University of Twente Student Theses

Login

Perceived tenure insecurity within deprivation: from a geospatial perspective

Dufitimana, Esaie (2021) Perceived tenure insecurity within deprivation: from a geospatial perspective.

[img] PDF
4MB
Abstract:Most cities in the Global South accommodate many people living in urban deprivation with a high level of perceived tenure insecurity due to the persisting threats of eviction and unwillingly loss of their land and other properties. There is an urgent need for up-to-date information to address this problem for promoting tenure security for all urban dwellers and make cities and human settlements safe and inclusive towards achieving the Sustainable Development Goals (SDGs). However, such information is often rare, outdated or incomplete. Though recent studies on urban deprivation have been applying earth observation as an alternative way to obtain timely information on urban deprivation, further application of earth observation to understand tenure insecurity has not been explored. In addition, the variation of perceived tenure insecurity within urban deprived areas and its relationship to spatial characteristics of these areas has not been investigated. Therefore, this research analyses the potential of earth observation-based information and other spatial information to measure and predict the variation of perceived tenure insecurity in urban deprived areas based on the city of Kigali, Rwanda. The research started by identifying variation of perceived tenure insecurity based on household survey data. These data were analysed through Multiple Correspondence Analysis (MCA) to obtain the indices representing that variation. The research applied hierarchical clustering to validate the latter, allowing the creation of four clusters corresponding to very high, high, moderate and low trends of perceived tenure insecurity. Hence, the research spatially mapped these clusters and evaluated their spatial distribution across the study area. Furthermore, the research extracted the land cover information and texture features from the VHR Google Earth image and additional spatial information such as slope and zoning plan maps, as indicators based on four buffer areas around households: 10, 15, 20 and 25 meters. Later, the research undertook several modelling processes using a random forest regression model to understand the relationship between image-based spatial indicators alongside other spatial information as indicators and the variation of perceived tenure insecurity. In addition, the study analysed the importance of each indicator for measuring and predicting the variation of perceived tenure insecurity. Findings revealed that respondents with a similar variation of perceived tenure insecurity are spatially concentrated due to the common factors inducing their perceptions. Moreover, the findings revealed that spatial characteristics from VHR images and other spatial information have the potential to capture the variation of perceived tenure insecurity in urban deprived areas. Furthermore, the research found that textural features present high importance in capturing such variation due to their capacity to capture the spatial arrangement of objects in the image. Moreover, additional spatial information describing location has significant prediction importance. These findings can assist municipalities and stakeholders to make evidence-based decisions for unjust city development. Besides, the research is a basis for further researches concerning the spatial measurement of tenure security trends and monitoring the implementation of the SDGs, especially goal 1(target 1.4) on tenure security for all and goal 11 addressing the issues of safe, inclusive and sustainable cities and human settlements.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/88718
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page