Multitemporal analysis of land subsidence induced by open-pit mining activity using improved combined scatterer interferometry with deep learning algorithm optimization.

Mine operational safety is an important aspect of maintaining the operational continuity of a mining area. In this study, we used the InSAR time series to analyze land surface changes using the ICOPS (improved combined scatterers with optimized point scatters) method. This ICOPS method combines persistent scatterers (PS) with distributed scatterers (DS) to increase surface deformation analysis's spatial coverage and quality. One of the improvements of this study is the use of machine learning in postprocessing, based on convolutional neural networks, to increase the reliability of results. This study used data from the Sentinel-1 SAR C-band satellite during the 2016-2022 observation period at the Musan mine, North Korea. In the InSAR surface deformation time analysis, the maximum average rate of land subsidence was approximately > 15.00 cm per year, with total surface deformation of 170 cm and 70 cm for the eastern dumping area and the western dumping area, respectively. Analyzing the mechanism of land surface changes also involved evaluating the geological conditions in the Musan mining area. Our research findings show that combining machine learning and statistical methods has great potential to enhance the understanding of mine surface deformation.

Convolutional neural network (CNN) with metaheuristic optimization algorithms for landslide susceptibility mapping in Icheon, South Korea.

Landslides are a geological hazard that can pose a serious threat to human health and the environment of highlands or mountain slopes. Landslide susceptibility mapping is an essential tool for predicting and mitigating landslides. This study aimed to investigate the application of deep learning algorithms based on convolutional neural networks (CNNs) with metaheuristic optimization algorithms, namely the grey wolf optimizer (GWO) and imperialist competitive algorithm (ICA), to landslide susceptibility mapping. The study area was Icheon City, South Korea, for which an accurate landslide inventory dataset was available. The landslide inventory map was prepared and randomly divided into datasets of 70% for training and 30% for validation. Additionally, 18 landslide-related factors, including geo-environmental and topo-hydrological factors, were considered as predictive variables. The models were compared using area under the curve (AUC) values in receiver operating characteristic (ROC) curve analysis. The validation results showed that optimized models based on CNN-GWO (AUC = 0.876, RMSE = 0.08) and CNN-ICA (AUC = 0.852, RMSE = 0.09) outperformed the standalone CNN model (AUC = 0.847, RMSE = 0.12). Nevertheless, the CNN model outperformed previous research that used a machine learning algorithm alone. Thus, the deep learning algorithm with optimization algorithms proposed in this study can generate more suitable models for landslide susceptibility mapping in the study area due to its improved accuracy.