Functional Urban Areas & Big Data
Interactive Map of Functional Urban Areas in Uzbekistan Built with Mobile Phone Data
Habidatum & World Bank, Degree of Urbanization Methodology by the European Commission and OECD
In this article, we present a recent analysis conducted by Habidatum, in collaboration with the World Bank, aimed at mapping and analyzing Functional Urban Areas (FUAs) in Uzbekistan. The study utilizes mobile phone data from a telecom operator Beeline Uzbekistan and is based on the European Commission’s & OECD Degree of Urbanization methodology.
Cities are more than just administrative boundaries. They are dynamic and complex systems that extend beyond their official borders and interact with other urban areas. To understand and manage cities better, we need to define and measure their functional urban areas (FUAs), which are the areas that are economically, socially and environmentally integrated with a core city.
FUAs can be identified by using different criteria, such as population density, land use, commuting patterns, or service provision. One of the most widely used and internationally comparable definitions of FUAs is the Degree of Urbanization (DEGURBA) methodology developed by the European Commission and the OECD. DEGURBA classifies local areas into three types: cities, towns, semi-dense areas and rural areas, based on their population size and density. FUAs are then defined as the combination of a city (the urban center) and its surrounding towns and semi-dense areas (the commuting zone) that have at least 15% of their employed residents working in the city.
DEGURBA is a useful tool for policymakers, researchers and practitioners who want to monitor and compare the performance of urban areas across different dimensions, such as environmental sustainability, social inclusion, economic development, or service provision. DEGURBA can also help to identify the challenges and opportunities that different types of urban areas face and to design and implement policies that are tailored to their specific needs and potentials.
However, applying DEGURBA to different contexts requires reliable and up-to-date data sources, which are not always accessible via regular statistical resources. This is where mobile phone data can offer new insights and possibilities.
Mobile phone data is a rich source of information about human mobility and behavior. By analyzing anonymized and aggregated records of mobile phone activity, we can map the spatial distribution and temporal variation of population density, identify the home and work locations of users, estimate the origin-destination flows of commuters, and infer the functionality of different urban areas.
Habidatum, a global urban data analytics company, has been using depersonalized mobile phone data to map and analyze FUAs in Uzbekistan, under the assignment of the World Bank. The project aims to inform a better understanding of both connectivity among urban centers in agglomerations surrounding target cities and the definition of functional urban areas in the country. Habidatum serves as a liaison for introducing mobile phone data into FUA analysis, making it more dynamic and granular. This involves negotiating data access, ensuring data anonymity and ethical use, developing analytical methodologies, conducting detailed analysis, and offering capacity-building support to the clients, in more detail:
• Providing technical guidance for the process of data access negotiations, data aggregation, transfer and management, ensuring data anonymity and its ethical use.
• Developing and reaching confirmation and approval of the analytical methodology.
• Conducting analysis and producing analytical outputs including metrics and visualizations covering commuting patterns, extent of functional urban areas, and functionality of different urban territories.
The project uses mobile phone data from Beeline Uzbekistan, one of the largest telecom operators in Uzbekistan. The depersonalized data includes information about the home location (night-time) and work location (day-time) of users. The data is adjusted to reflect the total population by using census data and population estimates.
The project applies DEGURBA methodology to identify urban and rural clusters based on home location data. Work location data is used to identify commuter flows between clusters. The project also explores changes in population density of different areas within FUAs throughout different hours of the day, and different days of the week to help understand what function different spaces serve and evaluate the sufficiency of services and infrastructure provided in different areas.
The project covers 6 target cities in Uzbekistan: Tashkent and Chirchiq; Samarkand; Namangan; Urgench (Xorazm region); Denau (Surxondaryo region). For each city, the project defines its FUA boundary (the urban center and the commuting zone), classifies its areas into 3 types by DEGURBA (urban center, urban and rural clusters) and 3 types by population dynamics (stable, acceptor, donor), then targets specific case studies and analyzes their transport connections, amenities and anchors for workers, consumers and residents.
The project produces an interactive map that allows users to explore the results of the analysis for each city. The map shows the FUA boundary for each city and the classification of areas within each FUA
The map can be accessed at https://uzbekistan-fua-map.habidatum.com
The project also provides preliminary recommendations for the development of FUAs based on the results of the analysis. For example, acceptor areas that attract workers but are not yet as populated and diverse as urban centers, may become the main targets in polycentric development strategy in the role of secondary centers (case of Samarkand). Or, donor areas that lose commuters and economic activities in competition with larger urban centers located far away, may be seen as development targets in terms of reducing commute, either through developing alternative job concentrations and workforce magnets or through developing high-speed connections with the existing far away urban centers (case of Chirchiq-Tashkent). Recommendations include creating multi-modal transport planning, building mixed-use development master plans, or introducing developer impact fees.
