Landslide Susceptibility Assessment Using Spatial Multi-Criteria Evaluation Model in Rwanda

Abstract

Landslides susceptibility assessment has to be conducted to identify prone areas and guide risk management. Landslides in Rwanda are very deadly disasters. The current research aimed to conduct landslide susceptibility assessment by applying Spatial Multi-Criteria Evaluation Model with eight layers of causal factors including: slope, distance to roads, lithology, precipitation, soil texture, soil depth, altitude and land cover. In total, 980 past landslide locations were mapped. The relationship between landslide factors and inventory map was calculated using the Spatial Multi-Criteria Evaluation. The results revealed that susceptibility is spatially distributed countrywide with 42.3% of the region classified from moderate to very high susceptibility, and this is inhabited by 49.3% of the total population. In addition, Provinces with high to very high susceptibility are West, North and South (40.4%, 22.8% and 21.5%, respectively). Subsequently, the Eastern Province becomes the peak under low susceptibility category (87.8%) with no very high susceptibility (0%). Based on these findings, the employed model produced accurate and reliable outcome in terms of susceptibility, since 49.5% of past landslides fell within the very high susceptibility category, which confirms the model’s performance. The outcomes of this study will be useful for future initiatives related to landslide risk reduction and management.

Comparative Analysis of Deterministic and Semiquantitative Approaches for Shallow Landslide Risk Modeling in Rwanda

“…The findings of the present study are in agreement with the results from previous studies (MIDIMAR, 2015; Nahayo et al., ; Nsengiyumva et al., ; Piller, ) that confirmed the western part of Rwanda as a landslide‐prone zone. Furthermore, all 10 districts are alarmingly susceptible to shallow landslide hazards.…”

“…Due to its steep and hilly slopes, the region is prone to various natural hazard risks, including shallow landslides, which are the most devastating. They are the most recurrent and dominant hazards in the subregion (Ministry of Disaster Management and Refugees [MIDIMAR], 2018;Nsengiyumva, Luo, Nahayo, Huang, & Cai, 2018b). These hazards cause massive destructions and losses of human lives, properties, infrastructure, and environment.…”

“…Furthermore, as reported by Rwanda's Ministry in Charge of Emergency Management, from 2011 to May 2015, 124 people were killed by landslides, 141 injured, and 897 houses destroyed in the study area (MIDIMAR, 2018). In May 2016 itself, landslides killed 35 people, and 26 were injured while 67 roads and 29 bridges were completely destroyed (Nsengiyumva et al, 2018b). Similarly, from January to mid October 2018, natural hazards including landslides killed 234 people, injured 218, destroyed 15,264 houses and 9,412 hectares of crops, damaged 31 roads and 52 bridges, completely destroyed 86 classrooms, and killed 797 livestock (MIDIMAR, 2018).…”

Estimating landslides vulnerability in Rwanda using analytic hierarchy process and geographic information system

“…The occurrence of landslides in mountainous regions is reported to be rainfall induced (Wenning et al 2017). Also, in Rwanda, more than 70% of the damages and losses from landslides are rainfall induced (Nsengiyumva et al 2018). Similarly, districts of western Rwanda with high monthly rainfall ( Figure 2H) are moderately, highly, and very highly vulnerable to landslides (Table 4, Figure 4).…”

“…These hazards affected millions of people (dead, injured, or homeless) and damaged hectares of croplands and infrastructures such as roads and bridges (Piller 2016;Wagesho and Claire 2016;Nsengiyumva et al 2018). The record differs by area and landslides are largely registered in the northwestern district (Nsengiyumva et al 2018). Specifically, high population density and poverty, high elevation, slope, and frequent torrential rainfall are the major landslides vulnerability drivers in western Rwanda (Dawson and Martin 2015;Piller 2016).…”