Impact of climate change on future flood susceptibility projections under shared socioeconomic pathway scenarios in South Asia using artificial intelligence algorithms

Journal of Environmental Management, Journal Year: 2024, Volume and Issue: 366, P. 121764 - 121764

Published: July 8, 2024

Language: Английский

---

Performance assessment of the landslide susceptibility modelling using the support vector machine, radial basis function network, and weight of evidence models in the N'fis river basin, Morocco

Geoscience Letters, Journal Year: 2022, Volume and Issue: 9(1)

Published: Oct. 15, 2022

Abstract Landslides in mountainous areas are one of the most important natural hazards and potentially cause severe damage loss human life. In order to reduce this damage, it is essential determine vulnerable sites. The objective study was produce a landslide vulnerability map using weight evidence method (WoE), Radial Basis Function Network (RBFN), Support Vector Machine (SVM) for N'fis basin located on northern border Marrakech High Atlas, area prone landslides. Firstly, an inventory historical landslides carried out based interpretation satellite images field surveys. A total 156 events were mapped area. 70% data from (110 events) used model training remaining 30% (46 validation. Next, fourteen thematic maps causative factors, including lithology, slope, elevation, profile curvature, slope aspect, distance rivers, topographic moisture index (TWI), position (TPI), faults, roads, normalized difference vegetation (NDVI), precipitation, land use/land cover (LULC), soil type, determined created available spatial database. Finally, susceptibility produced three models: WoE, RBFN, SVM. results validated several statistical indices receiver operating characteristic curve. AUC values SVM, WoE models 94.37%, 93.68%, 83.72%, respectively. Hence, we can conclude that SVM RBFN have better predictive capabilities than model. obtained could be helpful local decision-makers LULC planning risk mitigation.

Language: Английский

---

Implementation of Soil and Water Conservation in Indonesia and Its Impacts on Biodiversity, Hydrology, Soil Erosion and Microclimate

Applied Sciences, Journal Year: 2023, Volume and Issue: 13(13), P. 7648 - 7648

Published: June 28, 2023

Soil and water are natural resources that support the life of various creatures on Earth, including humans. The main problem, so far, is both can be easily damaged or degraded by human-induced drivers. threat damage degradation increasing due to rapid human population growth humans’ insatiable daily necessities. Indonesia has had experiences in soil conservation (SWC) programmes for a long time, which good lesson learned future strategy development. This article aims provide an overview benefits implementing SWC biodiversity, hydrology, erosion, microclimate sustainable ecological landscape management. Various vegetative mechanical techniques have been known implemented utilized improve strategies. It expected proper development will sustainability Forthcoming also incorporate local knowledge into their implementation. require coordination between stakeholders, i.e., communities, management authorities, policymakers, scientists, seamless integration varying fields levels governance. findings this study increased adaptation native plants rainfall conditions; infiltration improved hydrological characteristics; SWC, through vegetation techniques, played role lowering temperatures, humidity, reducing intensity levels.

Language: Английский

---

Implementation of random forest, adaptive boosting, and gradient boosting decision trees algorithms for gully erosion susceptibility mapping using remote sensing and GIS

Environmental Earth Sciences, Journal Year: 2024, Volume and Issue: 83(3)

Published: Feb. 1, 2024

Language: Английский

---

Spatial Prediction and Mapping of Gully Erosion Susceptibility Using Machine Learning Techniques in a Degraded Semi-Arid Region of Kenya

Land, Journal Year: 2023, Volume and Issue: 12(4), P. 890 - 890

Published: April 15, 2023

This study aimed at (i) developing, evaluating and comparing the performance of support vector machines (SVM), boosted regression trees (BRT), random forest (RF) logistic (LR) models in mapping gully erosion susceptibility, (ii) determining important conditioning factors (GECFs) a Kenyan semi-arid landscape. A total 431 geo-referenced points were gathered through field survey visual interpretation high-resolution satellite imagery on Google Earth, while 24 raster-based GECFs retrieved from existing geodatabases for spatial modeling prediction. The resultant exhibited excellent performance, although machine learners outperformed benchmark LR technique. Specifically, RF BRT returned highest area under receiver operating characteristic curve (AUC = 0.89 each) overall accuracy (OA 80.2%; 79.7%, respectively), followed by SVM 0.86; 0.85 & OA 79.1%; 79.6%, respectively). In addition, importance varied among models. best-performing model ranked distance to stream, drainage density valley depth as three most region. output susceptibility maps can efficient allocation resources sustainable land management area.

Language: Английский

---

An integrated approach based landslide susceptibility mapping: case of Muzaffarabad region, Pakistan

Geomatics Natural Hazards and Risk, Journal Year: 2023, Volume and Issue: 14(1)

Published: May 9, 2023

Landslides result in the devastation of property and loss lives. This study assesses landslide susceptibility by employing geographic information systems (GIS) machine learning techniques, that is, support vector (SVM) artificial neural network (ANN), with integration advanced optimization particle swarm (PSO). The landslide-inducing factors considered this include fault density, lithology, road slope, elevation, flow direction, aspect, earthquake intensity, curvature, Normalized Difference Water Index (NDWI), waterways rainfall, Vegetation (NDVI). resulting maps (LSMs) showed areas falling under high very class have higher rainfall levels, weak NDWI, direction. accuracy assessment techniques ANN an Area Under Curve (AUC) 0.81 performed better than SVM AUC 0.78 without optimization. Similarly, performance was also using PSO. During integrated modeling, PSO-ANN 0.87, whereas PSO–SVM 0.84. produced LSMs a similar trend terms percentage as models.

