Engelmann, Gregor
(2021)
LU(S)TI in the global South: an empirical analysis of land use and socio-economic transport interaction in Tanzania using mobile network data.
PhD thesis, University of Nottingham.
Abstract
The majority of rural-urban migration is filtered through slums: informally established, unplanned, and unrecognised by the government, scientists have a minimal understand- ing of the 200,000 that exist worldwide, never mind enough insight into the millions of individuals living there. This limited understanding often coincides with a more general absence of data in traditional urban planning approaches, leading to most cities seeing development, positive or otherwise, preceding planning.
Wesolowski and Eagle (2010) highlighted the key need to use models of human mobility to help guide effective spatial planning policies. Previous research has shown that thinking about the built environment alone cannot account for individual differences in behaviour, and that we must also consider factors such as socio-economic circumstance and context (which are far more likely to contain explanatory value than the geographies of points of interest, such as home and work locations of individuals alone). However, this remains a very difficult topic to study. Emerging economies are often characterised by institutions struggling to keep even demographic data streams up to date. Combined with ineffective data collection strategies, it is often realistic to expect stakeholders to retain an overview of the dynamics of urban systems. This gap causes many issues, but particularly in East Africa: expense and logistics restrict the ability to deploy sensor technologies; fast-changing environments reduce the utility of traditional household and census surveying; and even when raw data exists there are distinct skill gaps for data analysis.
To address this, this thesis extends nascent work, and systematically investigates the use of Call Detail Records (CDR) and Mobile Financial Service (MFS) transaction logs to model mobility, demographics, land use and their interplay. Data used was automatically generated as part of day-to-day operations of a major Tanzanian Mobile Network Operator. As part of this thesis, three empirical analyses are carried out to test the boundaries of inferring activity-based land use, predicting cell tower coverage level socio-economic levels and generating mobility metrics in the form of Origin-Destination matrices and synthetic daily activity plans for the Tanzanian port city of Dar Es Salaam. Further, shortcomings of CDR and MFS data, and ways to overcome these, are identified.
Empirical chapters form the basis for the identification of factors from the spatial dimension focused on assessing the impact of the built environment, socio-economic circum- stance and mobility behaviour allowing for the extension of traditional land use-transport interaction (LUTI) models, through the inclusion of socio-economic characteristics. This culminates in a new empirical LU(S)TI analysis for a sub-Saharan context. The metropolitan area of the port city of Dar es Salaam, Tanzania, is a pertinent case study area as it is facing similar challenges to many other fast-growing metropolitan areas in emerging economies globally.
| Item Type: |
Thesis (University of Nottingham only)
(PhD)
|
| Supervisors: |
Goulding, James
Perrat, Bertrand
Golightly, David |
| Keywords: |
CDR, mobile phone data, luti, land use transport interaction, Tanzania, mobility, land use, socioeconomics, poverty mapping |
| Subjects: |
H Social sciences > HT Communities. Classes. Races
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5101 Telecommunication |
| Faculties/Schools: |
UK Campuses > Faculty of Social Sciences, Law and Education > Nottingham University Business School |
| Item ID: |
64510 |
| Depositing User: |
Engelmann, Gregor
|
| Date Deposited: |
17 Mar 2021 10:28 |
| Last Modified: |
17 Mar 2021 10:30 |
| URI: |
https://eprints.nottingham.ac.uk/id/eprint/64510 |
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
