Search

Rainfall-induced landslide masses often change into disastrous debris flows and damage large areas in Nepal's mountainous region. The area covered by debris flow inundation or debris flow runout is a most essential component for landslide hazard assessments leading to development of land use plans. However, there is presently no tool that can allow to assess and predict debris flow runout from the initial landslide on the watershed scale in Nepal. In this paper, a method and simulation tool are developed for the assessment and prediction of debris flow runout in Kulekhani Watershed, Nepal. The Flow-R model with user-defined landslide-susceptible areas was chosen for debris flow runout analysis. The Flow-R model with various algorithms has the capability to analyze debris flow inundation with limited input information. Two recent debris flow events are taken as case studies to identify the appropriate algorithms for runout analysis of the study area. After comparison of the observed and simulated results for debris flow runout, the algorithms proposed by Holmgren (Hydrol Process 8:327–334, 1994) (modified) are found suitable for the study watershed. These algorithms are employed for debris flow inundation analysis in the study area with pre-defined landslide sources plus debris flow inundation map in GIS environment. The results obtained from this modeling for the debris flow area induced by 540 mm of rainfall in 24 h period was 2.68% of the watershed, which is comparable to previously observed debris flow area in the study watershed. The findings of this research and GIS simulation tool developed will enable the optimization of planning and investment for land development in the studied region. [ABSTRACT FROM AUTHOR]

Landslides cause extensive loss of life and property in the Nepal Himalaya. Since the late 1980s, different mathematical models have been developed and applied for landslide susceptibility mapping and hazard assessment in Nepal. The main goal of this paper is to apply fuzzy logic to landslide susceptibility mapping in the Ghurmi-Dhad Khola area, Eastern Nepal. Seven causative factors are considered: slope angle, slope aspect, distance from drainage, land use, geology, distance from faults and folds, soil and rock type. Likelihood ratios are obtained for each class of causative factors by comparison with past landslide occurrences. The ratios are normalized between zero and one to obtain fuzzy membership values. Further, different fuzzy operators are applied to generate landslide susceptibility maps. Comparison with the landslide inventory map reveals that the fuzzy gamma operator with a γ-value of 0.60 yields the best prediction accuracy. Consequently, this operator is used to produce the final landslide susceptibility zonation map. [ABSTRACT FROM AUTHOR]