Language: Английский

---

Gully erosion mapping susceptibility in a Mediterranean environment: A hybrid decision-making model

International Soil and Water Conservation Research, Journal Year: 2023, Volume and Issue: 12(2), P. 279 - 297

Published: Oct. 7, 2023

Gully erosion is one of the main natural hazards, especially in arid and semi-arid regions, destroying ecosystem service human well-being. Thus, gully susceptibility maps (GESM) are urgently needed for identifying priority areas on which appropriate measurements should be considered. Here, we proposed four new hybrid Machine learning models, namely weight evidence -Multilayer Perceptron (MLP- WoE), –K Nearest neighbours (KNN- - Logistic regression (LR- Random Forest (RF- mapping exploring opportunities GIS tools Remote sensing techniques El Ouaar watershed located Souss plain Morocco. Inputs developed models composed dependent (i.e., points) a set independent variables. In this study, total 314 points were randomly split into 70% training stage (220 gullies) 30% validation (94 sets identified study area. 12 conditioning variables including elevation, slope, plane curvature, rainfall, distance to road, stream, fault, TWI, lithology, NDVI, LU/LC used based their importance mapping. We evaluate performance above following statistical metrics: Accuracy, precision, Area under curve (AUC) values receiver operating characteristics (ROC). The results indicate RF- WoE model showed good accuracy with (AUC = 0.8), followed by KNN-WoE 0.796), then MLP-WoE 0.729) LR-WoE 0.655), respectively. provide information valuable tool decision-makers planners identify where urgent interventions applied.

Language: Английский

---

Exploring deep learning models for roadside landslide prediction: Insights and implications from comparative analysis

Physics and Chemistry of the Earth Parts A/B/C, Journal Year: 2024, Volume and Issue: unknown, P. 103741 - 103741

Published: Sept. 1, 2024

Language: Английский

---

Spatial implementation of frequency ratio, statistical index and index of entropy models for landslide susceptibility mapping in Al-Balouta river basin, Tartous Governorate, Syria

Geoscience Letters, Journal Year: 2022, Volume and Issue: 9(1)

Published: Dec. 26, 2022

Abstract Landslide vulnerability prediction maps are among the most important tools for managing natural hazards associated with slope stability in river basins that affect ecosystems, properties, infrastructure and society. events hazardous patterns of instability coastal mountains Syria. Thus, main goals this research to evaluate performance three different statistical outputs: Frequency Ratio (FR), Statistical Index (SI) Entropy (IoE) therefore map landslide susceptibility region To end, we identified a total 446 locations events, based on preliminary inventory derived from fieldwork high-resolution imagery surveys. In regard, 13 geo-environmental factors have high influence landslides were selected mapping. The results indicated FR method outperformed SI IoE models AUC 0.824 better adaptability, followed by 0.791. According SCAI values, although model achieved best reliability, other two also showed good capability determining susceptibility. result FR-based modelling 18.51 19.98% study area fall under very susceptible categories, respectively. generated method, about 36% is classified as having or sensitivity. whereas 14.18 25.62% “very susceptible” “high susceptible,” relative importance analysis demonstrated aspects, lithology proximity roads effectively motivated acceleration material influential both models. On hand, faults roads, along factor, influences formation events. As result, bivariate models-based mapping provided reliable systematic approach guide long-term strategic planning procedures area.

Language: Английский

---

Predictive machine learning for gully susceptibility modeling with geo-environmental covariates: main drivers, model performance, and computational efficiency

Natural Hazards, Journal Year: 2024, Volume and Issue: 120(8), P. 7211 - 7244

Published: March 12, 2024

Abstract Currently, machine learning (ML) based gully susceptibility prediction is a rapidly expanding research area. However, when assessing the predictive performance of ML models, previous frequently overlooked critical component computational efficiency in favor accuracy. This study aimed to evaluate and compare six commonly used algorithms modeling. Artificial neural networks (ANN), partial least squares, regularized discriminant analysis, random forest (RF), stochastic gradient boosting, support vector (SVM) were applied. The comparison was conducted under three scenarios input feature set sizes: small (six features), medium (twelve large (sixteen features). Results indicated that SVM most efficient algorithm with medium-sized set, outperforming other across all overall accuracy (OA) metrics (OA = 0.898, F 1-score 0.897) required relatively short computation time (< 1 min). Conversely, ensemble-based algorithms, mainly RF, larger reach optimal computationally demanding, taking about 15 min compute. ANN also showed sensitivity number features, but unlike its consistently decreased sets. Among geo-environmental covariates, NDVI, followed by elevation, TWI, population density, SPI, LULC, for Therefore, using involving these covariates modeling similar environmental settings strongly suggested ensure higher minimal time.

Language: Английский

---