Published in Discover Cities, the study by geographers Tanmoy Malaker and Qingmin Meng is one of the broadest examinations yet of how GeoAI is being used in urban analytics. The researchers combined bibliometric mapping with a systematic review of journal literature indexed in Web of Science and Scopus. Their search initially returned 24,272 records containing combinations of urban and geospatial-AI terminology. After progressively narrowing the search, removing irrelevant publication types, eliminating duplicates and manually checking the remaining papers, the researchers identified 100 articles for detailed analysis. The unusually strict filtering was designed to focus on studies that actually applied AI to urban geospatial problems rather than merely mentioning smart cities or machine learning.

The review traces GeoAI to a convergence of three traditionally distinct disciplines: geography, geospatial technologies and artificial intelligence. Geographic information systems provide the spatial framework; remote sensing supplies observations of the Earth’s surface; and machine-learning algorithms identify patterns or make predictions from those observations. Machine learning allows computer models to learn relationships from data without being given every rule explicitly. Deep learning, a specialized form of machine learning, uses multilayered neural networks to extract increasingly complex features, such as edges, rooftops, road layouts or vegetation patterns, from images. In urban research, these systems can connect the visual appearance of a neighbourhood with measurable outcomes such as heat exposure, property value, pedestrian comfort or flood vulnerability.

The term GeoAI became prominent after a 2017 research gathering, although the underlying idea is far older. Artificial-intelligence methods were already being discussed in geography during the 1980s and 1990s. What changed recently was the explosion in both the volume and resolution of urban data. Modern satellites can distinguish individual vehicles, buildings and other small objects. Street-level imagery provides repeated views of sidewalks, façades, trees and public spaces. Smartphones generate mobility traces, while social-media posts can reveal people’s perceptions and reactions in near real time. These sources are heterogeneous: they differ in scale, format, accuracy and meaning. GeoAI’s central promise is to fuse them into a common spatial analysis, transforming photographs, text, sensor readings and movement records into evidence about city systems.

That promise is reflected in the field’s rapid growth. The review found a marked increase in publications after 2020, with 2024 producing the largest number of studies in the dataset. China led the national research contributions with 33 papers, followed by the United States with 21 and Singapore with 18. European countries collectively accounted for 39 publications across roughly 17 nations. The research locations were not always close to the researchers conducting the work. In some cases, scholars based in Africa studied cities in Europe or other regions where high-quality geospatial datasets were easier to obtain. Open satellite imagery and online mapping resources have lowered geographical barriers to research, but they have not eliminated the deeper inequality created by uneven data availability.

Across the 100 studies, the urban environment was the largest application category, with 13 papers, followed by urban development and urban hazards, each with 12. Other applications included transportation, urban morphology, landscape, housing, data-driven urban decision-making, blue-green infrastructure, land-use and land-cover change, and urban safety. Researchers used GeoAI to classify land surfaces, extract building footprints, predict traffic, assess air and noise pollution, identify flood zones, study urban heat, evaluate public spaces and examine how neighbourhood design influences people’s feelings of safety or psychological restoration. Some studies combined satellite imagery with street-level photographs; others linked environmental measurements with socioeconomic indicators or anonymous mobile-phone data.

The dominant pattern was not the use of one supposedly miraculous algorithm, but the construction of hybrid analytical workflows. Sixty-four of the reviewed studies used combinations of artificial intelligence, machine learning, deep learning and conventional geospatial methods. Nineteen primarily emphasized machine learning, 10 focused on deep learning, and seven relied on established pretrained AI models. The categories overlap because the technologies are nested: deep learning is part of machine learning, while machine learning is generally treated as part of the broader AI family. Hybrid systems are attractive because urban problems usually require several stages of processing. A model may first use a convolutional neural network to identify objects in satellite imagery, then combine those results with GIS layers, spatial statistics and socioeconomic data to predict a neighbourhood-level outcome.

