Article
Environmental Sciences
Chuen-Ming Huang, Chyi-Tyi Lee, Liu-Xuan Jian, Lun-Wei Wei, Wei-Chia Chu, Hsi-Hung Lin
Summary: This study combines deep learning and fuzzy theory concepts to analyze landslide susceptibility and compares the results with logistic regression. The fuzzy neural network performs better than LR in training events but drops in validation events due to bias in landslide identification.
Article
Environmental Sciences
Senem Tekin, Tolga Can
Summary: This study assessed the landslide susceptibility in the Buyuk Menderes watershed using artificial neural network method and generated a susceptibility map. The map showed high accuracy and successfully predicted the locations of landslides in the study area.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2022)
Article
Geosciences, Multidisciplinary
Ali M. Rajabi, Mahdi Khodaparast, Mostafa Mohammadi
Summary: This study used an artificial neural network to conduct risk studies on landslides in the area affected by the Manjil-Rudbar earthquake in Iran in 1990. The results showed that the ANN method is relatively efficient for accurate prediction of landslides, covering 50% of the inventory map of the study area. The hazard map developed through Newmark displacement analysis was compared with other research findings, highlighting the effectiveness of the ANN approach.
Article
Engineering, Geological
Ajaya Pyakurel, Bhim Kumar Dahal, Dipendra Gautam
Summary: This study evaluated the adequacy of various machine learning models using data from the 2015 Gorkha earthquake in Nepal and assessed the importance of various landslide conditioning factors. The results showed that the Extremely Randomized Trees Classifier outperformed other machine learning algorithms in predicting earthquake-induced landslides.
SOIL DYNAMICS AND EARTHQUAKE ENGINEERING
(2023)
Article
Environmental Sciences
Xinyi Guo, Bihong Fu, Jie Du, Pilong Shi, Qingyu Chen, Wenyuan Zhang
Summary: This study compared landslide susceptibility models in seismic regions with different lithological features by selecting the Jiuzhaigou and Minxian earthquakes, finding that a coupling model is suitable for both rock and loess landslides.
Article
Environmental Sciences
Osman Orhan, Suleyman Sefa Bilgilioglu, Zehra Kaya, Adem Kursat Ozcan, Hacer Bilgilioglu
Summary: The main aim of this study was to generate and compare landslide susceptibility maps using five machine learning techniques. Five models were constructed with the help of landslide inventory and conditioning factors, and multiple evaluation metrics were used to validate their performance.
GEOCARTO INTERNATIONAL
(2022)
Article
Engineering, Environmental
Gabriele Amato, Matteo Fiorucci, Salvatore Martino, Luigi Lombardo, Lorenzo Palombi
Summary: The use of Artificial Neural Network (ANN) approaches has played an important role in predicting the distribution of effects triggered by natural forcing in the past decade. Geolocation accuracy, data numerosity, and spatial distribution are critical features of these approaches. This study tested the use of an ANN at a national scale in Italy to estimate earthquake-triggered landslides susceptibility. The ANN-based model showed good spatial prediction performance and revealed areas with high susceptibility in the central and southern/northern Apennines, as well as a division in susceptibility between the western and eastern sectors of the Alpine region.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2023)
Article
Engineering, Environmental
Ankit Tyagi, Reet Kamal Tiwari, Naveen James
Summary: This study presents a scientific method to identify the most significant landslide-causing parameters for an enhanced LSM analysis, and proposes a LSM model for the Himalayan region to improve landslide prediction accuracy.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2023)
Article
Environmental Sciences
Yufei Song, Wen Fan, Ningyu Yu, Yanbo Cao, Chengcheng Jiang, Xiaoqing Chai, Yalin Nan
Summary: This study proposes a new method for calculating the spatiotemporal probability of rainfall-induced landslides based on a Bayesian approach and develops a probabilistic-based early warning model at the regional scale. The results show that the proposed model has higher warning accuracy and economic benefits compared to the conventional model.
