Worldwide delineation of multi-tier city-regions

The study addresses a major evidence gap in how societies are spatially organized around multiple urban centers. While 92% of the global population lives in an urban center or within one hour’s travel, existing research typically focuses on large cities or single-center delineations, overlooking multi-tier interactions between cities and their hinterlands. Drawing on concepts from central place theory and recent advances such as the universal visitation law of human mobility, the authors propose a global, systems perspective that captures access to different tiers of urban centers (towns, small, intermediate and large cities) and the functional ties linking them to surrounding periurban and rural areas. The purpose is to systematically delineate city-regions worldwide for 2020 across 213 countries/territories, quantify primary and secondary city-regions, and analyze population access to services and opportunities across multiple tiers using travel-time-based catchments (1, 2, and 3 hours). This framework informs sustainable regional development, infrastructure planning, and equitable access to services.

There is no consensus definition of city-regions; most conceptualizations involve one or more core cities linked to a hinterland via functional ties (for example, markets, commuting, housing, and services). The work builds on central place theory (Christaller) and its ongoing relevance in regional science, and on recent empirical generalizations of human mobility such as the universal visitation law. Existing delineations—Functional Urban Areas (FUAs) and Urban–Rural Catchment Areas (URCAs)—typically link locations to a single urban center, limiting the ability to represent multi-center, multi-tier access. Prior studies on degree of urbanization and accessibility (OECD/EC and others) add nuance to urban/rural classifications but still lack a systematic, global, multi-tier analysis. This study extends that literature by allowing locations to access multiple tiers and by using precise travel-time catchments globally.

Study scope and data: The analysis covers 213 countries/territories in 2020. Urban centers are identified and classified into four tiers by population: towns (20,000–50,000), small cities (50,000–250,000), intermediate cities (250,000–1,000,000), and large cities (>1,000,000). Urban centers and populations are derived primarily from the Global Human Settlement Layer (GHSL), with robustness checks using multiple gridded population datasets (GHS-POP, GPW, LandScan, WorldPop). Each urban center receives a unique ID. Accessibility modeling: Travel-time to urban centers is computed using a least-cost path approach over a global cost surface representing roads, rail, navigable waterways, walkable terrain (with land cover), elevation/slope adjustments, and international border crossing delays. The computation uses eight-connectivity on a spherical grid (about 1 km resolution) with global wrap-around handling. Travel-time cutoffs of 1, 2, and 3 hours are analyzed, where 1 hour approximates commuting and 2–3 hours less frequent, more specialized trips. Catchments by activity tier: The authors assume a nested service hierarchy: higher-tier centers offer all lower-tier activities. They construct four activity grids corresponding to tiers of activities: (1) towns and above (basic activities), (2) small cities and above (higher-level services), (3) intermediate and above (more sophisticated/diversified services), (4) large cities only (most specialized services, e.g., airports). For each activity grid and cutoff, they delineate minimum-travel-time catchments for each urban center. Ties are resolved in favor of the smallest sufficient center for the required activity level to preserve lower-tier references. Patch construction: For each location, the closest center for each of the four activity tiers is recorded, yielding a four-digit type code (digits 0–4; 0 indicates no access within cutoff; 1–4 denote town to large city as provider). Unique combinations of providers across tiers define a “patch,” i.e., locations served by the same set of centers. Over 100,000 unique patches are identified globally. Patch IDs are persistent across travel-time cutoffs, enabling tracking of extent changes with different cutoffs. City-region delineation: City-regions are defined by grouping patches that share the same highest-tier urban center of reference within the cutoff. Primary city-regions are those whose patches all have the same highest-tier center as the city-region’s apex (no overlap with higher-tier centers). Secondary city-regions occur when at least one patch overlaps with a higher-tier center’s catchment; these are further categorized as satellite (no embedded lower tiers) or nested (with embedded lower tiers). The algorithm avoids double counting by distinguishing primary from embedded secondary systems. Country attribution uses GADM, assigning multicountry regions to the country with the largest surface share while also computing country-segment populations. Computation and tools: Spatial analyses and algorithms were implemented in Go, PHP, and Python; QGIS was used for prototyping and map production; final data cleaning/analysis in Excel; maps also produced in R. Code repositories: ITC-CRIB/city-regions and ITC-CRIB/globe-trotter. Outputs: For each travel-time cutoff, the study reports the number and types of primary city-regions by apex tier and computes area and population of urban cores, proximate catchments (<1 h from core edge), up to 2 h, and full catchments (up to 3 h).

