Prediction of gully erosion susceptibility mapping using novel ensemble machine learning algorithms

Arabameri, Alireza; Pal, Subodh Chandra; Costache, Romulus; Saha, Asish; Rezaie, Fatemeh; Danesh, Amir Seyed; Pradhan, Biswajeet; Lee, Saro; Hoang, Nhat‐Duc

doi:10.1080/19475705.2021.1880977

Cited by 53 publications

(15 citation statements)

References 132 publications

(138 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GESM prediction accuracy has been greatly improved using ML algorithms compared with traditional statistical methods. Typical ML algorithms used in GESM, such as by Arabamari et al [19], explored a new model based on the genetic algorithm-extreme gradient boosting (GE-XGBoost). Their results showed that this model was of great significance for predicting large-scale gully erosion susceptibility maps.…”

Section: Introductionmentioning

confidence: 99%

Gully Erosion Susceptibility Mapping in Highly Complex Terrain Using Machine Learning Models

Yang

Wang

Pang

et al. 2021

IJGI

View full text Add to dashboard Cite

Gully erosion is the most severe type of water erosion and is a major land degradation process. Gully erosion susceptibility mapping (GESM)’s efficiency and interpretability remains a challenge, especially in complex terrain areas. In this study, a WoE-MLC model was used to solve the above problem, which combines machine learning classification algorithms and the statistical weight of evidence (WoE) model in the Loess Plateau. The three machine learning (ML) algorithms utilized in this research were random forest (RF), gradient boosted decision trees (GBDT), and extreme gradient boosting (XGBoost). The results showed that: (1) GESM were well predicted by combining both machine learning regression models and WoE-MLC models, with the area under the curve (AUC) values both greater than 0.92, and the latter was more computationally efficient and interpretable; (2) The XGBoost algorithm was more efficient in GESM than the other two algorithms, with the strongest generalization ability and best performance in avoiding overfitting (averaged AUC = 0.947), followed by the RF algorithm (averaged AUC = 0.944), and GBDT algorithm (averaged AUC = 0.938); and (3) slope gradient, land use, and altitude were the main factors for GESM. This study may provide a possible method for gully erosion susceptibility mapping at large scale.

show abstract

Section: Introductionmentioning

confidence: 99%

Gully Erosion Susceptibility Mapping in Highly Complex Terrain Using Machine Learning Models

Yang

Wang

Pang

et al. 2021

IJGI

View full text Add to dashboard Cite

show abstract

“…Thanks to the water-induced behavior, many environmental factors are ideal for the origin and growth of gullies. The gullies are usually categorized as permanent and ephemeral, in keeping with stability and occurrence (Foster, 1986). Permanent gullies are wide in existence and otherwise the ephemeral gullies are found during the wet season because of the large size of the runoff and its related water activity (Garosi et al, 2019).…”

Section: Soil Erosionmentioning

confidence: 99%

“…The implementation of artificial intelligence (AI) approaches in the geographic information system has frequently applied in recent times to estimate the rate of catchment erosion (Korb & Nicholson, 2010). In addition, these AI techniques permitted in use of network analysis to integrate the qualitative information (Arabameri et al, 2021;Martınez et al, 2003;Pal et al, 2020). This model conducts in nonlinear feature and partners each neuron linked to a consequent neighbor and it is functionality varying to improve the network for effective results.…”

Section: Introductionmentioning

confidence: 99%

Water-induced erosion potentiality and vulnerability assessment in Kangsabati river basin, eastern India

Chakrabortty

Pal

Arabameri

et al. 2021

Environ Dev Sustain

Self Cite

View full text Add to dashboard Cite

The large-scale water-induced erosion is one of the most determining elements on land degradation in subtropical monsoon-dominated region. From this large-scale erosion, the fertility of the agricultural land has been decline consequently. So, estimation of the amount of erosion and its accurate prediction is necessary for escaping from this hazardous situation. In this study, the application of evidential belief function (EBF), spatial logistic regression (SLR) and ensemble of EBF and SLR to estimate the erosion potentiality with the help of ArcGIS and Soil and water assessment tool (SWAT). The average annual soil erosion has been estimated with the help of revised universal soil loss equation (RUSLE) and geographical information system (GIS). Apart from this to evaluate the importance of morphotectonic parameters on soil erosion, the correlation between erosion potentiality and average annual soil erosion has been quantified. In large-scale erosion, there is a direct impact of storm rainfall event in monsoon period in the entire subtropical region. Here, in erosion potentiality assessment, the optimal capacity of ensemble EBF-SLR is higher than the single alone methods, i.e., EBF and SLR. So, the mentioned approaches can be applied in soil erosion research in subtropical environment with considering the erosion causal parameters. This type of information can be helpful to the decision-maker and stakeholders to take proper initiative to reducing the rate of erosion. The main task of the future researcher is to implement this method more accurate ways with considering more reliable variables and slight modifications of the approaches in keeping in the view of regional environment.

show abstract

“…Gradient Boosted Machine (GBM) has also been used in soil degradation/erosion susceptibility studies in different forms (Arabameri et al, 2020(Arabameri et al, , 2021Sahin 2020) and was first introduced by Friedman et al (2000). Gradient boosting is also an ensemble model, comprising of decision trees.…”

Section: Classification Algorithmsmentioning

confidence: 99%

Designing Grazing Susceptibility to Land Degradation Index (GSLDI) in Hilly Areas

Minea

Ciobotaru

Ioana-Toroimac

et al. 2022

Preprint

View full text Add to dashboard Cite

Evaluation of grazing impact on land degradation processes is a difficult task due to heterogeneity and complex interacting factors involved. In this research, we designed a new methodology based on a predictive index of GSLDI (Grazing Susceptibility to Land Degradation Index) built on artificial intelligence to assess land degradation susceptibility in areas affected by small ruminants (SRs) of sheep and goats grazing impacts. The data for model training, validation, and testing consisted of sampling points (erosion and no-erosion) taken from aerial imagery. Seventeen environmental factors (e.g., DEM derivatives, small ruminants’ stock), and 55 subsequent attributes (e.g., classes/features) were assigned to each sampling point. The impact of SRs stock density over the land degradation process has been evaluated and estimated with two extreme SRs’ density scenarios: missing (0), and double density (overstocking). We applied the GSLDI methodology on the Curvature Subcarpathians, a region that experiences the highest erosion rates in Romania, and found that SRs grazing is not the major contributor to land degradation, accounting for only 4.6%. This methodology could be replicated in other steep slopes grazing areas as a tool to assess and predict areas susceptible to land degradation, and to establish common strategies for good land-use practices.

show abstract

Prediction of gully erosion susceptibility mapping using novel ensemble machine learning algorithms

Cited by 53 publications

References 132 publications

Gully Erosion Susceptibility Mapping in Highly Complex Terrain Using Machine Learning Models

Gully Erosion Susceptibility Mapping in Highly Complex Terrain Using Machine Learning Models

Water-induced erosion potentiality and vulnerability assessment in Kangsabati river basin, eastern India

Designing Grazing Susceptibility to Land Degradation Index (GSLDI) in Hilly Areas

Contact Info

Product

Resources

About