Article
Engineering, Environmental
Ananta Man Singh Pradhan, Yun-Tae Kim
Summary: This study developed a new methodological approach to assess the stability of hillslopes at the catchment scale by utilizing GIS-based pseudo-static model and deep learning neural network method integrating extreme rainfall and seismic force variables. The accuracy of the model was assessed to be high, producing reliable landslide susceptibility maps helpful for researchers working on landslide management strategies.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2021)
Article
Engineering, Environmental
Zhiqiang Yang, Chong Xu, Xiaoyi Shao, Siyuan Ma, Lei Li
Summary: This study utilizes a convolutional neural network (CNN-3D) with a spatial-channel attention block to generate landslide susceptibility maps (LSM) for Jiuzhaigou, China, and Iburi, Japan. The results demonstrate that the CNN-based model outperforms machine learning-based models in predicting landslides, generating more accurate and smoother LSM. In particular, CNN-3D performs the best in utilizing 3D information and addressing non-linear relationships.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2022)
Article
Chemistry, Multidisciplinary
Haoran Fang, Yun Shao, Chou Xie, Bangsen Tian, Yu Zhu, Yihong Guo, Qing Yang, Ying Yang
Summary: Earthquakes cause landslides and change the distribution of landslide risk. This study aims to update landslide susceptibility maps after the Jiuzhaigou earthquake using the DS-InSAR technique. The results show that the DS-InSAR technology effectively improves the performance of the landslide susceptibility mapping model and helps track future landslide susceptibility changes. Ecological restoration plays a role in reducing landslide susceptibility.
APPLIED SCIENCES-BASEL
(2022)
Article
Chemistry, Multidisciplinary
Turgut Pura, Peri Gunes, Ali Gunes, Ali Alaa Hameed
Summary: An earthquake is a natural event that can cause significant damage, loss of life, and economic effects. This study focuses on earthquake prediction using the RNN method and incorporates the calculation of b and d values for improved performance. The importance of this study lies in its detection of earthquakes in the Marmara region, classification of seismic data, and generation of future predictions using artificial neural networks.
APPLIED SCIENCES-BASEL
(2023)
Article
Environmental Sciences
Mahyat Shafapourtehrany, Fatemeh Rezaie, Changhyun Jun, Essam Heggy, Sayed M. Bateni, Mahdi Panahi, Haluk Ozener, Farzin Shabani, Hamidreza Moeini
Summary: This study used deep learning models and remote sensing data to generate landslide susceptibility maps, showing that areas with steep slopes, high rainfall, and soil wetness are more susceptible to landslides. This contributes to a better understanding of deep learning applications in the field of natural hazards.
Article
Engineering, Environmental
Kumari Sweta, Ajanta Goswami, Bipin Peethambaran, I. M. Bahuguna, A. S. Rajawat
Summary: Landslide susceptibility zonation is crucial for sustainable development and disaster management in mountainous areas. In this study conducted in the Himalayan region of India, artificial intelligence techniques, specifically the artificial neural network model, were utilized to create a landslide susceptibility map for the area.
BULLETIN OF ENGINEERING GEOLOGY AND THE ENVIRONMENT
(2022)
Article
Environmental Sciences
Ebu Bekir Aygar, Candan Gokceoglu
Summary: This study focuses on the challenges brought by crossing active faults during tunnel construction, proposing a special design and inner lining scheme to address the issues. The effectiveness of the design was verified through numerical analyses.
ENVIRONMENTAL EARTH SCIENCES
(2021)
Article
Materials Science, Characterization & Testing
N. Yesiloglu-Gultekin, C. Gokceoglu
Summary: Basalt is widely used in construction for its properties. This study aims to develop non-linear prediction models for uniaxial compressive strength (UCS) and elasticity modulus (E-i) by using simple and non-destructive test results. The performance of different algorithms were assessed and compared using various metrics. The results show that the ANFIS model performs slightly better in predicting UCS, while the ANN model is the most successful in predicting E-i. The models using porosity and sonic velocity as input parameters exhibit the highest correlation with observed data for predicting UCS, while the models with three inputs perform best for predicting E-i.
JOURNAL OF NONDESTRUCTIVE EVALUATION
(2022)
Article
Multidisciplinary Sciences
Candan Gokceoglu
Summary: This study aims to develop models to predict the rate of penetration (ROP) of tunnel boring machines (TBMs). By analyzing data from the longest railway tunnel in Turkey, it is found that the performances of artificial neural network (ANN) models are considerably better than those of multiple regression equations. The ANN models developed in this study can be reliably used for deep tunnel construction in metamorphic rock medium. However, the performances of multiple regression equations need improvement.
SN APPLIED SCIENCES
(2022)
Article
Computer Science, Interdisciplinary Applications
Abidhan Bardhan, Navid Kardani, Abdel Kareem Alzo'ubi, Pijush Samui, Amir H. Gandomi, Candan Gokceoglu
Summary: This study presents a comparative analysis of hybrid machine learning models for estimating the compression index (C-c) of clay. The proposed ANFIS-PSO model outperforms other models and shows high potential as an alternative to the actual oedometer test in civil engineering projects.
