University of Twente Student Theses

Login

Analyzing urban development patterns in a conflict zone : a case study of Kabul

Chaturvedi, Vineet (2019) Analyzing urban development patterns in a conflict zone : a case study of Kabul.

[img] PDF
3MB
Abstract:To measure the development of a city in conflict zone is a daunting task because of the lack of access to the field data and security issues. This research aims to analyse the patterns of growth and the factors influencing the growth in the city of Kabul with the help of high-resolution earth observation-based data using logistic regression model. The growth is analysed from 2001 to 2017. The research explores the possibility to analyse settlement patterns by extracting image features from high resolution aerial photograph and terrain features as input to evaluate the capabilities of Random Forest classifier. The complexity of urban morphology in the city of Kabul made it difficult to classify the different settlement patterns due to the similarities in the texture of the planned and unplanned settlements and low contrast between the built and non-built. The texture similarity composed of localities consisting of mud houses and barren land also made it difficult to separate built-up from other classes. The classification consisted of 7 classes which includes 4 categories of built-up class, water, vegetation and barren land. The built-up is classified on the basis of built-up densities into medium density planned, low density planned, hillside settlements unplanned and medium density unplanned. The overall accuracy achieved is 50% with lowest accuracies for classes medium density planned and low density planned. To improve the accuracy of the classification the 2 built-up classes for the planned settlements are merged and the 2 built-up classes for unplanned are merged and the overall accuracy achieved is 61%. The unplanned settlements have grown by nearly 4.5 times from 2001 to 2017 whereas planned settlements have grown by 1.25 times. The growth in unplanned settlements is mostly towards the west and north west parts of the city and the growth in planned settlements is in the central and eastern parts of the city. The research further analyses the factors that influence the growth patterns in a conflict zone using binary logistic regression model. To perform the regression the change in the planned and unplanned settlements in Kabul is measured using the classified aerial photograph of 2017 and the Built-up layer extracted from IKONOS for the year 2001. The results of the regression reveal that population density and military bases are important factor for both planned and unplanned settlements. Slope is also one of the factors for the growth of unplanned settlements, which means there is growth of unplanned settlements on steeper slopes on the hillside.
Item Type:Essay (Master)
Faculty:ITC: Faculty of Geo-information Science and Earth Observation
Subject:43 environmental science
Programme:Geoinformation Science and Earth Observation MSc (75014)
Link to this item:https://purl.utwente.nl/essays/81553
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page