The bibliometric analysis, performed with the software VOSviewer, revealed the intellectual structure of the field by measuring how often key terms appeared together. GeoAI was the most connected term, as expected, but artificial intelligence, machine learning, deep learning and remote sensing formed the strongest surrounding network. Remote sensing had especially prominent links with classification, GIS, street-view imagery, the built environment and green space. The researchers found that 47 studies combined image analysis with geospatial modelling, while 42 focused primarily on geospatial models. Eight concentrated on image processing and three on social sensing. Imagery was the main data source in 33 studies, and 30 used multiple sources, reinforcing the conclusion that modern urban GeoAI is fundamentally image-rich and data-fusion driven.

The algorithms involved range from familiar statistical tools to highly specialized neural networks. Convolutional neural networks are widely used to detect spatial features in images, while U-Net architectures segment satellite scenes and extract building footprints or land-cover classes. Graph neural networks represent streets, buildings or neighbourhoods as connected nodes and edges, allowing models to learn relationships that ordinary image grids may miss. PointNet can process three-dimensional point clouds for urban reconstruction and digital-twin applications. Random forests, gradient-boosting models and support-vector machines remain valuable for classification, prediction and identifying which variables influence a result. Geographically weighted regression and related spatial methods account for the fact that relationships can vary from one location to another, while space-time cubes organize observations simultaneously by place and time. Natural-language processing and large language models are beginning to add information about public sentiment, human experience and mobility behaviour, although these applications remain a small part of the reviewed literature.

The review’s most important message is that technical sophistication does not automatically produce trustworthy urban intelligence. Many deep-learning models require large collections of accurately labelled training data, yet such data are expensive, incomplete or concentrated in wealthy cities. Conventional algorithms may also struggle with spatial autocorrelation, in which nearby observations resemble one another, and spatial heterogeneity, in which the same relationship changes from one district to another. A model trained in one city can therefore perform poorly elsewhere. The researchers also identify a persistent shortage of socioeconomic and behavioural information. Satellite images can reveal roofs, roads and vegetation, but they cannot by themselves explain income, exclusion, cultural practices or how residents experience a place. Without those dimensions, a system may optimize the visible city while misunderstanding the human one.

Interpretability is another barrier. Deep neural networks can recognize patterns with remarkable accuracy, but their internal reasoning is difficult to inspect. That “black box” problem matters when a prediction influences zoning, infrastructure investment, emergency response or access to public services. Explainable AI methods attempt to show which image regions or variables affected a decision, yet spatial explanations remain incomplete. Computational cost is also significant: processing high-resolution imagery, sensor streams and large space-time datasets can demand hardware and expertise unavailable to smaller institutions. The review therefore calls for models that are more efficient, transferable and transparent, alongside spatially adapted explainability tools that can show not only what a model predicted but where, why and for whom the prediction may be unreliable.

GeoAI is moving toward a more human-centred phase, the researchers conclude. Future systems could combine satellite and street-view imagery with mobility records, environmental sensors, social-media language, public-health data and behavioural models in unified urban intelligence platforms. Generative AI and large language models may help translate unstructured text into spatial information, simulate alternative planning scenarios or support dynamic digital twins of cities. Yet the authors argue that interdisciplinary collaboration will be essential: geographers, urban planners, computer scientists, data scientists and social researchers must work together to determine which questions are worth asking and whose experiences are represented. If that balance can be achieved, GeoAI could become more than a faster way to map cities. It could help communities anticipate hazards, distribute resources more fairly and design urban environments that are not only efficient and sustainable, but genuinely responsive to the people who live in them.

Traditionally, city boundaries have been drawn based on administrative lines, but these often fail to capture the functional and lived realities of urban regions. As urban areas morph, sprawl, and reconfigure, administrative borders can lag behind socio-economic and physical changes, making them inadequate for many contemporary analyses. Alternative methods utilize various proxy datasets, each reflecting different facets of urbanity, such as population density, built-up areas as detected through satellite imagery, and night-time light emissions as captured by remote sensing technologies. These proxies, while informative, come embedded with unique assumptions and biases, necessitating a clear rationale for their selection.

