Predicting gully initiation: comparing data mining techniques, analytical hierarchy processes and the topographic threshold

Abstract

Predicting gully initiation at catchment scale was done previously by integrating a geographical information system (GIS) with physically based models, statistical procedures or with knowledge‐based expert systems. However, the reliability and validity of applying these procedures are still questionable. In this work, a data mining (DM) procedure based on decision trees was applied to identify areas of gully initiation risk. Performance was compared with the analytic hierarchy process (AHP) expert system and with the commonly used topographic threshold (TT) technique. A spatial database was used to test the models, composed of a target variable (presence or absence of initial points) and ten independent environmental, climatic and human‐induced variables. The following findings emerged: using the same input layers, DM provided better predictive ability of gully initiation points than the application of both AHP and TT. The main difference between DM and TT was the very high overestimation inherent in TT. In addition, the minimum slope observed for soil detachment was 2°, whereas in other studies it is 3°. This could be explained by soil resistance, which is substantially lower in agricultural fields, while most studies test unploughed soil. Finally, rainfall intensity events >62.2 mm h‐1 (for a period of 30 min) were found to have a significant effect on gully initiation. Copyright © 2012 John Wiley & Sons, Ltd.

Application of Decision-Tree Model to Groundwater Productivity-Potential Mapping

“…Additionally, the leaf holds a probability vector indicating the probability of the feature's indicating a groundwater-productive area. New points are classified by navigating from the root of the tree to a leaf according to the outcome of tests along the path [27]. Figure 4 shows the general structure of a decision tree, which consists of three elements: node, condition, and production.…”

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

“…Rainfall temporal distribution influences runoff hydraulics and soil moisture: the former regulates flow erosivity while the latter before rainfall events influences both generation of runoff (Descroix et al, 2002;Castillo et al, 2003;Capra et al, 2012) and soil resistance to erosion (Bocco, 1991;Nachtergaele et al, 2002;Poesen et al, 2003;Torri et al, 2006;Bou Kheir et al, 2007;Magliulo, 2012;Torri et al, 2012b). Volume and velocity of concentrated flow are controlled by topographic attributes such as contributing drainage area, slope curvature and slope steepness (Moore et al, 1988;Vandekerckhove et al, 1998Vandekerckhove et al, , 2000aPoesen et al, 2003;Valentin et al, 2005;Zucca et al, 2006;Gómez Gutiérrez et al, 2009a, b;Kakembo et al, 2009;Nazari Samani et al, 2009;Capra et al, 2012;Svoray et al, 2012;Chaplot, 2013;Conoscenti et al, 2013). Depth and cross-sectional morphology of gullies are regulated by erodibility of soil horizons (Ireland et al, 1939;Imeson and Kwaad, 1980;Poesen et al, 2003) and characteristics of geological substrata (Vandekerckhove et al, 2000a;Zucca et al, 2006;Conoscenti et al, 2008;Conforti et al, 2010;Lucà et al, 2011;Marzolff et al, 2011;El Maaoui et al, 2012).…”

“…In this sense, an important contribution is given by several investigations focusing on the assessment of a topographic threshold that has to be exceeded for the initiation of a gully (Montgomery and Dietrich, 1992;Desmet et al, 1999;Kakembo et al, 2009;Nazari Samani et al, 2009;Svoray et al, 2012). Moreover, a suitable choice for predicting the location of gullies is the adoption of an inferential approach that allows an investigator to assess the spatial probability of gully occurrence within a given area, on the basis of statistical relationships established between environmental controlling variables and the spatial distribution of gullies.…”

Application of Data Mining Using Artificial Neural Network: Survey

“…For the identification and risk of gully initiation a procedure known as Data mining which is based on decision trees is applied. In this article the comparison of DM technique is shown with many other procedures like expert system and topographic threshold method (TT) [42]. The results show that DM technique provides more accurate data than that of the other methods.…”

“…Because there are always assumption in all modeling approaches. So with the help of simpler model give us a good result just like complex one In this article [42] the author is dealing with the prediction of gully initiation. In past predicting gully initiation was prepared with the help of GIF scheme with knowledge base expert system, physical based system or statistical procedures.…”