Enhancing landslide disaster prediction by evaluating non landslide area sampling in machine learning models for Spiti Valley India

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 10, 2025

Language: Английский

---

Slope stability and surface displacement analysis of the Kuther Landslide in the Dehar Watershed, Himachal Himalaya, Northern India

Discover Geoscience, Journal Year: 2025, Volume and Issue: 3(1)

Published: Jan. 7, 2025

The road network in the Himalayan terrain is vital for India's socio-economic development and national security, yet complexities of topography, geological structures, diverse lithology, neotectonics make planning maintaining these routes a challenging task. Population growth expanding construction have caused slope destabilization, mass wasting, movement across terrain. Field-based stability assessments are essential characterization stabilization, helping planners predict select suitable strategies roads other infrastructure. This study presents comprehensive displacement analysis Kuther landslide, situated along Jawali to Dehar Watershed, Kangra District, Himachal Pradesh. An integrated approach was used, combining surface monitoring with COSI-Corr technique through methods such as rock rating, kinematic analysis, continuous rating. Findings indicate that highly susceptible planar failure, some instances wedge failure. ranges from class III IV, indicating conditions partially stable stable, while reveals creeping at an average rate 4 mm/year. highlights critical need analyses ensure safety vulnerable areas their communities.

Language: Английский

---

Effective prediction of earthquake-induced slope displacements, considering region-specific seismotectonic and climatic conditions

Natural Hazards, Journal Year: 2025, Volume and Issue: unknown

Published: March 18, 2025

Language: Английский

---

Landslide Susceptibility Mapping in Complex Topo‐Climatic Himalayan Terrain, India Using Machine Learning Models: A Comparative Study of XGBoost, RF and ANN

Geological Journal, Journal Year: 2025, Volume and Issue: unknown

Published: March 20, 2025

ABSTRACT Landslides present a significant danger to both infrastructure and human lives in the challenging terrain of Himalayas. Therefore, it is crucial accurately map areas prone landslides facilitate informed decision‐making proactive planning, allowing for effective management this hazard. Since landslide occurrences are accentuated by floods through toe erosion, wildfires research aims integrate machine learning techniques with analysis multiple hazards, such as forest fires, novel conditioning factors create comprehensive susceptibility. Geospatial was conducted examine relationship between 19 elements, including related flood fire susceptibility, which contribute occurrence landslides. This study tested efficacy three models mapping landslide‐prone areas: eXtreme Gradient Boost (XGBoost), Random Forest (RF) Artificial Neural Network (ANN). These can identify complex correlations patterns among resulting more accurate regions A regression performed evaluate multicollinearity confirm association dependent independent variables. The revealed variance inflation factor within acceptable bounds, providing validation correlation. ROC–AUC curve approach used assess models' accuracy. Among tested, XGB exhibited highest accuracy at 94%, followed RF 92% ANN 77%. results offer insightful information about how combine data from various hazard forecast work be instrumental local authorities disaster organisations prioritising resources, implementing mitigation plans enhancing resilience against threats.

Language: Английский

---

Landslide-induced vulnerability of road networks in Lahaul and Spiti, India: a geospatial study

Bulletin of Engineering Geology and the Environment, Journal Year: 2025, Volume and Issue: 84(6)

Published: May 24, 2025

Language: Английский

---

Landslide susceptibility analysis in the Bhilangana Basin (India) using GIS-based machine learning methods

Geosystems and Geoenvironment, Journal Year: 2024, Volume and Issue: 3(2), P. 100253 - 100253

Published: Jan. 13, 2024

Landslides are frequent natural hazards in mountainous regions, and harshly upset people's lives livelihoods. In the present study, we have carried out an analysis of seven GIS-based machine-learning techniques; asses their performance for landslide susceptibility mapping (LSM) Bhilangana Basin, Garhwal Himalaya. A inventory consisting 423 polygons was prepared using repeated field investigations, multi-dated satellite images periods between 2000 to 2022. The dataset classified into two groups: training (70%) test (30%), 12 predictive variables were used LSM. methods produce LSM boosted regression tree (BRT), Fisher discriminant (FDA), generalized linear model (GLM), multivariate adaptive splines (MARS), model-architect (MDA), random forest (RF), support vector machine (SVM). sensitivity these models predict susceptible areas area under curve (AUC) method. RF (AUC = 0.988) has given highest precision indicating best performance. Though MARS (0.974), SVM (0.965), MDA (0.952) also performed adequately (all AUC values above 0.95), however, it is recommended that highly suitable region.

Language: Английский

---

Concepts of Disasters and Research Themes: Editorial Message

Springer eBooks, Journal Year: 2024, Volume and Issue: unknown, P. 3 - 39

Published: Jan. 1, 2024

Language: Английский

---

Landslides: Unfolding Slope Disasters in Hilly Terrains

Journal of the Geological Society of India, Journal Year: 2024, Volume and Issue: 100(5), P. 619 - 621

Published: May 1, 2024

Language: Английский

---

Predictive Modeling of Landslide Susceptibility in Soft Soil Canal Regions: A Focus on Early Warning Systems

International Journal of Advanced Computer Science and Applications, Journal Year: 2023, Volume and Issue: 14(12)

Published: Jan. 1, 2023

The Mekong Delta (MD) has suffered significant losses in land resources, economic damage, and human property casualties due to recent landslides. An early warning system for landslides is a valuable tool identifying the effectiveness timely detection of changes soil promptly determine solutions minimize damage caused by an area. In this study, we apply machine learning approach based on Long Short-Term Memory (LSTM) algorithm experiment with landslide events soft MD. Horizontal pressure, change inclination angles sensor pile mass sliding both x y directions, levels are determined deformation displacement along riverbank, considered candidate factors inputs model. Data from established used train model, creating training testing dataset 374,415 samples. accuracy classification threshold proposed be measured using average F1 score derived precision recall values. optimal prediction results gleaned observational window 4 minutes 30 seconds project roughly 2 hours into future. validation process resulted recall, precision, F1-score stands at 0.8232 remarkably low standard deviation about 1%. successful application research can help identify abnormal leading riverbank loading, thereby conditions developing reliable information provide managers ability suggest protect lives, residents infrastructures.

Language: Английский

---

Predicting Rainfall by Fuzzy Logic

Advances in computational intelligence and robotics book series, Journal Year: 2024, Volume and Issue: unknown, P. 99 - 117

Published: April 29, 2024

Water is vital to all living things; water life. According the UNDP scarcity affects more than two billion people and it projected rise as temperatures do due climate change. The chapter presents some preliminary results of rainfall prediction for case Mexico City by considering input variables, temperature (T) wind speed (WS). A fuzzy logic rule-based approach was employed in analysis. presented model has potential not only predict but also drought. Moreover, been highlighted that becomes necessary address droughts designing implementing drought disaster management systems mitigate impact such events. Therefore, prediction, an early warning, plays a key role measures achieve this. More generally, hoped approaches herein may contribute better understanding impacts regions, cities, communities prone hazard.

Language: Английский

---