| No | Cities | Population |
|---|---|---|
| 1 | Lagos | 15,945,912 |
| 2 | Kano | 4,348,481 |
| 3 | Ibadan | 3,874,908 |
| 4 | Abuja | 3,839,646 |
| 5 | Port Harcourt | 3,480,101 |
| 6 | Benin City | 1,904,631 |
| 7 | Onitsha | 1,623,382 |
| 8 | Uyo | 1,329,284 |
| 9 | Nnewi | 1,239,186 |
| 10 | Aba | 1,188,803 |
| 11 | Kaduna | 1,187,398 |
| 12 | Ikorodu | 1,093,308 |
| 13 | Ilorin | 1,030,498 |
Methodology for global urban indicators in Nigerian cities
1 Introduction
This methodological paper will present the basic guidelines for estimating transportation and mobility benchmarks in Nigeria’s most populated cities. The cities are as follows:
This system of indicators is designed with the goal of facilitating comprehensive and comparable analysis across various cities. It is meticulously constructed to ensure uniformity in the assessment metrics, providing a standardized lens through which different urban environments can be evaluated and compared. This standardized framework allows consistent tracking and assessment over time, which is critical in understanding trends and making data-driven decisions.
Key areas that these indicators target include:
Mobility patterns: The indicators measure the mobility behavior of the population, focusing not only on general patterns but also on the share of sustainable mobility. This could involve analysis of public transportation use, cycling, walking, and other forms of low-impact commuting.
Urban sprawl and its expansion: This component monitors the growth and spread of urban areas. Tracking urban sprawl can help city planners understand the evolving demands on infrastructure and services.
Socioeconomic and demographic indicators: These metrics aim to provide a comprehensive picture of the population’s socioeconomic conditions. They could encompass metrics such as income distribution, employment rates, internet accessibility, population density, age distribution, and other relevant factors.
Proxies for environmental pollution: The system estimates the level of environmental pollution using suitable proxies, which could include measurements of air quality, water quality, and waste production, among others.
Hotspots for vehicular congestion: The indicators identify and monitor areas of high traffic congestion. This data can inform strategies to alleviate congestion and improve transportation efficiency.
Accessibility to infrastructure and public services: This area focuses on the extent to which individuals can access key infrastructural amenities and public services. It measures factors such as the distance to health care facilities, availability of public transportation, access to clean water, and the coverage of educational institutions, among others.
These indicators’ homogeneity is crucial, as it enables meaningful comparisons not just over time within a single city but also across different cities. This can lead to more informed decision-making and more effective strategies for urban development and management.
2 Inputs
In order to establish the indicators that will serve as a point of comparison between the different cities, four global thematic groups are established which, although they respond to an organizational criterion, are strongly interrelated.
| Indicator | Dataset | Provider | Requires_DDP | |
|---|---|---|---|---|
| Environment and Climate Change | Historic air quality | Air quality | Open Source Data | No |
| Historical temperatures | Temperatures | tomorrow.io | Yes | |
| Satellite imagery | API Engine | No | ||
| Nighttime lights from satellite sensors | Light Imagery | Terrain | Yes | |
| Socioeconomic and Gender | Gender distribution | HRPDM | Meta | No |
| Household income | Meta Relative Wealth Index / Internet Speed | Meta / Ookla | No / Yes | |
| Transportation and Mobility | Number of trips between H3 polygons and administrative regions | Veraset Movement | Veraset | Yes |
| Average distances | Veraset Movement | Veraset | Yes | |
| Average speeds per polygon and administrative region | Veraset Movement / Waze Irregularities | Veraset / Waze | Yes | |
| Optimal routing | Open Source API | OSM | No | |
| Vehicular congestion, by street, polygon, and region | Irregularities / Alerts / Jams | Waze | Yes | |
| Urban Development | Population | High Resolution Population Density Maps | Meta / Kontur | No |
| Building Footprints, both static and dynamic | Building Footprints | OSM / Google / Microsoft | No | |
| Number of hospitals, schools, primary, secondary, and tertiary streets | API | OSM | No |
Global thematic groups and indicators
These indicators offer a comprehensive set of tools for analyzing urban environments, allowing for detailed comparisons over time and between different cities. This multifaceted approach will yield a deeper understanding of the complex dynamics within urban systems, which can guide more informed decision-making and more effective urban management strategies.
3 Methodology
This analysis will be conducted using the R programming language, with the aid of several specific libraries to handle, visualize, and analyze the data. Here is a sample of the libraries and their uses:
tidyverse: An essential collection of packages for data manipulation and visualization.sparkR: Provides an R interface to Apache Spark, a fast and general engine for large-scale data processing.apache.sedona: A cluster computing system for processing large-scale spatial data, includes spatial Resilient Distributed Datasets (RDD) and DataFrames.tmap: Provides an elegant and flexible way to create thematic maps.caret: Used for training and plotting classification and regression models.sf: Simplifies the handling of geospatial data (features as rows) in R.raster: Allows handling of raster (grid) data, for example satellite data.leaflet: Used for interactive mapping.dplyr: A fast, consistent tool for working with data frames, both in memory and out of memory.ggplot2: An implementation of the Grammar of Graphics, providing the ability to produce sophisticated and complex visualizations with relatively simple code.readr: Provides a fast and friendly way to read rectangular data (like csv, tsv, and fwf).