ARCHIVES OF COMPUTATIONAL METHODS IN ENGINEERING
(2022)
Article
Environmental Sciences
C. Gokceoglu, B. Unutmaz
Summary: Waste dams are commonly used for storing mining byproducts and their construction is of high importance due to potential harm to the environment and nearby individuals. This study utilizes 3D finite-element analysis to investigate a nickel-ore waste dam in Turkey and examines the effects of slope instability on the main dam body. The results suggest that after slope rehabilitation, deformations significantly decreased and the waste dam became safer.
ENVIRONMENTAL EARTH SCIENCES
(2022)
Article
Environmental Sciences
Beste Tavus, Sultan Kocaman, Candan Gokceoglu
Summary: This study evaluated the flood damage mapping performances of two satellite Earth Observation sensors, Sentinel-1 and Sentinel-2, using the Random Forest supervised classification method and various feature types in the Sardoba Reservoir area. The results show that the fusion of S1 and S2 data exhibits high classification accuracy for flooded areas, particularly in separating inundated vegetation.
SCIENCE OF THE TOTAL ENVIRONMENT
(2022)
Article
Engineering, Civil
Ebu Bekir Aygar, Servet Karahan, Suat Gullu, Candan Gokceoglu
Summary: This study investigates the suitability and seismic sustainability of rigid support systems in the Dogancay T1 tunnel using analytical and numerical methods. The results show that the proposed support system can be successfully implemented in the tunnel.
TRANSPORTATION INFRASTRUCTURE GEOTECHNOLOGY
(2022)
Article
Construction & Building Technology
Candan Gokceoglu, Ebu Bekir Aygar, Hakan A. A. Nefeslioglu, Servet Karahan, Suat Gullu
Summary: The T26 tunnel encountered serious problems and excessive deformations during excavation, leading to suspension of the project. The issues were addressed through redesign and the use of the New Austrian Tunneling Method. This study aims to describe the problems encountered in the T26 tunnel and discuss the advantages and disadvantages of TBM and NATM methods for tunnels with difficult ground conditions.
Article
Engineering, Geological
C. Gokceoglu, C. Bal, C. H. Aladag
Summary: Prediction of tunnel boring machine (TBM) performance is still a challenging research subject. In this study, geological and geotechnical parameters were used to predict TBM performance. The random forest algorithm showed superior performance compared to other methods.
GEOTECHNICAL AND GEOLOGICAL ENGINEERING
(2023)
Letter
Construction & Building Technology
C. Gokceoglu
CONSTRUCTION AND BUILDING MATERIALS
(2023)
Article
Chemistry, Multidisciplinary
Nazli Tunar Ozcan, Candan Gokceoglu
Summary: This study discusses ground improvement techniques using jet grouting to prevent liquefaction of marine sediments in the construction of a FSRU terminal in Saros Bay, Turkey. It presents a practical quality control procedure for offshore grouting operations and assesses the performance of jet columns. The results show that the jet grout applications following this procedure are suitable and effective for improving offshore soils.
APPLIED SCIENCES-BASEL
(2023)
Article
Engineering, Geological
Gokhan Tacim, Evren Posluk, Candan Gokceoglu
Summary: In the planning of tunnel support system, investigation of structure-tunnel interaction is crucial, especially in the area with unpredictable geological conditions. This study focuses on a single-track railway tunnel in a karstic limestone area, investigating structure-tunnel interaction and the importance of grouting. It is found that the presence of karstic caves and heavily fractured nature of the limestone pose challenges to the tunnel construction, which can be mitigated through grouting.
INTERNATIONAL JOURNAL OF GEO-ENGINEERING
(2023)
Proceedings Paper
Geography, Physical
I Yalcin, R. Can, S. Kocaman, C. Gokceoglu
Summary: This study proposes a Convolutional Neural Network (CNN) architecture to automatically identify discontinuities in rock masses. Close-range photogrammetry methods were used to produce orthophotos, and training data was generated through manual mensuration. Preliminary results show that the proposed method has a high capability for determining discontinuities, but there are issues with image quality and discontinuity identification.
XXIV ISPRS CONGRESS IMAGING TODAY, FORESEEING TOMORROW, COMMISSION II
(2022)
Proceedings Paper
Geography, Physical
G. Karakas, S. Kocaman, C. Gokceoglu
Summary: Generating accurate and up-to-date landslide susceptibility maps in landslide-prone areas is crucial for identifying future hazard potential. The accuracy and spatial resolution of the digital elevation models (DEMs) used as input are among the most important factors affecting the accuracy of the maps.
XXIV ISPRS CONGRESS: IMAGING TODAY, FORESEEING TOMORROW, COMMISSION III
(2022)