Landslides in the central Kenyan highlands : Risks and actors

Sammanfattning: Measures taken to prevent and mitigate landslides are often based on geotechnical solutions. However, to understand local effects of landslides, people’s perception, knowledge and experiences are important, as this influence the level of vulnerability, and the efficacy of adaptive and mitigating measures. This thesis documents perception, knowledge and experiences of farmers in central Kenya, with the overall objective to analyse people’s vulnerability to landslide events. Semi-structured interviews were conducted with farmers and authority representatives. The results show varying levels of knowledge about landslides among farmers, and varying perceptions concerning hazard-exposure. Triggering factors were generally known, but many farmers stated that landslides cannot be prevented because of lack of resources, and there is a general lack of confidence in advice given by local agricultural authorities. Some farmers regarded landslides as life-threatening hazards, while others meant that landslides were subordinate to other, more pressing issues. People with previous experience from landslides on their land claimed that they were financially affected for at least five years. Local agricultural extension officers, on the other hand, did not view landslide as an issue. Thus, although there is direct communication between famers and authorities, lack of a mutual problem identification is evident. Problems are defined by educated extension officers, rather than by the people potentially exposed to the problems. This inequality is also reflected on family level, between men and women, in spite of women being physically more present on the land than men, and therefore potentially more at risk for landslides. According to risk theory, people in the area are vulnerable, but net benefits create opportunities they would not have if they left their land. In addition, lack of economic resources engenders a need to prioritise between problems. Therefore, although conveying information about how to mitigate landslide hazards to farmers is important, issues of socio-economic character must also be addressed. There is no direct correlation between farmers’ perception and knowledge about landslides and the way they act, and external initiatives aimed at mitigating the effects of landslides must therefore include empowerment of farmers. Mapping and analysing landslide susceptibility using GIS may help with hazard preparedness and mitigation. Therefore, the thesis also aims to test the applicability of GIS-analysis for landslide risk assessment, using available data on topography, geology, soil, land use, hydrology and precipitation. Areas susceptible to landsliding could not be identified, mainly because of inadequate data availability and quality. A DEM with higher resolution, a more detailed basis for land use changes analysis and more information concerning historical landslides would be needed.