From Edge Expansion to Internal Infill: The Infrastructure Logic of Resource-Based Cities Behind Three Decades of Urban Evolution in Panjin

I. Thirty Years of Monitoring the Urban "Skin"

Urban impervious surface—land covered by impermeable materials such as asphalt and concrete—is the most direct physical marker of global urbanization. A study by a team from the Chinese Academy of Sciences, published in *Scientific Reports*, took Panjin City, Liaoning Province, China, as a sample and used deep learning combined with time-spectral-texture optimization methods to conduct pixel-level identification of urban expansion trajectories from 1990 to 2020. This study not only provides a set of data, but also reveals how resource-based cities adjust their spatial logic after long-term growth.

According to the study, the built-up area of Panjin increased from 312.75 square kilometers to 489.49 square kilometers, with an average annual expansion of 5.89 square kilometers. On the surface, these are ordinary figures in the wave of Chinese urbanization. But the key change lies in spatial form: the urban expansion model shifted from the early "leapfrog" and "edge" expansion to "infill" expansion after 2016. What does this mean? It means that Panjin has bid farewell to the extensive "spreading a big pancake" stage and begun to carry out density reorganization and functional renewal within an existing spatial framework.

For infrastructure researchers, this transformation has significance beyond a single city. It reflects the inevitable choice of China's resource-based cities under fiscal constraints, tightened land quotas, and the need for existing-stock renewal.

II. Spatial Dilemmas and Transformation Pressures of Resource-Based Cities

Panjin is a typical resource-based city, with its industrial system built on oil and gas resources. From the 1990s to the beginning of this century, resource-based cities often adopted a layout logic of "production first, life later." Industrial zones, residential areas, and public service facilities were separated from each other, relying on newly created land to meet growth demands. This model was effective during the resource extraction cycle, but once resources enter a decline phase, the city faces multiple pressures: declining fiscal revenue, slowing population growth, and inefficient use of existing space.

The study shows that the spatial compactness of Panjin's impervious surface has been declining, indicating that the early urban form tended to be dispersed. A dispersed urban structure implies higher infrastructure costs: longer pipeline networks, wider roads, larger commuting distances, and lower accessibility to public services. For a city seeking transformation, this is an unsustainable development model.

The model shift that occurred around 2016 coincided precisely with China's macroeconomic policy adjustments. At the national level, "urban renewal" and "existing-stock planning" have been emphasized, land supply policies have been tightened, and local governments have begun to re-examine the potential of existing built-up areas. The case of Panjin shows that this policy transmission is taking effect even in northern resource-based cities.

III. From Remote Sensing Data to Infrastructure Decisions: The Evolution of Tools

Traditional urban expansion monitoring relies on statistical yearbooks and administrative division adjustment records, whose accuracy and timeliness are insufficient to support refined planning. The methods adopted in this study—pixel-level classification by deep learning, combined optimization of time-spectral-texture features, and piecewise linear regression to identify abrupt change points—represent a replicable research paradigm.This paradigm is of great significance to infrastructure investment institutions. In the Global South, many cities lack high-quality land use data, leading to uncertainties in infrastructure siting and financing decisions. Monitoring systems based on remote sensing imagery and artificial intelligence can provide continuous, consistent, and verifiable urban spatial data without relying on traditional statistical systems. This means that infrastructure investors can more accurately assess the real speed and form of urban expansion, thereby optimizing project sequencing and capital allocation.

For example, in the Panjin case, if investors observed in 2000 that peripheral expansion was dominant, they could infer strong demand for supporting infrastructure in new urban areas; after 2016, however, peripheral expansion waned, and infill-oriented redevelopment of old urban areas, pipeline upgrades, and smart transportation facilities became more reasonable investment directions. This data-driven investment logic is becoming the new common sense in the infrastructure sector.

4. A Mirror Comparison of Global Resource-Based Cities

The predicament Panjin faces has numerous analogues worldwide. Germany's Ruhr region underwent a long adjustment of urban form after the decline of the coal and steel industries; Alberta's oil cities in Canada rose and fell repeatedly with oil price fluctuations; and multiple resource-based cities in northeast China have been seeking a way out amid population outflows and inefficient land use. The study cites the examples of the Ruhr and Alberta to show that this is not a problem unique to China, but a spatial legacy of industrial civilization.

The difference lies in policy tools and planning capacity. The Ruhr region gradually completed its transition from industrial landscape to knowledge economy through urban renewal and green corridor construction; Panjin, for its part, attempts to increase land use intensity without expansion through planning constraints and market guidance. A comparison of the two shows that when the main drivers of urban growth shift from resource extraction to services and innovation, spatial compactness instead becomes a competitive advantage.

For infrastructure investors, resource-based cities mean both risks and opportunities. On the one hand, fiscal capacity is limited and payment capacity is questionable; on the other hand, there is substantial demand in areas such as old urban area renewal, ecological restoration, and transportation improvement. What matters is identifying whether a city has already crossed the stage of the "resource curse" and entered diversified development. Panjin's infill expansion after 2016 can be seen as a signal of improved urban governance capacity.

5. What Infill Expansion Means for Infrastructure

Compared with peripheral expansion, infill expansion has a more complex impact on infrastructure. Peripheral expansion is usually accompanied by new road networks and expansion of water supply and power systems, with relatively low construction costs and ease of adopting modular standard designs. Infill expansion, by contrast, inserts new development parcels into existing built-up areas, requiring connection to old pipeline networks and placing higher demands on traffic carrying capacity and public service density.This also changes the prioritization of infrastructure financing. In the edge-expansion phase, land transfer fees can cover most development costs, including infrastructure; in the infill phase, however, land value appreciation gains shrink, and early-stage infrastructure investment must rely more on government bonds, PPP models, or capital from public utility enterprises. For a city like Panjin, completing pipeline network renewal, road rehabilitation, and disaster prevention facility upgrades under tight fiscal constraints is a more complex proposition than "building one more road."

The "decline in compactness" trend identified by the research also reminds planners that infill expansion does not automatically bring improved spatial efficiency. If infill parcels lack public services and employment support, the city may still fall into "infill sprawl"—that is, the land is filled in, but functions become fragmented. Therefore, infrastructure investment must be synchronized with industrial layout and employment center planning in order to generate genuine spatial benefits.

6. A Generalizable Monitoring Model, and China's "Saving for a Rainy Day"

The real value of this research method lies in its replicability. China has more than a hundred resource-based cities, and globally there are over a thousand resource-based cities dependent on a single industry. Traditional monitoring methods cannot support large-scale, high-frequency spatial diagnosis. The combination of deep learning and remote sensing offers a low-cost pathway.

For example, in Africa and Southeast Asia, many expanding cities lack cadastral information, making infrastructure investment highly risky. If a similar method could be used to generate a unified picture of urban impervious surface change, international development agencies and private investors would be able to identify effective demand more precisely. The methodological framework of this study will help build the knowledge infrastructure for global urban infrastructure investment.

Returning to Panjin itself. Over thirty years, this city added approximately 177 square kilometers of built-up area, and its expansion slowed markedly after 2016, shifting to internal infill. This is not a sign of urban decline, but rather a reflection of a maturing urban development logic. For observers, it is a micro-level case of Chinese cities shifting from "quantity" to "quality."

7. Conclusion: The Future of Infrastructure Lies in Understanding the City's "Morphological Grammar"

A city's expansion pattern is the spatial projection of its political-economic structure. Panjin's story tells of how a resource-based city shifted from relying on newly added land to tapping the value of the existing stock, and how, in an era of increasingly rational infrastructure investment, it used data to support decision-making. When the Chinese Academy of Sciences team's impervious surface data is translated into planning strategies, the city gains the capacity for "self-adaptation."

For the global infrastructure industry, the Panjin case reminds us: urban growth in the next phase will no longer rely on large-scale edge development, but on a comprehensive project based on data diagnostics, renewal of the existing stock, and refined financing. The competitiveness of infrastructure lies not only in construction capacity, but also in the ability to understand and predict changes in urban form. Panjin's thirty years are precisely the touchstone of that ability.