GIS-based landslide susceptibility mapping using heuristic and bivariate statistical methods for Iva Valley and environs
- PDF / 7,401,422 Bytes
- 19 Pages / 547.087 x 737.008 pts Page_size
- 55 Downloads / 213 Views
GIS-based landslide susceptibility mapping using heuristic and bivariate statistical methods for Iva Valley and environs Southeast Nigeria O. H. Ozioko & O. Igwe
Received: 7 May 2019 / Accepted: 4 November 2019 # Springer Nature Switzerland AG 2020
Abstract Udi-Iva Valley region of Enugu state has the most concentration of landslide in Southeastern Nigeria. Detailed field investigations alongside satellite image studies were employed to delineate nine landslide conditioning factors. Lithology, elevation, slope, aspect, curvature, distance from drainage, distance from road, land cover, and distance from lineament have been chosen as the landslide causative factors in the study area. This study presents a susceptibility mapping of landslides involving a combined bivariate statistical: frequency ratio (FR) and heuristic analytical hierarchy process (AHP) approach integrated in GIS environment. Validation and cross-validation of the susceptibility maps thus produced was achieved with the aid of landslide density approach in combination with prediction rate curve to check for the uniformity in the class areas in the susceptibility model produced. The analytical hierarchy process (AHP) produced results in which the lithology and slope factors had highest weights of 0.17 and 0.14 respectively. A strong correlation was observed in the lithology and slope conditioning factors; this is evident in the results of the FR approach with 10.68 and 6.86 FR values respectively. The landslide susceptibility maps were classified into five classes: very low susceptibility, low susceptibility, medium susceptibility, high susceptibility and very high susceptibility. Prediction rate curve was used to assess the predictive potential of the landslide susceptibility models, the result showed O. H. Ozioko (*) : O. Igwe Department of Geology, University of Nigeria, Nsukka, Nigeria e-mail: [email protected]
area under curve values of 70.49% for AHP and 72.09% for FR method. The similarity in the landslide density distribution in the susceptibility class, indicates a correlation between the generated susceptibility model and field observations. The statistical and heuristic methods employed have produced positive results; this was confirmed by the fact that all the 300 landslides were found to have occurred within the high susceptibility and very high susceptibility zones respectively. Keywords AHP . GIS . Landslide prediction . Frequency ratio . Susceptibility . Validation
Introduction Landslides are one of the most common geoenvironmental disasters in the world that is responsible for the death of hundreds of people and billions of dollars’ worth in property damage annually. The major factors that contribute to landslides occurrence have been studied extensively. Varnes (1984) and Hutchinson (1995) found out that the most important inherent factors include geology, geomorphology, soil, land use land cover, and hydrologic conditions. The Iva Valley region (Fig. 1) has the highest density of landslide clusters in Southeastern Nigeria.
Data Loading...
