Conference Agenda

The Global South is urbanising rapidly as its cities attract rural inhabitants in search for a better living. Urban dynamics in the Global North are likely different as suburbanisation is still a major development, but there are also signs of reurbanisation. This begs the question how different urban expansion actually is in different regions around the globe.

Building on a newly developed spatial analysis framework we analyse urban expansion for a large set of capital cities around the world. Our framework characterises five different ways in which urban agglomerations expand in size. In addition to the commonly distinguished processes of infill, edge expansion, and outlying development, we also characterise two forms of urban development that we expect to be especially important in the fragmented urban landscape of the Global South: merging and joining. Merging is defined as the amalgamation of relatively small urban units into larger ones, while joining refers to the growing together of relatively large urban patches. In the latter case the total area of the initial urban patches is larger than the newly developed urban area that connects them.

In this research, we employ the latest release of the Global Human Settlement Layer (GHSL) that provides a highly detailed, and fairly consistent rasterised account of urban development patterns since 1975. The spatial analysis process of our urban patch expansion framework consists of five stages: 1) detecting infill in urban voids; 2) identifying individual urban patches; 3) calculating the number of neighbouring patches for initial urban patches; 4) calculating the share of new urban patch areas relative to the total area of a new patch; and 5) calculating the share of borders with initial patches in relation to the total border length of the newly developed area.

Our research uncovers the dominant development processes per region and assesses how this dominance changes over time. In a subsequent panel regression analysis, we study the importance of various driving forces in steering these developments. More specifically we look at the relevance of economic and demographic indicators in addition to variables that describe the size and spatial patterns of the selected urban agglomerations. By separating the analysis for capitals in the Global South and North, we assess whether development trajectories differ between these regions.

Since the late 1990s, research on activity spaces has shifted the focus of segregation studies be yond residential areas, incorporating both temporal and spatial dimensions [1]. These expanded measures of segregation complement traditional residential segregation indices by capturing real-world social interactions. This shift enables a transition from place-based to people-based measures of segregation, offering a more dynamic understanding of social divides. Early work by Hägerstrand introduced a methodological framework to explicitly define the geography of social interactions, inspiring a steady stream of research across the social and natural sciences [2]. His contributions elevated this framework into a distinct sub-discipline known as Time Geography. Although initially overlooked due to the significant challenges of data management and analysis, Hägerstrand’s approach has gained substantial traction in recent years. This resurgence is largely driven by advancements in technology, which have simplified the aggregation and analysis of individual trajectories, contextual factors, and infrastructural constraints across urban spaces [3]. For example, Alessandretti et al. [4] demonstrate, using mobile phone data, that despite the exploratory nature of human behavior, individuals tend to confine their activity spaces to a limited number of locations. This finding aligns with Dunbar’s assertion of a cognitive limit—approximately 100 stable social relationships that an individual can maintain [5]. In this study, we explore how segregation, measured through voice call interactions, varies across cities of different sizes and socioeconomic status (SES) using mobile phone data in Chile and Brazil. By merging SES information with a Call Detail Records (CDR) dataset from a major mobile phone provider, we assess segregation patterns in voice call interactions across a gradient of city sizes in Chile. We analyze 12.5 and 1,258 million anonymized voice calls in Chile and Brazil respectively (representing social events) between reliable users to construct voice call networks for cities of different sizes. These networks are tagged with the SES of callers, call duration, and urban distance between callers, allowing us to characterize the nature of directed interactions based on SES. The SES of each user is estimated by determining their most probable residential location. Segregation is evaluated using a modified exposure index (E) applied across SES quin tiles, providing insights into the dynamics of social interactions within and between socioeconomic groups. More specifically, we evaluate E between SES quintiles as following: Eαβ = N/ NαNβ ∑T (ntα ntβ / nt) [6]. Here, α and β are SES, N total population. T is defined as the social space constructed from the egonet of diameter equal to three. These egonets were built for each user with known SES in the network. While city networks are generally sparselikely due to the dataset representing only a fraction of the market shareseveral network indices help describe the social structure of voice call interactions across cities and over time. For instance, call reciprocity remains large and consistent across cities. Assortativity by SES is positive in all networks, indicating that users tend to connect with others of similar socioeconomic status.