Population-based proxies define cities based on concentrations of people, often relying on census data or other demographic sources. This approach can highlight areas of intense human activity and social interaction. However, it may neglect important spatial structures, such as industrial zones or unpopulated commercial districts, which are nonetheless integral parts of the urban fabric. Moreover, population data are typically collected at intervals as long as a decade, limiting their responsiveness to rapid urban dynamics. Consequently, population metrics provide a somewhat static and socially centered picture of a city’s boundaries.

Conversely, proxies relying on built-up area data leverage remote sensing imagery to detect physical infrastructure and developed land. This method captures the morphological aspect of urbanization, mapping where human construction dominates the landscape. Built-up area proxies often use high-resolution satellite data to discern artificial surfaces, offering fine spatial detail with high temporal frequency. While they effectively mirror the tangible footprint of urban expansion, these proxies may misclassify transient or non-urban structures, and often do not differentiate between residential, commercial, or industrial land uses, potentially conflating distinct urban functions within the boundary delineation.

Another widely used proxy is night-time light intensity, which measures the artificial illumination visible from space during the night. This method taps into the energetic and economic activities that illuminate urban nightscapes, providing a dynamic indicator of urban vibrancy and development. Night-time light data have increasingly become popular for mapping urban extents in places with limited census or infrastructure data. However, they can introduce artifacts such as blooming effects, where lights appear to spread beyond their actual sources, and may underrepresent low-income or peri-urban areas with less developed electrical infrastructure, skewing the perceived urban footprint.

Recognizing these inherent differences, Van Migerode and colleagues advocate for a tailored approach that moves beyond the one-size-fits-all mentality. Their framework encourages researchers to explicitly state the conceptualization of the city they aim to operationalize, considering whether their focus lies in socio-economic interactions, morphological characteristics, or energetic activity. By interrogating the theoretical underpinnings of urban conceptualization, researchers can better match proxy datasets to the phenomena under investigation, yielding more accurate and meaningful delineations.

This purpose-driven framework also addresses the multifaceted nature of cities as complex socio-technical systems. For studies exploring urban economic networks, population proxies may offer the most relevant boundaries, capturing hubs of human activity and interaction. In contrast, infrastructural planning or environmental assessments may benefit from built-up area proxies to define where physical developments occur. Similarly, investigations into urban energy consumption or night-time economies might find night-time light data more indicative of their spatial focus. Thus, choosing the proxy is not merely a technical decision but an epistemological one, deeply connected to how urbanity is understood.

The framework proposed promotes transparency in research methodologies by encouraging authors to articulate their rationales clearly. This practice can enhance reproducibility and comparability across studies, as differing urban delineations often lead to incongruous results in cross-city or temporal analyses. Moreover, explicit justification enables better critique and refinement of urban proxies, fostering a collective advancement in the science of urban boundary delineation.

Technological advances in remote sensing and data availability further complicate and enrich this field. High-resolution satellites, increased frequency of data capture, and improvements in light detection instruments have expanded the potential of morphological and night-time light proxies. At the same time, the proliferation of big data sources such as mobile phone records, social media check-ins, and GPS traces presents emerging avenues to define urban boundaries based on patterns of human mobility and interaction, though these too bring questions about representativeness and privacy.

Indeed, the use of multi-proxy approaches is gaining traction as a way to integrate complementary perspectives. Instead of relying on a single dataset, combining population, built-up area, and night-time light data can create hybrid delineations that balance social, physical, and energetic dimensions of urbanity. However, such integrative methods necessitate sophisticated statistical and machine learning techniques to reconcile often contrasting signals and to assign appropriate weights, underscoring the importance of a well-founded theoretical framework to guide the integration process.

The implications of selecting appropriate urban proxies extend beyond academic inquiry. Urban policy, infrastructure investment, disaster management, and sustainability planning all depend on accurate urban boundary definitions. For example, misestimating city size can lead to under- or over-provision of services, misallocation of fiscal resources, and inadequate risk assessments. As urban populations worldwide continue to grow and new urban forms emerge, such as mega-regions and polycentric metropolitan areas, the challenge of delineating city boundaries only intensifies, demanding innovative and flexible approaches.

