A new approach to landslide assessment using Depth-Number fractal model

Aims: Landslide classification using a fractal model at the Tajan river basin in northern Iran is the study intended as a new approach based on 142 landslide information data sets. Materials Methods: The obtained results were interpreted using the Depth–Number (Dp–N) fractal model and a relatively broad set of information available for each landslide class, consisting of the Digital Elevation Model (DEM), rainfall, land-use, geology (lithology and fault) and drainage network data. Findings: The log–log plot shows five classes for depth (weakly, moderately, highly, strongly, and extremely magnitude) which shows that the extremely magnitude landslides have depths higher than 19.95 m in the NE, middle, western, and SE parts of the Tajan basin. The strong (5-19.95 m) and high (2.4-5 m) magnitude landslides happened in the northern, NE, western, and NW parts. The results, which were matched up to land-use, drainage network, DEM, and fault allocation patterns, revealed an affirmative correlation between landslide classes and the particulars in the area. In addition, the coefficient of determination, R2, for each population shows that the classification has been done correctly using the Dp-N fractal model. Amounts of P-value obtained from paired samples t-test and ANOVA showed that the separate categories are incongruous and significantly different (sig=0.000). Conclusion: Results show that separating the populations of landslides based on a parameter as magnitude and the difference between the populations’ magnitude of landslides should be considered in landslide susceptibility zonation.