ORIGINAL ARTICLE Determination of importance for comprehensive topographic factors on landslide hazard mapping using artiﬁcial neural network Mutasem Sh. Alkhasawneh • Umi Kalthum Ngah • Lea Tien Tay • Nor Ashidi Mat Isa Received: 30 March 2013 / Accepted: 8 December 2013 Ó Springer-Verlag Berlin Heidelberg 2013 Abstract A landslide is one of the natural disasters that occur in Malaysia. In addition to the geological factor and the rain as triggering factor, topographic factors such as elevation, slope angle, slope aspect, and curvature are considered as the main causes of landslides. The study in this paper was conducted in three stages. The ﬁrst stage involved the extraction of extra topographic factors. Pre- vious landslide studies had identiﬁed only four of the topographic factors. However, eight new additional factors have also been identiﬁed in this study. They are general curvature, longitudinal curvature, tangential curvature, cross-section curvature, surface area, diagonal line length, surface roughness, and rugosity. At this stage, 13 factors were extracted from the digital elevation model. The sec- ond stage involved specifying the importance of each factor. The multilayer perceptron network and backpropa- gation algorithm were used to specify the weight of each factor. Results were veriﬁed using the receiver operating characteristics based on the area under the curve method in the third stage. The results indicated 76.07 % accuracy in predicting of landslides, with slope angle as the most important factor while the tangential curvature has the least importance. Keywords Landslide  Topographic factors  Neural networks  Weightage Introduction Landslides are among the most aggressive natural disasters that have caused the loss of lives and billions of dollars in damage worldwide annually. A landslide is a natural haz- ard that can occur within a speciﬁed period of time and a given area (Varnes 1984). It has become a subject of interest for studies; a considerable amount of research works have been conducted over the past years to identify the most important factors that cause landslides (Alkha- sawneh et al. 2012). However, different factors such as geological, topo- graphic, physical, and human causes (human disregard for sustainable developments) contribute to landslide occur- rence (Varnes 1984; Hutchinson 1995). The literature review of landslide-causing factors shows that topographic factors are linked strongly with landslide occurrence (Huot et al. 2003; Ercanoglu and Gokceoglu 2004; Ermini et al. 2005). This paper mainly focuses on the topographic fac- tors. As geological factors and rain perception are impor- tant factors which cause landslide, they are also added to improve the prediction of the MLP. The topographic factors can be divided into 13 factors: elevation, slope angle, slope aspects, plan, proﬁle, cross- section, tangential, longitude and general curvature, diagonal line lengths, surface area, roughness, and rug- osity. Recently, the digital elevation model (DEM) has been used as the basic data source to extract topographic data and is also one of the core database sources for several geographic information system (GIS) applications (Zhou and Liu 2004). Studies have been conducted on landslide susceptibility evaluation using DEM. A land- slide risk hazard map for Penang Island was produced using MLP neural network. Five training sites from Penang Island and nine different factors were involved in M. Sh. Alkhasawneh (&)  U. K. Ngah  L. T. Tay  N. A. M. Isa Imaging and Computational Intelligence (ICI) Group, School of Electrical and Electronic Engineering, Universiti Sains Malaysia, Engineering Campus, 14300 Nibong Tebal, Penang, Malaysia e-mail: m_sh_ka1@yahoo.com 123 Environ Earth Sci DOI 10.1007/s12665-013-3003-x