The use of mobile phone data offers new insights into commuter flows, helping explain why people commute to specific urban areas, even if they are distant. This data, combined with commute times and daytime versus nighttime population information, informs transportation improvements such as road upgrades and expanded public transit. These improvements aim to increase productivity by reducing unproductive commute time and improving job matches in integrated local labor markets.
The geoanalytics demonstrates the potential of mobile phone data for mapping and analyzing FUAs in Uzbekistan and beyond. It also shows how DEGURBA methodology can be applied and enriched with new data sources and analytical angles. The project provides a framework that can be replicable across other cities and countries.
The project builds on Habidatum’s prior experience in mapping agglomerations. Past relevant projects include mapping the boundaries of agglomerations based on travel time and geotagged social media data (2017), or mapping building density within agglomerations (2018).
To summarize, this project seeks to enhance our understanding of connectivity within urban centers, define functional urban areas, and provide actionable insights using mobile phone data. It contributes to policy development and sustainable urban growth and can be replicated in other cities and countries.
If you are interested in collaborating with us or learning more about our work, please contact us at ask@habidatum.com. We look forward to hearing from you.
Annex 1. About the interactive map
https://uzbekistan-fua-map.habidatum.com
This is a result of Habidatum’s mobile network (Beeline Uzbekistan) depersonalized data analysis under an assignment of the World Bank.
Background
Uzbekistan is undergoing a period of rapid urbanization and chaotic urban growth and expansion, while its cities are struggling to become the drivers of economic growth and improved well-being largely as they are poorly managed and undersupplied with infrastructure, basic services, and amenities. The government of Uzbekistan has recognized the importance of tackling the challenge of urban expansion and moving to a better-managed model of urban growth. As a part of this new approach, the World Bank is supporting the Government of Uzbekistan with a loan to finance the Medium-Size Cities Integrated Urban Development program which combines investment in infrastructure and livability in id-size cities with capacity building at the national level to support government’s effort of managing urbanization better. The initiative includes plans to develop a country Urbanization strategy that should underpin the National Medium-size cities program and define investment and policy priorities in urban development for the years to come. In order to develop a better understanding of how cities comprising the system of cities in Uzbekistan functionally interact and connect, and inform infrastructure and service provision investments as well as critical national reforms for urban development, the World Bank is carrying out a spatial analysis including an analysis of population mobility.
Description
The purpose of this report is to bring the attention of policymakers to the importance of integrating vast urbanized territories that emerged as a result of a chaotic urban build-up expansion in recent years around major urban centers of Uzbekistan and to inform infrastructure investment decisions that could support spatial integration and further the socio-economic development within these expansive urban agglomerations of FUAs (Functional urban areas).
For that, we define the boundaries of FUA around 6 largest urban centers: Tashkent and Chirchiq; Samarkand; Namangan; Urgench (Xorazm region); Denau (Surxondaryo region). based on the commuting patterns that indicate the extent of the catchment area of the city’s labor market. We then classify areas within each FUA. Area classes themselves may signal certain development needs and strategies.
For example, acceptor areas attracting workers but not yet as populated and diverse as urban centers, may become the main targets in polycentric development strategy in the role of secondary centers. Or, donor areas losing population in competition with larger urban centers located far away, may be seen as development targets in terms of reducing commute, either through developing alternative job concentrations and workforce magnets or through developing high-speed connections with the existing far away urban centers.
Definitions that we use are specified below:
- Data sources: mobile phone data or mobile network data or telecom data or telco data or cell phone data (synonyms) and relevant abbreviations; GIS data produced by online maps, such as Google or Open Street Map (includes road network mapping, venues, and addresses, main settlements); census and other official statistical sources.
- Local unit: administrative unit (district, region).
- Commuting zone: local units (districts) or grid-level areas, passing the commuting flow thresholds to the urban center of the FUA.
- Urban center: a set of contiguous, high-density (1,500 residents per square kilometer) grid cells with a population of 50,000 in the contiguous grid cells.