Urban expansion in rapidly developing Indian cities has intensified the competition between green and concrete spaces, impacting environmental sustainability and urban livability. Urbanisation often changes a city's spatial, social, economic, and ecological conditions. Urban Green Space refers to various vegetation types in urban areas, like forests, trees, shrubs, wetlands, parks, etc. Urban Green Space provides several environmental benefits and services, like air purification, temperature reduction, biodiversity promotion, and quality urban life. Urban Concrete Spaces are impervious surfaces like buildings and roads, primarily concrete, asphalt, etc. Increasing concrete surfaces indicate urban expansion and sprawl, resulting in a temperature change, reduced soil permeability and groundwater recharge, impacts on air pollution, an altered urban environment, and urban flooding. This study investigates the spatio-temporal dynamics of green and concrete spaces in Coimbatore, covering 257 sq. km, India's 16th largest urban agglomeration and the second largest city in Tamil Nadu, following Chennai. Coimbatore has experienced notable urbanisation, particularly in the last few decades, including a substantial population and urban area increase, leading to land use, land cover changes, and industrial growth. The study spans between 2017 and 2025, utilising high spatial and temporal resolution, multi-spectral PlanetScope (3m) satellite imagery, an excellent source for vegetation monitoring in cloudy areas, as it increases the chance of acquiring a cloudless image.

The analysis employs machine learning and geoprocessing tools to extract and classify land use and land cover features. A multi-year temporal dataset was processed using GIS-based tools to map and quantify the changes in green and concrete spaces. The results reveal changes in green cover and built-up surfaces across the study period. The findings underscore significant urbanisation trends, indicating the conversion of natural landscapes into impervious zones, with implications such as heat stress, biodiversity loss, and challenges to environmental resilience. The study offers critical insights for urban planners, environmental regulators, and policymakers to improve green infrastructure planning for mitigating urban heat island effects and promoting sustainable urban development in metropolitan cities like Coimbatore, India.

Background and need: A recent World Bank study indicates that hazardous flood zones in human settlements more than doubled from 1985 to 2015. Rapid urbanization has led to a constant rise in impermeable surfaces, blocking natural drainage systems and exacerbating surface runoff. These urbanization-induced changes amplify local vulnerabilities and manifest as risks, including recurrent and often severe urban flooding. In this context, nature-based solutions have reappeared as an opportunity to create resilient cities. Under this umbrella concept, the blue-green infrastructure (BGI) occurring as natural blue and green spaces has significant potential to mitigate urban flooding. The urbanization process, coupled with the fragmentation of blue and green spaces, has reduced BGI’s capacity to absorb excess water during extreme rainfall events. Therefore, the hydro-ecological function of blue-green infrastructure in urban areas and their regional catchments needs to be considered while planning for urban development.

Study area and Data: Multiple studies have reported that over the last two decades, Indian metropolitan cities such as Chennai, Delhi, Hyderabad, Bangalore, and Mumbai have not just expanded but have also been experiencing recurrent urban flooding. The study area for this research is Pune and Bangalore, which are among Indian’s 10 biggest Urban agglomerations. In 2024, the Pune Municipal Corporation (PMC) identified “flood-affected areas” based on recent monsoon impacts and citizen-reported data. Similarly, the Greater Bangalore Municipal Corporation (Bruhat Bengaluru Mahanagara Palike, BBMP) classified “flood-vulnerable locations” in 2023 using remote sensing data and urban flood modeling. These in situ urban flooding datasets were obtained to carry out field investigations across both cities in locations where urban densification and urban flooding were conflicting. Simultaneously, the datasets indicating regional characteristics were extracted for these conflicting areas.

Methodology: The data acquired during the field investigation have been analyzed using micro-level spatial profiling techniques. The profiling of urban form and blue-green space characteristics was done using spatial variables such as built-to-unbuilt area ratios, extent of greenness, and the condition and presence of surface water bodies. These variables were mapped through visual observations and on-site photographic documentation. At the same time, the elevation (Digital Elevation Model), soil (typology), and precipitation (total annual rainfall) in these conflicting areas were extracted from secondary data sources and overlayed.

Results: The two-pronged approach provides a comprehensive overview of the spatial qualities of blue-green spaces in conflicting areas, i.e. urban flooding hotspots occurring in dense urban areas. The results reveal evidence to comment on the patterns of alterations of blue-green spaces in rapidly densifying urban areas. It also highlights spatial mismatches/ correspondence between areas of high impermeability and (in)adequacy of green or open areas. These findings offer evidence to comment on the causal relationship between urban densification, the disruption or absence of blue-green infrastructure, and the occurrence of urban flooding.