In recent years, the low altitude economy has emerged as a pivotal sector within strategic emerging industries, characterized by its extensive industrial chain and high technological intensity. This study focuses on the Pearl River Delta (PRD) urban agglomeration, a key region in China for the development of the low altitude economy, where local industrial chains are relatively complete. Our research aims to elucidate the spatial distribution of low altitude economy manufacturing and the spatiotemporal correlation patterns between its industrial and innovation chains, providing theoretical insights and policy support for fostering integrated production-innovation development. By analyzing enterprise data up to February 2025, we employ advanced spatial analysis techniques to uncover the dynamics of this rapidly evolving sector.
The low altitude economy encompasses economic activities conducted within low-altitude airspace, leveraging aerial resources for various applications. In 2021, China’s national planning documents formally incorporated the low altitude economy, leading to the establishment of a dedicated administration in 2024 to accelerate its growth. The PRD region, with its robust manufacturing base, technological accumulation, and innovation capabilities, serves as an ideal case study for examining how production and innovation chains interact spatially. Our investigation addresses critical questions: Do enterprises in different segments of the low altitude economy exhibit spatiotemporal co-agglomeration? What patterns of leading or following relationships exist? And what factors influence these patterns?
To address these questions, we collected data on enterprises in the low altitude economy manufacturing sector, categorizing them into two key segments: components and aircraft. Enterprises were further classified as production-oriented or innovation-oriented based on whether they obtained high-tech enterprise certification. Data sources included Tianyancha and the High-tech Enterprise Certification Management Network, with manual verification to ensure accuracy. Our analytical framework integrates kernel density analysis, co-location quotient analysis with spatial-temporal windows, and unordered multinomial logistic regression models to explore spatial agglomeration, spatiotemporal correlations, and influencing factors.
Kernel density analysis is employed to visualize the spatial concentration of enterprises, particularly in the aircraft segment, which represents the core of the low altitude economy. The kernel density function estimates the probability density of spatial elements within a neighborhood, producing a smooth surface that reveals distribution trends. The formula for kernel density is given by:
$$f(x) = \frac{1}{nh} \sum_{i=1}^{n} K\left(\frac{x – X_i}{h}\right)$$
where $K$ is the kernel function, $h$ is the bandwidth, $n$ is the number of points, and $X_i$ are the spatial locations of enterprises. This method helps identify high-density zones, such as those in Guangzhou and Shenzhen, indicating core areas of low altitude economy activity.
To assess spatiotemporal correlations, we utilize the co-location quotient (CLQ) model with spatial-temporal windows. This approach measures the degree to which one type of enterprise is attracted to another in both space and time. The local indicator of co-location quotient (LCLQ) for a point $A_i$ relative to type $B$ is defined as:
$$LCLQ_{A_i \to B} = \frac{N_{A_i \to B}}{N_B / (N – 1)}$$
where $N_{A_i \to B}$ is the weighted average number of type $B$ points within the neighborhood of $A_i$, $N_B$ is the total number of type $B$ points, and $N$ is the total number of points. The weights $w_{ij}$ are computed using a Gaussian kernel function:
$$w_{ij} = \exp\left(-0.5 \times \frac{d_{ij}^2}{d_{ib}^2}\right)$$
Here, $d_{ij}$ is the distance between points $A_i$ and $j$, and $d_{ib}$ is the bandwidth distance. By incorporating temporal windows based on enterprise establishment dates, we classify spatiotemporal correlation patterns into four types: co-agglomeration, leading, following, and independent agglomeration. For instance, if $LCLQ_{A_i \to B, t_{i1}} > 1$ and $LCLQ_{A_i \to B, t_{i0}} > 1$, it indicates co-agglomeration; if $LCLQ_{A_i \to B, t_{i1}} > 1$ and $LCLQ_{A_i \to B, t_{i0}} < 1$, it signifies a leading pattern.