- Functional urban area (FUA): funcational urban area is a combination of a densely inhabited urban center and a commuting zone, itdentified above. The boundary of the FUA forms a single contiguous area.
- Area class: class of the area (cluster), based on two distinct classifications:
- Degree of urbanization:
- Urban center: see above;
- Urban cluster: contiguous (using eight-point contiguity) 1km grid cells with a density of at least 300 inhabitants per km2, with a collective population of at least 5 000 in the cluster;
- Rural cluster: contiguous (using eight-point contiguity) 1km grid cells with a density of at least 300 inhabitants per km2, with a collective population of at least 500 in the cluster;
- Population dynamics (based on the difference between nighttime and daytime population):
- Stable cluster/area: area, where % difference between nighttime and daytime population is less than a set threshold;
- Acceptor cluster/area: area, where the nighttime population is larger;
- Donor cluster/area: area, where the daytime population is larger.
Spatially, the general pattern for all FUA is the prevalence of donor clusters surrounding the urban center. Acceptors and stable areas can both be adjacent to urban center (as in the case of Samarkand and some others, see further descriptions below) or be located further away (smaller-sized clusters, forming local labor market nodes).
It is important to note that each cluster class presence and development plays its role in balanced FUA development:
- Donor of daytime population: concentrates labor force, may need better connection to jobs (acceptors, centers), local job alternatives (supply of services and associated jobs), and comfortable amenities supply that would allow people consume services not only in the places of work (centers and acceptors) but also in their living areas. Developing fast and easy transportation services between these areas is also essential;
- Areas where the delta between daytime and nighttime population is not significant (stable clusters) may be considered as targets for commercial services development, both for the local residents and workers, which would not only increase the comfort of life, but also generate more jobs, so that a stable cluster can gradually turn into an acceptor/secondary center;
- Acceptor clusters, in their turn, already are attracting workforce, and with further development of commercial density and diversity, may complement urban centers as secondary nodes of socio-economic activity, and be the drivers of polycentric FUA development.
How to use the map:
Search and explore Uzbekistan’s Functional Urban Areas: their
boundaries, area types by the degree of urbanization (urban-rural), and
commuting patterns (acceptor-donor). Use search, move the map,
zoom in and out, and hover the mouse on the areas of your interest to
see their boundaries. If you have any questions, reach out to us via
email: ask@habidatum.com
Annex 2. About Beeline Uzbekistan involvement
Our team of analysts has diligently processed an extensive volume of data comprising over 330 million lines, focusing on the mobility patterns of residents across six prominent regions. The outcomes of our research shed light on the residents of Uzbekistan, predominantly concentrated in the vibrant city of Tashkent and its environs, with Namangan and Samarkand securing the second and third positions respectively. This alignment with the overall population distribution in these regions further validates our findings.
To ensure the utmost accuracy and reliability, we meticulously analyzed the behavior of more than 70% of our active user base. Leveraging advanced tools and methodologies, we constructed these comprehensive models within a remarkably efficient two-week timeframe. The sheer scale of our data processing endeavors is evident in the creation of an interactive map, commissioned by the World Bank, which showcases the socio-economic activity within these six major urban areas. This formidable dataset encompasses a substantial volume of approximately 800 GB, comprising around 10 billion rows.
During the analytical phase, we employed rigorous scrutiny and rigorous exclusion criteria to capture precise mobility patterns. Our considerations encompassed multiple variables, such as various modes of transportation, daily activity levels, and their correlations with weekends, holidays, leisure pursuits, and household responsibilities. By meticulously verifying hypotheses, our analysts were able to ascertain the accuracy and significance of the discovered patterns. Furthermore, the utilization of pre-existing predictive models, such as the HOME/OFFICE behavior framework, proved invaluable in augmenting the depth of our analysis.
To present our findings with utmost clarity and comprehensibility, we leveraged interactive visualization tools equipped with specialized filters. These tools facilitated the creation of visually engaging geospatial representations, enabling a nuanced understanding of the intricate dynamics of resident movements. Additionally, our dedicated team developed numerous custom scripts, tailored to extract the requisite data in the desired formats, thereby enabling multi-dimensional analyses across various parameters.
This groundbreaking research not only illuminates the mobility patterns of Uzbekistan’s residents but also serves as a testament to the power of data-driven insights in informing critical decision-making processes. As we continue to refine our methodologies and explore new avenues, our commitment to delivering accurate and actionable information remains unwavering.
