Deep Learning and Machine Learning Models for Landslide Susceptibility Mapping with Remote Sensing Data

Abstract

Karakoram Highway (KKH) is an international route connecting South Asia with Central Asia and China that holds socio-economic and strategic significance. However, KKH has extreme geological conditions that make it prone and vulnerable to natural disasters, primarily landslides, posing a threat to its routine activities. In this context, the study provides an updated inventory of landslides in the area with precisely measured slope deformation (Vslope), utilizing the SBAS-InSAR (small baseline subset interferometric synthetic aperture radar) and PS-InSAR (persistent scatterer interferometric synthetic aperture radar) technology. By processing Sentinel-1 data from June 2021 to June 2023, utilizing the InSAR technique, a total of 571 landslides were identified and classified based on government reports and field investigations. A total of 24 new prospective landslides were identified, and some existing landslides were redefined. This updated landslide inventory was then utilized to create a landslide susceptibility model, which investigated the link between landslide occurrences and the causal variables. Deep learning (DL) and machine learning (ML) models, including convolutional neural networks (CNN 2D), recurrent neural networks (RNNs), random forest (RF), and extreme gradient boosting (XGBoost), are employed. The inventory was split into 70% for training and 30% for testing the models, and fifteen landslide causative factors were used for the susceptibility mapping. To compare the accuracy of the models, the area under the curve (AUC) of the receiver operating characteristic (ROC) was used. The CNN 2D technique demonstrated superior performance in creating the landslide susceptibility map (LSM) for KKH. The enhanced LSM provides a prospective modeling approach for hazard prevention and serves as a conceptual reference for routine management of the KKH for risk assessment and mitigation.

Integrating Machine Learning Ensembles for Landslide Susceptibility Mapping in Northern Pakistan

“…Landslide inventory maps have several objectives, ranging from the documentation of diverse landslide types to identifying geographical locations within a specific region. These play a significant role in supplying fundamental data for the formulation of models related to landslide risk or susceptibility [44,45]. Moreover, these maps quantify the limits of mass movements, determine statistical indexes for the frequency and spatial distribution of failures of a slope, and regress the consequences of particular landslide-triggering events, i.e., intense rainfall, rapid snowmelt, seismic activity, etc.…”

A Multichannel-Based Deep Learning Framework for Ocean SAR Scene Classification

“…Moreover, SAR data have been used in diverse applications, including fish species recognition [8], urban land classification [9], marine spatial planning [10], and earthquake disaster prediction [11]. To further broaden the versatility of SAR technology and highlight the potential to contribute to an array of scientific and practical domains, geophysical researchers have employed machine learning and deep learning techniques to support fish species recognition, military objective protection, and natural hazard prevention [12][13][14]. However, SAR images inherently exhibit speckle noise due to the attenuation of echo signals, which is characterized by a distribution of granular patterns [15].…”