Van Migerode and colleagues’ contribution is thus timely and significant, offering a pathway to sharpen analytical precision amid the complexity of urban systems. Their emphasis on purpose-driven selection and methodological clarity serves as a call to action for urban scientists, geographers, planners, and policymakers alike to critically evaluate and transparently report the foundations of their city boundary definitions. This paradigm shift has the potential to yield urban analyses that are not only methodologically robust but also genuinely reflective of the multifarious realities cities embody.

The study also raises questions about future directions in urban boundary research. How can emerging technologies, such as artificial intelligence and real-time data streams, be harnessed to dynamically update urban boundaries? What ethical considerations arise when integrating new data types? Can standardized frameworks be developed without sacrificing contextual sensitivity? These pressing inquiries underscore that delineating cities is as much an evolving scientific endeavor as it is a reflection of changing societal values and technological landscapes.

In sum, the research by Van Migerode, Poorthuis, and Derudder advances the discourse on urban boundary delineation by positioning the proxy selection process as fundamentally purpose-driven, rather than convenient or conventional. Their framework champions the alignment between conceptual goals and empirical data, ensuring that proxies serve as accurate mirrors of the urban phenomena under study. As cities continue to shape the future of humanity, refining how we define their edges remains a critical step in understanding and managing urban complexity.

This nuanced approach promises to enhance the rigor and relevance of urban studies, enriching everything from demographic assessments and infrastructure planning to environmental monitoring and economic development strategies. Through clarity and intentionality in proxy selection, researchers can unlock new insights into urban form and function, enabling smarter decisions that respond effectively to the challenges and opportunities of urban life.

Ultimately, the quest to delineate city boundaries is a reflection of our broader efforts to comprehend the urban condition in all its dimensions. As this study reveals, the tools we choose shape not only our maps but also our understanding of what cities are and what they could become.

Subject of Research: Proxy datasets for delineating city boundaries and their suitability based on purpose-driven selection.

Article Title: Proxy datasets to identify city boundaries.

Article References:
Van Migerode, C., Poorthuis, A. & Derudder, B. Proxy datasets to identify city boundaries. Nat Cities (2026). https://doi.org/10.1038/s44284-026-00464-6

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s44284-026-00464-6

At the heart of this study is the use of extensive building footprint data generated from high-resolution satellite imagery provided by Maxar and meticulously processed through the DigitizeAfrica platform. This comprehensive mapping of building outlines forms the foundation for analyzing urban structure with unprecedented granularity. Complementing this are detailed street network datasets extracted from OpenStreetMap, carefully curated to exclude irregular pedestrian pathways, ensuring that the analysis captures functional urban environments critical for infrastructure access.

Population distribution, a crucial dimension in evaluating urban and peri-urban dynamics, was estimated using two authoritative raster-based datasets: LandScan and WorldPop. These datasets provide global population grids at resolutions fine enough to be meaningfully downscaled to street-block levels. The researchers employed a sophisticated allocation method that apportions population counts proportionally based on building area within delineated blocks, thereby aligning demographic data with the physical urban fabric, an approach that significantly enhances the precision of population density estimates.

The delineation of urban, peri-urban, and rural areas was grounded in geometries from the Global Human Settlement Layer (GHSL) Urban Centre Database, with peri-urban zones defined as 10-kilometer buffers around urban extents. This spatial stratification acknowledges the fluid and transitional nature of urban expansion in African contexts, recognizing peri-urban areas as critical arenas where infrastructure deficits often manifest most acutely.

A unique aspect of the study is the generation of spatial blocks—polygonal units defined by contiguous street networks supplemented by natural boundaries such as coastlines and rivers. These blocks serve as the basic analytical units, offering a lens through which the relationship between urban morphology and socio-economic indicators can be explored. The authors highlight the exclusion of footpaths and minor trails from this network to maintain focus on the infrastructure most relevant to formal urban accessibility.

To validate the richness and accuracy of the building footprint data, the team conducted cross-verifications with satellite imagery across various regions, taking into account the likelihood that very small structures may not be residential in nature. Additionally, the completeness of OpenStreetMap street networks was systematically assessed, reinforcing the reliability of the subsequent spatial analyses.

The research further integrates socio-demographic insights from the Demographic and Health Survey (DHS) program, encompassing 238 administrative regions within 22 countries, with 40 unique surveys conducted over a decade (2010-2021). This broad dataset provided nearly 370 subnational observations, enabling rigorous correlation analyses between the spatial block characteristics and indicators such as slum population proportions and health outcomes. Principal component analysis across 67 measured indicators distilled multidimensional data into actionable insights regarding informal settlement distribution and urban complexity.

One of the paper’s notable contributions is the validation of a metric denoted as k, an estimator designed to quantify informal settlement prevalence. This metric was rigorously calibrated against self-reported slum maps from nine urban areas, demonstrating its robustness in capturing the spatial footprint and heterogeneity of underserved urban locales. Such advancements in measurement are pivotal in addressing the long-standing opacity surrounding informal settlements, often excluded from official statistics.

Beyond the empirical findings, the study pioneers an open, interactive visualization platform accessible at millionneighborhoods.africa, which empowers users to explore multi-scale urban characteristics across the continent. This tool presents an intuitive interface allowing users to transition seamlessly from a continental overview to granular analyses at the street-block level. Features include dynamic displays of urban complexity k, building counts, population estimates, and land usage types, all overlaid with satellite imagery for real-time visual validation.

This visualization caters to a broad audience, ranging from urban planners and policymakers to researchers and civil society actors, fostering transparency and enabling data-driven decision-making. By situating population density estimates within tangible urban forms at high resolutions, the platform sheds light on the spatial distribution of infrastructure needs, potentially guiding targeted interventions in areas of acute deficit.

Importantly, the research underscores the significance of peri-urban zones as liminal spaces where informal settlements frequently expand amidst the absence of adequate infrastructure. These zones, often overlooked in traditional urban planning frameworks, emerge as critical targets for infrastructural investments that can alleviate poverty and improve living conditions, ensuring sustainable urban growth trajectories.

The integration of multisource datasets—building footprints, comprehensive street networks, population grids, and socio-demographic surveys—exemplifies a holistic approach to urban analysis, leveraging the strengths of remote sensing, crowd-sourced mapping, and institutional data repositories. This methodology offers a replicable model adaptable to other regions confronting similar urbanization and informality challenges globally.

This groundbreaking work not only advances scientific understanding of urbanization patterns in sub-Saharan Africa but also delivers practical tools and metrics to quantify and visualize the infrastructural deficits contributing to informal settlement formation. Its implications extend beyond academic discourse, charting pathways for responsive urban governance and inclusive policy frameworks vital for dealing with the continent’s explosive urban transformation.

As the global community increasingly recognizes the centrality of sustainable and equitable urbanization to development agendas, studies of this caliber illuminate the pathways through which data-driven insights can catalyze effective solutions. By bridging satellite-derived spatial data with ground-truth demographic and health indicators, Bettencourt and Marchio’s research sets a new standard for spatial urban analysis and underscores the urgency of addressing infrastructure deficits as a cornerstone to solving the informal settlement crisis in sub-Saharan Africa.

Subject of Research:
Infrastructure deficits and informal settlements in sub-Saharan Africa analyzed through high-resolution spatial data and demographic surveys.

Article Title:
Infrastructure deficits and informal settlements in sub-Saharan Africa

Article References:
Bettencourt, L.M.A., Marchio, N. Infrastructure deficits and informal settlements in sub-Saharan Africa.
Nature (2025). https://doi.org/10.1038/s41586-025-09465-2

Image Credits:
AI Generated