance was evaluated using a random forest (RF) model. Groundwater drawdown, land use and land cover, elevation, and lithology were the most important LSCFs. Using the k-fold CV approaches, twelve LS susceptibility maps (LSSMs) were prepared as each fold employed all three models (ANN-bagging, ANN, and bagging). The LS susceptibility mapping showed that between 5.7% and 12.6% of the plain had very high LS susceptibility. All three models produced LS susceptibility maps with acceptable prediction accuracies and goodness-of-fits, but the best maps were produced by the ANN-bagging ensemble method. Overall, LS risk was highest in agricultural areas with high groundwater drawdown in the flat lowlands on quaternary sediments (Qcf). Groundwater extraction rates should be monitored and potentially limited in regions of severe or high LS susceptibility. This investigation details a novel methodology that can help environmental planners and policy makers to mitigate LS to help achieve sustainability.