These libraries, among others, will be utilized in various stages of data preparation, exploration, visualization, and modeling throughout the analysis.
3.1 Environment and Climate Change
Historic air quality: Historical air quality data is obtained from open source databases. This data is processed to provide regular air quality readings for each administrative region and H3 polygon.
Historical temperatures: Temperature data is sourced from tomorrow.io. Historical data is processed to provide regular temperature readings and trends over time for each administrative region and H3 polygon.
Satellite imagery: Satellite imagery is obtained from Google’s API Engine. The imagery is processed to extract valuable geospatial information like land cover, vegetation, and built-up areas.
Nighttime lights from satellite sensors: Light imagery is obtained from Terrain. The imagery is processed to identify the distribution and intensity of nighttime lighting, which can be used as a proxy for human activity and development.
3.2 Socioeconomic and Gender
Gender distribution: Gender distribution data is obtained from Meta’s High Resolution Population Density Maps. The data is processed to provide gender distribution figures for each administrative region and H3 polygon.
Household Income: The Relative Wealth Index predicts the relative standard of living within countries using de-identified connectivity data, satellite imagery and other nontraditional data sources. The data is provided for 93 low and middle-income countries at 2.4km resolution. The Ookla Broadband Speed dataset provides an extensive view of network performance from the world’s largest source of crowdsourced network tests.
3.3 Transportation and Mobility
Number of trips between H3 polygons and administrative regions: This indicator is derived from the Veraset Movement dataset. It involves processing the raw data to count the number of trips between H3 polygons and between administrative regions, using geographic information system (GIS) techniques to match trip origin and destination points with the appropriate geospatial zones.
Average distances: This metric is also derived from the Veraset Movement dataset. It involves calculating the distance between the start and end points of each trip and then computing the average trip distance.
Average speeds per polygon and administrative region: This indicator combines data from the Veraset Movement dataset and Waze Irregularities. It requires determining the distance and time of each trip to calculate speed, and then computing average speeds for each H3 polygon and administrative region.
Optimal routing: Using Open Source API from OpenStreetMap (OSM), optimal routing can be determined by using GIS techniques and routing algorithms, considering factors like road distance, travel time, and road conditions.
Vehicular congestion, by street, polygon, and region: Data from Waze’s Irregularities, Alerts, and Jams is processed to identify areas of high traffic congestion. This data is then aggregated at the street, H3 polygon, and administrative region levels.
3.4 Urban Development
Population: Population data is obtained from High Resolution Population Density Maps provided by Meta and Kontur. This data is processed to provide population figures for each administrative region and H3 polygon.
Building Footprints, both static and dynamic: Data from OSM, Google, and Microsoft is processed to identify the locations and dimensions of buildings within each city. Changes over time are tracked to monitor urban development.
Number of hospitals, schools, primary, secondary, and tertiary streets: Data is obtained from OSM’s API. The locations of hospitals, schools, and streets are identified and counted within each H3 polygon and administrative region.
4 Results
4.1 Socioeconomic and Gender
Household wealth measurements are predominantly derived from face-to-face surveys undertaken by the U.S. Agency for International Development. These comprehensive surveys encompass data from 1,457,315 unique households located in 66,819 villages spread across 56 low and middle-income countries around the globe.
In a bid to supplement this foundational understanding of wealth distribution, spatial markers are employed. These markers seamlessly associate each village with various non-traditional data sources. The sources include satellite imagery, mobile network data, topographical maps, and privacy-preserving connectivity data provided by Facebook. The amalgamation of these diverse datasets aims to paint a comprehensive and multi-dimensional portrait of each village’s inherent characteristics.
To delve deeper and extract tangible patterns from the non-traditional data, deep learning techniques, coupled with other computational algorithms, are utilized. This computational approach is pivotal in transforming the raw data into a discernible quantitative feature set representative of each village. The significance of this transformation is the illumination of attributes not readily evident from sole reliance on survey data.
Building on these meticulously derived quantitative features, a supervised machine learning model is then trained. The primary objective of this model revolves around the prediction of the relative wealth of every inhabited 2.4 km² grid cell on the planet. What stands out about this model is its capability to provide insights into regions where actual data might be scant or non-existent, effectively bridging knowledge gaps with predictive analytics.
The Relative Wealth Index can be combined with some basic population indicators like gender and age distribution, making deep indicators that can be used to understand the socioeconomic status of a city. For example, at country level:
The geospatial scope can be easily expanded or cropped to include cities and administrative regions, according to the needs of the analysis.
The possibility of connecting the Relative Wealth Index with other indicators, such as the number of hospitals, schools, and streets, allows for a more comprehensive understanding of the socioeconomic status of a each city. Combining this workflow with the global indicators and the largest cities in Nigeria, we can obtain a more detailed view and benchmark of every city.
For example, the following map shows the Relative Wealth Index for the city of Lagos, Nigeria:
