China's economic focus has shifted from high-speed growth to high-quality development. High-quality tourism development, crucial for sustainable economic and social progress, demands optimized industrial structure, efficiency, and ecological environment. Guangdong Province, a leading economic and tourism region in China, demonstrates significant growth in tourism revenue and tourist numbers but also reveals substantial disparities in tourism efficiency and development levels among its 21 cities. This study uses Guangdong Province as a case study to analyze these disparities and explore pathways to promote coordinated development and enhance regional competitiveness. The research addresses the need for a comprehensive understanding of tourism quality, encompassing both efficiency and development levels, which are often treated separately. It tackles limitations in existing research by incorporating both spatial and temporal dimensions in analyzing regional differences and spatial structures, recognizing the dynamic and interactive nature of tourism regions. The study aims to measure tourism development levels across Guangdong's cities, dissect tourism efficiency using DEA-BCC and Malmquist productivity index, and analyze the dynamic evolution of local tourism spatial structures using LISA time paths and spatiotemporal leaps. Finally, it constructs a coupling coordination model to examine the relationship between tourism efficiency and development levels.

Existing research on tourism efficiency has evolved from single-indicator evaluations (e.g., management, operational, transportation efficiency) to comprehensive assessments (e.g., ecological, poverty alleviation, regional efficiency). Methodologies have transitioned from qualitative to quantitative analyses using models like DEA, SBM Malmquist, and DEA-SNA. Studies have also shifted in scale, from national or provincial levels to more localized areas. While research on regional tourism differences and spatial structures is extensive, limitations exist. Studies often lack a spatial or temporal dimension, failing to capture the dynamic evolution of regional disparities. This study builds upon existing research by integrating both spatial and temporal perspectives to better understand the complexities of regional tourism development.

The study employs an exploratory spatiotemporal data analysis framework to characterize the structural characteristics of tourism efficiency and development level across 21 cities in Guangdong Province from 2000 to 2020. **1. Data Sources:** Data were primarily sourced from the "Guangdong Provincial Statistical Yearbook," "Tourism Yearbook," and "Statistical Bulletin of National Economic and Social Development." Missing data were supplemented using indicator smoothing methods. Vector data of administrative divisions were extracted from the National Bureau of Surveying and Mapping Geographic Information. **2. Tourism Efficiency Measurement:** The Data Envelopment Analysis (DEA-BCC) model was used to measure the tourism efficiency of each city. Input indicators included the number of employees in the tertiary industry (proxy for labor) and the number of 3A or higher-grade tourist attractions, star-rated hotels, and travel agencies (proxy for capital). Output indicators were total tourist arrivals and total tourism revenue (deflated using the consumer price index). **3. Tourism Development Level Measurement:** The tourism development level was measured using a composite index based on total tourism headcount and revenue, normalized to facilitate comparisons. **4. Dynamic Analysis (Malmquist Index):** The Malmquist index model decomposed total factor productivity changes into changes in pure technical efficiency, scale efficiency, and technological progress, revealing the dynamic changes in tourism efficiency. **5. Spatial Analysis (LISA):** Local Indicators of Spatial Association (LISA) were used to analyze the spatial patterns of tourism efficiency and development level. LISA time paths (length and curvature) and spatiotemporal leaps were calculated to understand the dynamic evolution of local spatial structures. The length of the LISA time path reflects the dynamic characteristics while curvature reflects the fluctuation characteristics of the local spatial structure. **6. Coupling Coordination Degree Model:** A coupling coordination degree model was constructed to assess the interaction and coordination between tourism efficiency and development level. This model quantitatively measures the degree of interaction between the two systems using a formula that incorporates both indices, and is categorized into five types: severe disorder, moderate disorder, basic coordination, moderate coordination, and high coordination. The coupling coordination degree was calculated using a coupled coordination model that combined the coupling degree and a comprehensive coordination index.

The study's key findings include: **1. Spatial Differences in Tourism Efficiency:** Significant spatial differences in comprehensive tourism efficiency were observed across the 21 cities, concentrated mainly in the Pearl River Delta (PRD) region. The overall average tourism efficiency showed a fluctuating downward trend. **2. Static Characteristics:** Scale efficiency played a supporting role in comprehensive efficiency, while technical efficiency exerted an influencing and constraining role. The tourism development level index varied significantly, with a majority of cities demonstrating room for growth. **3. Dynamic Characteristics:** Total factor productivity showed an overall increasing trend, but significant fluctuations were observed due to events like SARS, the financial crisis, and haze events. Technological progress was identified as the main driver of positive growth in tourism efficiency, while pure technical efficiency fluctuated significantly, indicating inefficient resource utilization. Scale efficiency remained largely unchanged. **4. LISA Time Path Analysis:** The local spatial structure of tourism efficiency was found to be less stable than the tourism development level. The PRD region showed higher LISA time path curvatures for both efficiency and development level compared to peripheral regions. **5. LISA Spatiotemporal Leap Analysis:** The local spatial structure of tourism efficiency was unstable, with cities easily changing their relative positions. In contrast, the local spatial structure of the tourism development level was more stable, indicating path dependence. **6. Coupling Coordination:** The overall coupling coordination degree between tourism efficiency and development level gradually improved from 2000 to 2020, shifting from severe disorder to basic coordination. High coupling coordination was concentrated in the PRD region. However, some cities with initially high levels experienced regression, highlighting the need for balanced regional development.

The findings address the research question by demonstrating the complex interplay between tourism efficiency and development level in Guangdong Province. The spatial and temporal variations highlight the need for regionally specific strategies to improve tourism quality and sustainability. The fluctuating trends in efficiency and the significant role of technological progress underscore the need for continuous innovation and investment in the sector. The unequal distribution of tourism efficiency and development levels emphasizes the limitations of relying solely on PRD-centric development strategies, revealing the need for strategies promoting balanced growth across the province. The uneven distribution of the coupling coordination degree reveals a lack of synergy in urban forces and the need for accelerating the construction of core city clusters and promoting collaborative tourism development.

This study contributes to understanding the complex relationship between tourism efficiency and development level in a dynamic spatial context. The significant spatial disparities and temporal fluctuations necessitate regionally tailored strategies for achieving high-quality tourism development. Future research could delve into the specific drivers of spatial heterogeneity and explore the effectiveness of policy interventions designed to promote balanced regional growth. Further refinement of the evaluation indices, considering sustainability, transportation, and policy factors, would enhance the robustness of future studies.

The study's limitations include the reliance on readily available data, potentially limiting the comprehensiveness of the analysis. The proxy indicators used for labor and capital inputs might not fully capture the nuances of tourism resource utilization. Future studies could benefit from more detailed and disaggregated data, allowing for a deeper investigation of the driving forces behind the observed trends.