Deep learning-based landslide susceptibility mapping

Abstract

Landslides are considered as one of the most devastating natural hazards in Iran, causing extensive damage and loss of life. Landslide susceptibility maps for landslide prone areas can be used to plan for and mitigate the consequences of catastrophic landsliding events. Here, we developed a deep convolutional neural network (CNN-DNN) for mapping landslide susceptibility, and evaluated it on the Isfahan province, Iran, which has not previously been assessed on such a scale. The proposed model was trained and validated using training (80%) and testing (20%) datasets, each containing relevant data on historical landslides, field records and remote sensing images, and a range of geomorphological, geological, environmental and human activity factors as covariates. The CNN-DNN model prediction accuracy was tested using a wide range of statistics from the confusion matrix and error indices from the receiver operating characteristic (ROC) curve. The CNN-DNN model was evaluated comprehensively by comparing it to several state-of-the-art benchmark machine learning techniques including the support vector machine (SVM), logistic regression (LR), Gaussian naïve Bayes (GNB), multilayer perceptron (MLP), Bernoulli Naïve Bayes (BNB) and decision tree (DT) classifiers. The CNN-DNN model for landslide susceptibility mapping was found to predict more accurately than the benchmark algorithms, with an AUC = 90.9%, IRs = 84.8%, MSE = 0.17, RMSE = 0.40, and MAPE = 0.42. The map provided by the CNN-DNN clearly revealed a high-susceptibility area in the west and southwest, related to the main Zagros trend in the province. These findings can be of great utility for landslide risk management and land use planning in the Isfahan province.

Figure 1

The main CNN–DNN architecture.

Figure 2

Location map of the study region using ArcGIS 10.4.1 software package.

Figure 3

The geomorphologic factors used in the analysis: (a) altitude variation, (b) slope aspect, (c) slope curvature, (d) profile curvature using ArcGIS 10.4.1 software package.

Figure 4

The geologic factors used in the analysis: (a) geo-units, (b) distance to faults, (c) land-use, (d) soil type, (e) hydrologic variation, (f) slope dip using ArcGIS 10.4.1 software package.

Figure 5

The environmental factors used in the analysis: (a) climate, (b) watershed, (c) drainage pattern, (d) vegetation using ArcGIS 10.4.1 software package.

Figure 6

The human-activity related factors used in the analysis: (a) distance to roads, (b) distance to cities using ArcGIS 10.4.1 software package.

Figure 7

The processing flowchart of the proposed model.

Figure 8

Pearson's coefficient for each information layer.

Figure 9

Landslide susceptibility map for the proposed model using ArcGIS 10.4.1 software package.

Figure 10

The OA and loss function values were obtained for the applied model.

Figure 11

The SGD optimiser results for the proposed model.

Figure 12

The RMSprop optimiser results for the proposed model.

Figure 13

The Adagrad optimiser results for the proposed model.

Figure 14

ROC results for the CNN–DNN model and the benchmark methods.