Identification of Landslide-Prone Areas and Mitigation Measures along the Proposed Chaumala–Garbha Road, Kailali, Nepal

Abstract

This study identifies landslide-prone sections and appropriate mitigation measures along the proposed Chaumala–Garbha Road, a 22.317 km alignment (Chainage 0+000–22+317) in Kailali and Doti districts, Sudurpashchim Province, Nepal. Primary data (field observation, GPS survey, landslide inventory and photographic documentation) and secondary data (DEM, geology, land use/land cover and rainfall) were compiled at 500 m chainage intervals (46 stations) within a 1 km road-corridor buffer. A GIS-based Analytical Hierarchy Process (AHP) integrating eight conditioning factors – slope, geology, rainfall, land use/land cover, drainage, road proximity, aspect and elevation – was used to derive factor weights (slope 31.2%, geology 20.1%, rainfall 14.6%, land use/land cover and road proximity 10.4% each, drainage 6.8%, elevation 3.9%, aspect 2.6%) with a consistency ratio of 1.3%, well within Saaty's acceptable limit. The resulting Landslide Susceptibility Index classified 11.9 km (53.6%) of the corridor as Moderate, 9.7 km (43.4%) as High and 0.7 km (3.0%) as Low susceptibility, with no, very low or very high sections. The longest continuous high-risk stretches occur at chainage 4+550–10+700, 11+450–13+400 and 17+250–18+600. Receiver Operating Characteristic (ROC) validation returned an Area Under the Curve of 0.81, indicating good predictive accuracy. Field investigation attributed recurring failures mainly to inadequate drainage, haphazard road excavation, over-steepened cut slopes (70–80° against the 34–45° recommended for sandy soil), absence of stabilization structures and deforestation, corroborated by a focus group discussion and an in-depth interview with local stakeholders. Structural measures (retaining walls, breast walls, gabion walls and rock bolting) are recommended for high-susceptibility sections, and non-structural measures (surface drains, slope trimming and bioengineering) for moderate- and low-susceptibility sections. The study demonstrates that a GIS-based AHP framework provides a validated, corridor-level basis for prioritizing landslide mitigation investment on rural mountain roads.