Machine learning models for gully erosion susceptibility assessment in the Tensift catchment, Haouz Plain, Morocco for sustainable development

Journal of African Earth Sciences, Год журнала: 2024, Номер 213, С. 105229 - 105229

Опубликована: Март 11, 2024

Gully erosion is a widespread environmental danger, threatening global socio-economic stability and sustainable development. This study comprehensively applied seven machine learning (ML) models including SVM, KNN, RF, XGBoost, ANN, DT, LR, evaluated gully susceptibility in the Tensift catchment predict it within Haouz plain, Morocco. To ensure reliability of findings, employed robust combination inventory, sentinel images, Digital Surface Model. Eighteen predictors, encompassing topographical, geomorphological, environmental, hydrological factors, were selected after multicollinearity analyses. The revealed that approximately 28.18% at very high risk erosion. Furthermore, 15.13% 31.28% are categorized as low respectively. These findings extend to where 7.84% surface area highly risking erosion, while 18.25% 55.18% characterized areas. gauge performance ML models, an array metrics specificity, precision, sensitivity, accuracy employed. highlights XGBoost KNN most promising achieving AUC ROC values 0.96 0.93 test phase. remaining namely RF (AUC = 0.89), LR 0.80), SVM 0.81), DT 0.86), ANN 0.78), also displayed commendable performance. novelty this research its innovative approach combat through cutting edge offering practical solutions for watershed conservation, management, prevention land degradation. insights invaluable addressing challenges posed by region, beyond geographical boundaries can be used defining appropriate mitigation strategies local national scale.

Язык: Английский

---

Random Forest–based gully erosion susceptibility assessment across different agro-ecologies of the Upper Blue Nile basin, Ethiopia

Geomorphology, Год журнала: 2023, Номер 431, С. 108671 - 108671

Опубликована: Март 27, 2023

Several environmental factors are known to influence the spatial distribution and susceptibility of gully erosion, yet relative importance interaction these remain little understood in Ethiopia. In this study, we integrated detailed field investigations with high-resolution remote sensing products assess erosion identify its controlling using Random Forest (RF) model six representative watersheds across contrasting (highland, midland, lowland) agro-ecological environments Upper Blue Nile basin Data for 20 were extracted from datasets at eight different pixel resolutions ranging 0.5 30 m a geographic information system environment. About 70 % dataset each watershed randomly selected training validation purposes, respectively. Multicollinearity correlation analyses performed variables collinearity problems explain their statistical relationships among other variables. RF predicted factors. The showed outstanding performance when finest-resolution used. Elevation, height above nearest drainage, runoff curve number-II, distance streams, drainage density, soil type, land use/land cover found be most important gullies all watersheds, irrespective treatment conditions settings. Thus, susceptible was low-lying grazing cultivated lands sensitive high runoff-generation capacity located within short horizontal vertical distances networks. Therefore, basin- watershed-scale management strategies should give priority areas. identification hydrologic parameter predicting direct excess rainfall, as one novel finding which will useful developing improved process-based models.

Язык: Английский

---

Understanding land degradation induced by gully erosion from the perspective of different geoenvironmental factors

Journal of Environmental Management, Год журнала: 2022, Номер 315, С. 115181 - 115181

Опубликована: Апрель 29, 2022

Язык: Английский

---

Determining the most appropriate drought index using the random forest algorithm with an emphasis on agricultural drought

Natural Hazards, Год журнала: 2022, Номер 115(1), С. 923 - 946

Опубликована: Сен. 2, 2022

Язык: Английский

---

Quantitative analysis of the impact of climate change and oasification on changes in net primary productivity variation in mid-Tianshan Mountains from 2001 to 2020

Ecological Indicators, Год журнала: 2023, Номер 154, С. 110820 - 110820

Опубликована: Авг. 21, 2023

Net primary productivity (NPP) has been substantially changed under the intense oasification in urban agglomerations on northern slopes of mid-Tianshan Mountain (UANSTM) and climate change. However, temporal variations NPP remain unclear, relative contribution change annual variation is still debate. By using remote sensing data, reanalysis modified Carnegie–Ames-Stanford Approach (CASA) model, a machine learning method, we explored spatial–temporal UANSTM region quantified to from 2001 2020. Our study indicated that: (1) presents an overall increasing trend most presented decreasing mainly due cropland conversion area; (2) oasification-dominated area concentrated built-up land cropland; (3) during 2001–2020, increased by about 5.4 Tg·C, climatic increase were (73.1% 26.9%, respectively); (4) water-related factors was main driver region.

Язык: Английский

---

Gully Erosion Susceptibility Assessment in the Kondoran Watershed Using Machine Learning Algorithms and the Boruta Feature Selection

Sustainability, Год журнала: 2021, Номер 13(18), С. 10110 - 10110

Опубликована: Сен. 9, 2021

Gully erosion susceptibility mapping is an essential land management tool to reduce soil damages. This study investigates gully based on multiple diagnostic analysis, support vector machine and random forest algorithms, also a combination of these models, namely the ensemble model. Thus, map in Kondoran watershed Iran was generated by applying models occurrence non-occurrence points (as target variable) several predictors (slope, aspect, elevation, topographic wetness index, drainage density, plan curvature, distance streams, lithology, texture use). The Boruta algorithm used select most effective variables modeling susceptibility. area under receiver operating characteristic curve (AUC), characteristics, true skill statistics (TSS) were assess model performance. results indicated that had best performance (AUC = 0.982, TSS 0.93) compared others. factors region topological, anthropogenic, geological. methodology this can be other regions control mitigate phenomenon biophilic regenerative techniques at locations influential factors.

Язык: Английский

---

Analysis of gully erosion susceptibility and spatial modelling using a GIS-based approach

Geoderma, Год журнала: 2022, Номер 420, С. 115869 - 115869

Опубликована: Апрель 13, 2022

Язык: Английский

---

Soil erosion prediction using Markov and CA-Markov chains methods and remote sensing drought indicators

Ecological Informatics, Год журнала: 2023, Номер 78, С. 102386 - 102386

Опубликована: Ноя. 25, 2023

Язык: Английский

---

Influence of geological conditions on gully distribution in the Dry–hot Valley, SW China

CATENA, Год журнала: 2022, Номер 214, С. 106274 - 106274

Опубликована: Апрель 6, 2022

Язык: Английский

---

Potential risk of soil erosion on the Tibetan Plateau during 1990–2020: Impact of climate change and human activities

Ecological Indicators, Год журнала: 2023, Номер 154, С. 110669 - 110669

Опубликована: Июль 21, 2023

With global warming and increasing anthropogenic activities, the ecosystems on Tibetan Plateau are becoming increasingly fragile, exacerbating risk of soil erosion in area. However, interacting effects induced by numerous processes occurring at different times spaces challenging to quantify using conventional single-process assessment models (e.g., water erosion, wind freeze–thaw). Consequently, our understanding complicated state plateau with respect potential for is limited. Therefore, we created a methodological framework multi-criteria decision-making (MCDM) evaluate various driven climate change human activity under current topographical circumstances this region. The results showed that model was reliable, estimated accuracy receiver operating characteristic curve training validation datasets 0.721. majority areas (60.69%) were very-low or low-risk levels, while 17.55% (mainly southeast along surrounding high mountains) risk. In general, average significantly increased during 1990–2020, exerted more pressure land surface than activities did. dramatically 28.15% total area, found be concentrated plateau's southern, eastern, central, northern regions. findings study provide basis local ecological environmental protection resource management propose new protocol forecast prevent worldwide.

Язык: Английский

---