Mapping erosion susceptibility by a multivariate statistical method: A case study from the Ayvalık region, NW Turkey

Akgün, Aykut; Türk, Necdet

doi:10.1016/j.cageo.2010.09.006

Cited by 126 publications

(43 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…An ROC curve plots true positive rate TP (sensitivity) against false positive rate FP (1 − specificity) for all possible cut-off values; sensitivity is computed as the fraction of cells hosting gullies that were correctly classified as susceptible, while specificity is derived from the fraction of cells free of gullies that were correctly classified as not-susceptible. The closer the ROC curve to the upper left corner (AUC = 1), the higher the predictive performance of the model; a perfect discrimination between positive and negative cases produces an AUC value equal to 1, while a value close to 0.5 indicates inaccuracy in the model (Fawcett, 2006;Akgün and Türk, 2011). In relation to the computed AUC value, Hosmer and Lemeshow (2000) classify a predictive performance as acceptable (AUC N 0.7), excellent (AUC N 0.8) or outstanding (AUC N 0.9).…”

Section: Logistic Regression Analysis and Model Accuracy Evaluationmentioning

confidence: 99%

“…Such a method was adopted by Meyer and Martínez-Casasnovas (1999), who predicted the occurrence of gullies in vineyards of north-east Spain by using sub-basins as elementary sampling units. However, the increasing resolution of digital elevation models (DEMs), available at various scales, stimulated the investigators to adopt grid cells of the same size of the DEM pixels as mapping units for the spatial prediction of water erosion processes (see Bou Kheir et al, 2007;Conoscenti et al, 2008;Gómez Gutiérrez et al, 2009a,b;Conforti et al, 2010;Magliulo, 2010;Akgün and Türk, 2011;Lucà et al, 2011;Märker et al, 2011;Magliulo, 2012;Conoscenti et al, 2013). In this study, we analysed the susceptibility conditions to gully erosion by using, as mapping units, grid cell units (CLUs) and slope units (SLUs): the CLUs simply correspond to the 5 × 5 m pixels of the DEM available for the investigated area.…”

Section: Mapping Unitsmentioning

confidence: 99%

“…The experiment was carried out in a small basin of central-northern Sicily, which extends for 9.5 km 2 and is affected by intense erosion. Logistic regression was selected as the multivariate statistical method for assessing gully erosion susceptibility because of the following reasons: i) it can work with a variety of types of independent variables such as categorical, binary, ordinal or continuous; ii) it is free of data distribution constraints (Bai et al, 2010); iii) it is also robust when input data are auto-correlated (Ohlmacher and Davis, 2003;Mathew et al, 2009), as often happens when dealing with environmental attributes; and iv) only a few studies have tested logistic regression for assessing gully erosion susceptibility (Meyer and Martínez-Casasnovas, 1999;Akgün and Türk, 2011;Lucà et al, 2011) and further investigations are needed.…”

Section: Introductionmentioning

confidence: 99%

“…Hence, the calculated probability values can be used to generate maps of gully erosion susceptibility, showing how proneness to this phenomenon changes spatially. Different statistical methods have been recently adopted to predict spatial distribution of gully erosion on various spatial scales (see Table 1), from the field scale to the continental scale (Meyer and Martínez-Casasnovas, 1999;Hughes et al, 2001;Chaplot et al, 2005;Bou Kheir et al, 2007;Geissen et al, 2007;Conoscenti et al, 2008;Gómez Gutiérrez et al, 2009a;Conforti et al, 2010;Magliulo, 2010;Akgün and Türk, 2011;Lucà et al, 2011;Märker et al, 2011;Magliulo, 2012;Conoscenti et al, 2013).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Gully erosion susceptibility assessment by means of GIS-based logistic regression: A case of Sicily (Italy)

et al. 2014

View full text Add to dashboard Cite

This research aims at characterizing susceptibility conditions to gully erosion by means of GIS and multivariate statistical analysis. The study area is a 9.5 km 2 river catchment in central-northern Sicily, where agriculture activities are limited by intense erosion. By means of field surveys and interpretation of aerial images, we prepared a digital map of the spatial distribution of 260 gullies in the study area. In addition, from available thematic maps, a 5 m cell size digital elevation model and field checks, we derived 27 environmental attributes that describe the variability of lithology, land use, topography and road position. These attributes were selected for their potential influence on erosion processes, while the dependent variable was given by presence or absence of gullies within two different types of mapping units: 5 m grid cells and slope units (average size = 2.66 ha). The functional relationships between gully occurrence and the controlling factors were obtained from forward stepwise logistic regression to calculate the probability to host a gully for each mapping unit. In order to train and test the predictive models, three calibration and three validation subsets, of both grid cells and slope units, were randomly selected. Results of validation, based on ROC (receiving operating characteristic) curves, attest for acceptable to excellent accuracies of the models, showing better predictive skill and more stable performance of the susceptibility model based on grid cells.

