Interpretability of simple RNN and GRU deep learning models used to map land susceptibility to gully erosion

The Science of The Total Environment, Год журнала: 2023, Номер 904, С. 166960 - 166960

Опубликована: Сен. 9, 2023

Язык: Английский

---

Exploring forest fire susceptibility and management strategies in Western Himalaya: Integrating ensemble machine learning and explainable AI for accurate prediction and comprehensive analysis

Environmental Technology & Innovation, Год журнала: 2024, Номер 35, С. 103655 - 103655

Опубликована: Май 5, 2024

Forest fires pose a significant threat to ecosystems and socio-economic activities, necessitating the development of accurate predictive models for effective management mitigation. In this study, we present novel machine learning approach combined with Explainable Artificial Intelligence (XAI) techniques predict forest fire susceptibility in Nainital district. Our innovative methodology integrates several robust — AdaBoost, Gradient Boosting Machine (GBM), XGBoost Random Deep Neural Network (DNN) as meta-model stacking framework. This not only utilises individual strengths these models, but also improves overall prediction performance reliability. By using XAI techniques, particular SHAP (SHapley Additive exPlanations) LIME (Local Interpretable Model-agnostic Explanations), improve interpretability provide insights into decision-making processes. results show effectiveness ensemble model categorising different zones: very low, moderate, high high. particular, identified extensive areas susceptibility, precision, recall F1 values underpinning their effectiveness. These achieved ROC AUC above 0.90, performing exceptionally well an 0.94. The are remarkably inclusion confidence intervals most important metrics all emphasises robustness reliability supports practical use management. Through summary plots, analyze global variable importance, revealing annual rainfall Evapotranspiration (ET) key factors influencing susceptibility. Local analysis consistently highlights importance rainfall, ET, distance from roads across models. study fills research gap by providing comprehensive interpretable modelling that our ability effectively manage risk is consistent environmental protection sustainable goals.

Язык: Английский

---

Using an interpretable deep learning model for the prediction of riverine suspended sediment load

Environmental Science and Pollution Research, Год журнала: 2024, Номер 31(22), С. 32480 - 32493

Опубликована: Апрель 24, 2024

Язык: Английский

---

Interpreting optimised data-driven solution with explainable artificial intelligence (XAI) for water quality assessment for better decision-making in pollution management

Environmental Science and Pollution Research, Год журнала: 2024, Номер 31(30), С. 42948 - 42969

Опубликована: Июнь 17, 2024

Язык: Английский

---

A GIS-based modified PAP/RAC model and Caesium-137 approach for water erosion assessment in the Raouz catchment, Morocco

Environmental Research, Год журнала: 2024, Номер 251, С. 118460 - 118460

Опубликована: Фев. 20, 2024

Язык: Английский

---

Towards a more comprehensive scenario analysis: Response of soil erosion to future land use and climate change in the Central Yunnan Urban Agglomeration, China

Journal of Environmental Management, Год журнала: 2025, Номер 376, С. 124523 - 124523

Опубликована: Фев. 16, 2025

Rapid urbanization and climate change exacerbate soil erosion globally, threatening ecosystem services sustainable development. However, current predictive studies on future often lack comprehensive consideration of the interactions between land use change. This study proposed a scenario analysis framework that integrated four Shared Socioeconomic Pathways (SSPs) from CMIP6 with bespoke land-use scenarios (Inertial Development (IDS), Urban Priority (UDPS), Ecological Protection (EPPS), Farmland (FPPS)) to create 16 scenarios, allowing for more nuanced understanding potential trajectories. The results indicated (1) compared baseline period (2000-2020), in Central Yunnan Agglomeration (CYUA) would improve, albeit significant differences among scenarios. most notable improvement was under EPPS + SSP1-2.6 (ScC1). (2) lower Jinsha, upper Nanpan, Red river basins were high-risk areas CYUA, each dominated by different factors, necessitating differentiated control measures. (3) Land-use jointly influenced direction development, lightest occurring heaviest FPPS. largest decrease occurs SSP1-2.6, whereas smallest SSP5-8.5. (4) Climate had impact than change, reduction rates modulus area relative past 20 years being 9% 3.77%, respectively, approximately eight times magnitude recommends reducing carbon emissions, enhancing vegetation cover, controlling slope development effectively mitigate risk CYUA promote regional method provides new insights into global small-scale predictions.

Язык: Английский

---

Water erosion in soybean farming areas under different soil covers and simulated rainfall patterns in the Southern Amazon, Brazil

CATENA, Год журнала: 2025, Номер 254, С. 108959 - 108959

Опубликована: Март 23, 2025

Язык: Английский

---

Advanced Modeling of Forest Fire Susceptibility and Sensitivity Analysis Using Hyperparameter-Tuned Deep Learning Techniques in the Rajouri District, Jammu and Kashmir

Advances in Space Research, Год журнала: 2025, Номер unknown

Опубликована: Май 1, 2025

Язык: Английский

---

Quantitative assessment of morphometry and GIS integrated RUSLE model-based soil loss estimation from Pahuj river basin, central India

DELETED, Год журнала: 2024, Номер 90(4), С. 1049 - 1066

Опубликована: Май 10, 2024

In this study the morphometric indices of Pahuj river basin (PRB) were evaluated by applying remote sensing and GIS. The Shuttle Radar Topographic Mission (SRTM) based 30 m digital elevation (DEM) data was used in order to extract parameters using standard methods. PRB covering an area (3648 km2) is controlled homogenous lithology geological structures. drainage density indicates that permeable soil with coarse texture dominantly occurring large low-lying flat areas basin. Contrary high gradient consist impermeable hard granitic rocks Neoarchean Precambrian age a low quantity soil. value elongation ratio form factor reveal elongated show peak flows. To assess erosion susceptibility, attributes Revised Universal Soil Loss Equation (RUSLE) model integrated GIS estimate loss from results rainfall erosivity (R-factor) along pattern indicate upper catchment relatively exhibits intensity than middle lower region. findings (R), erodibility (K), topographic (LS), crop management (C) factors infer quite area. ruggedness number Melton (4.16) imply moderately rugged less prone erosion, particularly relief effective practices water conservation will enhance storage capacity prevent sediment PRB. research may be helpful resolve crisis can such drought-prone

Язык: Английский

---

Assessment of soil erosion risk and vulnerability in the transboundary Sio-Malaba-Malakisi watershed in Kenya and Uganda

Journal of Environmental Management, Год журнала: 2024, Номер 370, С. 122916 - 122916

Опубликована: Окт. 15, 2024

Язык: Английский

---