Novel hybrid models to enhance the efficiency of groundwater potentiality model

Applied Water Science, Год журнала: 2022, Номер 12(4)

Опубликована: Март 9, 2022

Abstract The present study aimed to create novel hybrid models produce groundwater potentiality (GWP) in the Teesta River basin of Bangladesh. Six ensemble machine learning (EML) algorithms, such as random forest (RF), subspace, dagging, bagging, naïve Bayes tree (NBT), and stacking, coupled with fuzzy logic (FL) a ROC-based weighting approach have been used for creating integrated GWP. GWP was then verified using both parametric nonparametric receiver operating characteristic curves (ROC), empirical ROC (eROC) binormal curve (bROC). We conducted an RF-based sensitivity analysis compute relevancy conditioning variables modeling. very high potential regions were predicted 831–1200 km 2 521–680 areas based on six EML models. Based area under ROC, NBT (eROC: 0.892; bROC: 0.928) model outperforms rest GPMs considered next step turned into crisp layers membership function, approach. Subsequently four operators assimilate layers, including AND, OR, GAMMA0.8, GAMMA 0.9, well GAMMA0.9. Thus, we created FL model. results eROC bROC showed that 0.9 operator outperformed other operators-based terms accuracy. According validation outcomes, performance. will aid enhancing efficiency preparing viable planning management.

Язык: Английский

---

Enhancing community resilience in arid regions: A smart framework for flash flood risk assessment

Ecological Indicators, Год журнала: 2023, Номер 153, С. 110457 - 110457

Опубликована: Июнь 15, 2023

This paper presents a novel framework for smart integrated risk management in arid regions. The combines flash flood modelling, statistical methods, artificial intelligence (AI), geographic evaluations, analysis, and decision-making modules to enhance community resilience. Flash is simulated by using Watershed Modelling System (WMS). Statistical methods are also used trim outlier data from physical systems climatic data. Furthermore, three AI including Support Vector Machine (SVM), Artificial Neural Network (ANN), Nearest Neighbours Classification (NNC), predict classify occurrences. Geographic Information (GIS) utilised assess potential risks vulnerable regions, together with Failure Mode Effects Analysis (FMEA) Hazard Operability Study (HAZOP) methods. module employs the Classic Delphi technique appropriate solutions control. methodology demonstrated its application real case study of Khosf region Iran, which suffers both drought severe floods simultaneously, exacerbated recent climate changes. results show high Coefficient determination (R2) scores SVM at 0.88, ANN 0.79, NNC 0.89. FMEA indicate that over 50% scenarios risk, while HAZOP indicates 30% same rate. Additionally, peak flows 24 m3/s considered occurrences can cause financial damage all techniques study. Finally, our research findings practical decision support system compatible sustainable development concepts resilience

Язык: Английский

---

Geospatial modelling of floods: a literature review

Stochastic Environmental Research and Risk Assessment, Год журнала: 2023, Номер 37(11), С. 4109 - 4128

Опубликована: Июль 10, 2023

Язык: Английский

---

Research Progress and Prospects of Urban Flooding Simulation: From Traditional Numerical Models to Deep Learning Approaches

Environmental Modelling & Software, Год журнала: 2024, Номер unknown, С. 106213 - 106213

Опубликована: Сен. 1, 2024

Язык: Английский

---

Unveiling global flood hotspots: Optimized machine learning techniques for enhanced flood susceptibility modeling

Journal of Hydrology Regional Studies, Год журнала: 2025, Номер 58, С. 102285 - 102285

Опубликована: Март 4, 2025

Язык: Английский

---

Selecting optimal conditioning parameters for landslide susceptibility: an experimental research on Aqabat Al-Sulbat, Saudi Arabia

Environmental Science and Pollution Research, Год журнала: 2021, Номер 29(3), С. 3743 - 3762

Опубликована: Авг. 13, 2021

Язык: Английский

---

Groundwater potentiality mapping using ensemble machine learning algorithms for sustainable groundwater management

Frontiers in Engineering and Built Environment, Год журнала: 2021, Номер 2(1), С. 43 - 54

Опубликована: Окт. 28, 2021

Purpose The present study aims to construct ensemble machine learning (EML) algorithms for groundwater potentiality mapping (GPM) in the Teesta River basin of Bangladesh, including random forest (RF) and subspace (RSS). Design/methodology/approach RF RSS models have been implemented integrating 14 selected condition parametres with inventories generating GPMs. GPM were then validated using empirical bionormal receiver operating characteristics (ROC) curve. Findings very high (831–1200 km 2 ) potential areas (521–680 predicted EML algorithms. (AUC-0.892) model outperformed based on ROC's area under curve (AUC). Originality/value Two new constructed GPM. These findings will aid proposing sustainable water resource management plans.

Язык: Английский

---

Developing Robust Flood Susceptibility Model with Small Numbers of Parameters in Highly Fertile Regions of Northwest Bangladesh for Sustainable Flood and Agriculture Management

Sustainability, Год журнала: 2022, Номер 14(7), С. 3982 - 3982

Опубликована: Март 28, 2022

The present study intends to improve the robustness of a flood susceptibility (FS) model with small number parameters in data-scarce areas, such as northwest Bangladesh, by employing machine learning-based sensitivity analysis and an analytical hierarchy process (AHP). In this study, nine most relevant elements (such distance from river, rainfall, drainage density) were chosen conditioning variables for modeling. FS was produced using AHP technique. We used empirical binormal receiver operating characteristic (ROC) curves validating models. performed Sensitivity analyses random forest (RF)-based mean Gini decline (MGD), decrease accuracy (MDA), information gain ratio find out sensitive variables. After performing analysis, least eliminated. re-ran rest enhance model’s performance. Based on previous studies weighting approach, general soil type, river/canal (Dr), land use/land cover (LULC) had higher factor weights 0.22, 0.21, 0.19, 0.15, respectively. without well study. According RF-based ratio, factors slope, elevation, while curvature density less parameters, which excluded re-running just vital parameters. Using ROC curves, new yields AUCs 0.835 0.822, It is discovered that predicted may be maintained or increased removing factors. This will aid decision-makers developing management plans examined region.

Язык: Английский

---

Flood vulnerability and buildings’ flood exposure assessment in a densely urbanised city: comparative analysis of three scenarios using a neural network approach

Natural Hazards, Год журнала: 2022, Номер 113(2), С. 1043 - 1081

Опубликована: Апрель 5, 2022

Язык: Английский

---

Assessing landscape ecological vulnerability to riverbank erosion in the Middle Brahmaputra floodplains of Assam, India using machine learning algorithms

CATENA, Год журнала: 2023, Номер 234, С. 107581 - 107581

Опубликована: Окт. 9, 2023

Язык: Английский

---