To analyze the factors influencing these patterns, we apply an unordered multinomial logistic regression model. The model expresses the log-odds of an enterprise belonging to a specific spatiotemporal pattern relative to a reference category. The equation is:
$$\ln\left(\frac{P(y_a = s | \mathbf{x})}{P(y_a = z | \mathbf{x})}\right) = \alpha_s + \sum_{k=1}^{K} \beta_{sk} x_k$$
where $P(y_a = s | \mathbf{x})$ is the probability of enterprise $a$ belonging to pattern $s$, $z$ is the reference pattern, $\alpha_s$ is the intercept, and $\beta_{sk}$ are coefficients for predictors $x_k$. Predictors include industrial zone等级, cluster type, distance to airports, land price, economic vitality, and enterprise size. Odds ratios are used to interpret the results, indicating how these factors affect the likelihood of different patterns.
Our analysis reveals that the spatial distribution of low altitude economy manufacturing in the PRD urban agglomeration exhibits a distinct dual-core structure centered on Guangzhou and Shenzhen. Both production-oriented and innovation-oriented enterprises are concentrated in various industrial functional zones and central urban areas. Kernel density maps highlight high-density clusters in national-level development zones and high-tech parks, such as the Guangzhou Science City and Shenzhen Bay area. The spatial pattern underscores the role of existing industrial infrastructure and policy support in shaping the low altitude economy landscape.
The spatiotemporal correlation analysis shows that production-oriented enterprises often lead innovation-oriented ones in the aircraft segment. Specifically, production-oriented firms tend to attract innovation-oriented firms over time, but the reverse is less common. This pattern suggests that production activities drive initial agglomeration, with innovation following suit. In terms of industrial chain dynamics, the components segment frequently follows the aircraft segment, but the aircraft segment shows a tendency toward independent agglomeration. This indicates a hierarchical relationship where core manufacturing activities dictate the spatial evolution of supporting industries.
For instance, in Guangzhou, production-oriented enterprises in the aircraft segment lead to the co-agglomeration of innovation-oriented firms, facilitated by local universities and research institutes. In Shenzhen, however, while production leads innovation, the latter does not significantly attract additional production activities, possibly due to policy barriers in high-cost areas. These findings highlight the nuanced interplay between production and innovation in the low altitude economy.
The regression results identify several key factors influencing spatiotemporal patterns. National-level industrial zones significantly promote co-agglomeration and leading patterns, with odds ratios exceeding 2.5 for various modes. Economic vitality, measured by POI density of commercial services, also plays a crucial role, with higher vitality increasing the likelihood of correlated patterns. Land price and distance to airports have moderate effects, while cluster type and enterprise size show varied impacts. The table below summarizes the odds ratios for significant predictors across different patterns:
| Variable | Co-agglomeration | Leading Pattern | Following Pattern | Independent Agglomeration |
|---|---|---|---|---|
| National Industrial Zone | 6.643*** | 2.627*** | 3.700*** | 3.810*** |
| Economic Vitality | 1.245*** | 1.728*** | 1.372*** | 1.775*** |
| Land Price | 1.177*** | 1.222*** | 1.176* | 1.180*** |
| Airport Distance | 0.778 | 0.478*** | 1.126 | 0.500*** |
| Enterprise Size | 0.381* | 1.064* | 1.102 | 1.246* |
Note: ***, **, * indicate significance at 1%, 5%, and 10% levels, respectively. Odds ratios greater than 1 indicate increased likelihood of the pattern.
These results underscore the importance of policy-driven industrial zones and vibrant economic environments in fostering the integrated development of the low altitude economy. For example, national-level zones provide infrastructure and incentives that attract both production and innovation activities, leading to synergistic clusters. Conversely, areas with lower economic vitality or inadequate infrastructure may struggle to achieve such integration.
In conclusion, the low altitude economy in the PRD urban agglomeration demonstrates initial stages of production-innovation and industrial-innovation chain co-agglomeration. However, the overall level of regional coordination requires enhancement. Policy recommendations include establishing specialized low altitude economy industrial parks, strengthening existing industrial zones, and supporting leading enterprises in collaboration with research institutions. By optimizing spatial布局 and fostering innovation ecosystems, the PRD can further solidify its position as a hub for the low altitude economy, driving sustainable economic growth and technological advancement.
The low altitude economy represents a transformative opportunity for regional development, and our study provides a framework for understanding its spatial dynamics. Future research could explore longitudinal data to track evolving patterns or compare the PRD with other regions to identify best practices. As the low altitude economy continues to expand, continued focus on integrated policies and infrastructure will be essential for maximizing its potential.