show abstract

Section: Logistic Regression Analysis and Model Accuracy Evaluationmentioning

confidence: 99%

Section: Mapping Unitsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Gully erosion susceptibility assessment by means of GIS-based logistic regression: A case of Sicily (Italy)

et al. 2014

View full text Add to dashboard Cite

show abstract

“…A ROC curve plots true positive rate TP (sensitivity) against false positive rate FP (1−specificity), for all possible cutoff values; sensitivity is computed as the fraction of unstable cells that were correctly classified as susceptible, while specificity is derived from the fraction of stable cells that were correctly classified as nonsusceptible. The closer the ROC curve to the upper left corner (AUC=1), the higher the predictive performance of the model; a perfect discrimination between positive and negative cases produces an AUC value equal to 1, while a value close to 0.5 indicates inaccuracy in the model (Akgün and Türk 2011;Fawcett 2006;Nandi and Shakoor 2009;Reineking and Schröder 2006). In relation to the computed AUC value, Hosmer and Lemeshow (2000) classify a predictive performance as acceptable (AUC>0.7), excellent (AUC>0.8), or outstanding (AUC>0.9).…”

Section: Validationmentioning

confidence: 99%

Forward logistic regression for earth-flow landslide susceptibility assessment in the Platani river basin (southern Sicily, Italy)

et al. 2013

View full text Add to dashboard Cite

Forward logistic regression has allowed us to derive an earth-flow susceptibility model for the Tumarrano river basin, which was defined by modeling the statistical relationships between an archive of 760 events and a set of 20 predictors. For each landslide in the inventory, a landslide identification point (LIP) was automatically produced as corresponding to the highest point along the boundary of the landslide polygons, and unstable conditions were assigned to cells at a distance up to 8m. An equal number of stable cells (out of landslides) was then randomly extracted and appended to the LIPs to prepare the dataset for logistic regression. A model building strategy was applied to enlarge the area included in training the model and to verify the sensitivity of the regressed models with respect to the locations of the selected stable cells. A suite of 16 models was prepared by randomly extracting different unoverlapping stable cell subsets that have been appended to the unstable ones. Models were finally submitted to forward logistic regression and validated. The results showed satisfying and stable error rates (0.236 on average, with a standard deviation of 0.007) and areas under the receiver operating characteristic (ROC) curve (AUCs) (0.839 for training and 0.817 for test datasets) as well as factor selections (ranks and coefficients). As regards the predictors, steepness and large-profile and local-plan topographic curvatures were systematically selected. Clayey outcropping lithology, midslope drainage, local and midslope ridges, and canyon landforms were also very frequently (from eight to 15 times) included in the models by the forward selection procedures. The model-building strategy allowed us to produce a performing earth-flow susceptibility model, whose model fitting, prediction skill, and robustness were estimated on the basis of validation procedures, demonstrating the independence of the regressed model on the specific selection of the stable cells.

show abstract

Flow field analysis and particle erosion of tunnel‐slope systems under coupling between runoff and fast (slow) seepage

Zhang,

Song,

Zhang

et al. 2023

Deep Underground Science and Engineering

View full text Add to dashboard Cite

The presence of particles on the surface of a tunnel slope renders it susceptible to erosion by water flow, which is a major cause of soil and water loss. In this study, a nonlinear mathematical model and a mechanical equilibrium model are developed to investigate the distribution of flow fields and particle motion characteristics of tunnel slopes, respectively. The mathematical model of flow fields comprises three parts: a runoff region, a highly permeable soil layer, and a weakly permeable soil layer. The Navier‒Stokes equation controls fluid motion in the runoff region, while the Brinkman‐extended Darcy equation governs fast and slow seepage in the highly and weakly permeable soil layers, respectively. Analytical solutions are derived for the velocity profile and shear stress expression of the model flow field under the boundary condition of continuous transition of velocity and stress at the fluid‒solid interface. The shear stress distribution shows that the shear stress at the tunnel‐slope surface is the largest, followed by the shear stress of the soil interface, indicating that particles in these two locations are most vulnerable to erosion. A mechanical equilibrium model of sliding and rolling of single particles is established at the fluid‒solid interface, and the safety factor of particle motion (sliding and rolling) is derived. Sensitivity analysis shows that by increasing the runoff depth, slope angle, and soil permeability, the erosion of soil particles will be aggravated on the tunnel‐slope surface, but by increasing the particle diameter, particle‐specific gravity, and particle stacking angle, the erosion resistance ability of the tunnel‐slope surface particles will be enhanced. This study can serve as a reference for the analysis of surface soil and water loss in tunnel‐slope systems.

show abstract

Mapping erosion susceptibility by a multivariate statistical method: A case study from the Ayvalık region, NW Turkey

Cited by 126 publications

References 32 publications

Gully erosion susceptibility assessment by means of GIS-based logistic regression: A case of Sicily (Italy)

Gully erosion susceptibility assessment by means of GIS-based logistic regression: A case of Sicily (Italy)

Forward logistic regression for earth-flow landslide susceptibility assessment in the Platani river basin (southern Sicily, Italy)

Flow field analysis and particle erosion of tunnel‐slope systems under coupling between runoff and fast (slow) seepage

Contact Info

Product

Resources

About