• Coverage and tiers: 30,079 urban centers globally: 18,619 towns; 9,440 small cities; 1,538 intermediate; 482 large.
• City-region counts (1 h cutoff): 4,210 primary and 25,869 secondary city-regions. Of the primary: 1,524 single-tier and 2,686 multi-tier; among multi-tier, 1,751 two-tier, 627 three-tier, 308 four-tier. As cutoff increases to 3 h, primary city-regions consolidate to 1,403 (with more four-tier and fewer other types), reflecting reduced fragmentation and broader access to higher-tier services.
• Accessibility of urban centers: 92% of the global population lives in or within 1 h of an urban center (≥20,000 people); only 8% are >1 h away. Within 1 h, two-thirds of the world’s population has easier access to small or intermediate cities than to large ones.
• Access to multiple tiers: Of 7.8 billion people (2020), 41% (3.2B) have access to multiple urban tiers within 1 h; 57% and 64% within 2 h and 3 h, respectively (≈4.5B and 5.0B). The abstract notes 4.7B within 3 h.
• Distribution by closest center (1 h): 55% live in/closest to a town or small city; 37% to an intermediate or large city. Summing combinations that include intermediate and/or large tiers, about two-thirds of the world’s population live within 1 h of at least one intermediate or large city.
• Urban hierarchy patterns: 34% of the global population is part of city-regions with at least one intermediate or large city within 1 h. 28% belong to two-tier regions; 13% to three- or four-tier ones (rising to 32% and 33% at 3 h, covering ≈5.0B people).
• Town/small-city populations’ higher-tier access: Among those living in or closest to towns or small cities, 20% have easier access to intermediate cities vs. 10% to large cities within 1 h (35% vs. 17% within 3 h), highlighting the bridging role of intermediate cities.
• National contrasts: Infrastructure and geography strongly shape accessibility. Ethiopia: 76% live >1 h from any intermediate/large city; 21% >1 h from any urban center; only 5% have access to a large city within 1 h; 70% can access towns/small cities within 1 h. France exhibits near-comprehensive coverage by three-tier regions within 3 h. Primary city-regions at 3 h: France 4 (of 223 centers), Ethiopia 20 (of 358), Nigeria 19, Pakistan 15. Ethiopia shows many primary regions centered on small cities, indicating limited connectivity.
• Tier membership: The smaller the center, the more likely it belongs to a higher-tier region: 95% of towns, 78% of small cities, and 56% of intermediate cities are embedded within a higher-tier city-region (1 h). The most common configuration is two-tier regions centered on small cities with one or more towns (1,560 within 1 h).

The findings demonstrate a globally prevalent, multi-tier structure of city-regions, consistent with central place theory and contemporary mobility regularities. Allowing locations to reference multiple urban centers across tiers reveals extensive interconnections among urban cores and their hinterlands. The consolidation of primary regions at longer travel times indicates that less frequent, specialized activities are effectively centralized in higher-tier cities, while frequent, basic needs are met by lower tiers. Policy implications are substantial: national strategies should account for the diversity of accessibility patterns by geography and infrastructure, eschewing one-size-fits-all approaches centered solely on primate cities. Intermediate and small cities, and in some cases towns, often engage far larger shares of surrounding populations than large cities do for daily or frequent activities, especially in lower-income countries. Prioritizing infrastructure and services that strengthen these nodes—and their radial connections via secondary/tertiary roads and transport—can enhance inclusive regional development, improve access to services (health, education), and support resilient, polycentric urban systems. The dataset provides a platform to integrate with socioeconomic and environmental data to guide land-use planning, watershed management, and cross-boundary environmental governance.

This study delivers the first systematic, global delineation of multi-tier city-regions by integrating population-based urban tiers with precise travel-time catchments and multi-center access. It shows that most people live within reach of urban services and that multi-tier linkages—especially involving intermediate and small cities—are critical for connecting hinterlands to opportunities. The publicly available dataset and tools enable country-level diagnostics and planning. Future research should incorporate place-specific service inventories, economic specializations, and governance contexts; refine tier thresholds where appropriate; consider multimodal/temporal variations in accessibility; and examine the dynamics and resilience of polycentric regions and secondary city-regions across development trajectories.

• Assumption that larger cities provide more specialized activities; actual service portfolios vary across countries and within tiers.
• Urban population as a proxy for functions does not capture economic composition, infrastructure quality, or institutional roles.
• Sensitivity to parameterization: results depend on travel-time cutoffs (1/2/3 h) and the population thresholds defining tiers.
• Global uniform criteria trade off local nuance shaped by history, culture, economy, and environment.
• Accessibility estimates rely on global cost surfaces and assumptions (network coverage, speeds, terrain effects, border delays) that may omit local constraints.
• Does not analyze historical drivers or governance shaping current urban hierarchies and linkages.
• Despite testing multiple population datasets, residual uncertainties in gridded populations and urban extents remain.