Landslide susceptibility and influencing factors analysis in Rwanda
- PDF / 4,756,064 Bytes
- 28 Pages / 439.37 x 666.142 pts Page_size
- 86 Downloads / 376 Views
Landslide susceptibility and influencing factors analysis in Rwanda Richard Mind’je1,2,3,4 · Lanhai Li1,2,3,5,6 · Jean Baptiste Nsengiyumva1,3 · Christophe Mupenzi4 · Enan Muhire Nyesheja4 · Patient Mindje Kayumba3,4 · Aboubakar Gasirabo1,3,4 · Egide Hakorimana1,3,4 Received: 17 May 2019 / Accepted: 7 December 2019 © Springer Nature B.V. 2019
Abstract Rwanda as a landlocked country has been recurrently facing tremendous and devastating landslides having serious impacts on the environment and socioeconomic development. Analyzing landslide susceptibility and its influencing factors could ease landslide disaster control and, hence, contributes to environmental sustainability. Unfortunately, detailed information on influencing factors per respective classes and their level of correlation to the probability of landslide occurrence have not been fully analyzed in Rwanda. The purpose of this study is to reveal the spatial correlation between landslide occurrence and different classes of influencing factors using the frequency ratio (FR) approach with geographical information system (GIS) and remote sensing techniques. Initially, a landslide inventory map was prepared using 423 landslide locations that were randomly split into 75% of training datasets (318 points) and 25% (105 points) to validate the model. A multicollinearity analysis was performed among ten influencing factors using the tolerance and variance inflation factor method. These factors include elevation, slope, aspect, distance to roads, distance to rivers, the normalized difference vegetation index, land use land cover, stream power index, rainfall and soil texture. The analysis revealed no multicollinearity among these factors, and therefore, all of them were suitable for modeling process. The FR model evaluated the relationship between landslide incidence and influencing factors in their respective classes and then generated the landslide susceptibility map using GIS which revealed the western, northern and some parts of southern province as the most susceptible areas owing to the slope, LCLU, rainfall and elevation as the main influencing factors. The generated susceptibility map was validated using the area under curve which portrayed 81.2% and 84.6% for success and prediction rate, respectively. To conclude, the results of this study are essential for future initiatives regarding landslide risk reduction toward sustainability in Rwanda. However, these can be broad and may reflect the most landslide influencing factors projected in other central-east African regions. Keywords Disaster · Influencing factors · Landslide · Rwanda · Susceptibility
* Lanhai Li [email protected] Extended author information available on the last page of the article
13
Vol.:(0123456789)
R. Mind’je et al.
1 Introduction Disaster risk management as an issue at stake worldwide shifts its emphasis from post-disaster to pre-disaster phases (Baas et al. 2008). Essential management activities in pre-disaster phases like risk assessment, hazard identification, preparedness and
Data Loading...
