The rapid development of the Guangdong-Hong Kong-Macao Greater Bay Area has led to significant changes in land use patterns, which have far-reaching impacts on the regional ecology, economy, and social structure. In order to deeply understand these changes and formulate scientific and reasonable land management policies, based on a total of 31 periods of land use data from 1990 to 2020, adopts a method combining K-means and a convolutional neural network (CNN). Firstly, K-means is used to preliminarily classify land use data and provide accurate training labels. Secondly, a convolutional neural network is used to extract and classify features to achieve high-precision land use identification. Finally, the classification results and land use dynamic degree are integrated to quantitatively analyze the characteristics of land use change in the study area. The results show that: (1) The optimal effect is obtained when the land use data is divided into four categories. The silhouette coefficient for K-means is 0.73, and the classification accuracy of the convolutional neural network is 89.71%, fully verifying the effectiveness of this method. (2) In the past 30 years, the land use change in the study area can be divided into four typical stages, namely, the initial stage of rapid urbanization (1990—1997), the acceleration urbanization period (1998—2004), the stable development period (2005—2009), and the optimization and sustainable development period (2010—2020). Significant differences in land use patterns exist across these stages. (3) The characteristics of land use change in the study area are mainly reflected in the continuous reduction of farmland area, the rapid expansion of urban and construction land, while ecological land such as forests and water bodies, although remaining relatively stable, also face certain pressures and challenges. The study emphasizes the innovative application of machine learning technology in land use research, noting that the method can effectively identify land use pattern changes under limited labeled data, demonstrating its significant value in complex classification research. The research results not only deepen the understanding of land use change in the Guangdong-Hong Kong-Macao Greater Bay Area, but also deepen the understanding of land use change in the Guangdong-Hong Kong-Macao Greater Bay Area. Moreover, it provides a scientific basis and practical guidance for regional land use planning, management and sustainable development, which is of great significance for promoting regional coordinated development.

Considering the urgent need to construct a new pattern of Chinese-style urbanization development in Guangdong Province, this study, based on urban land use data and statistical panel data from 122 districts and counties in Guangdong Province from 2000 to 2020, this study establishes a quantitative model to analyze the spatiotemporal differentiation characteristics of urban land expansion. It further integrates the Optimal Parameter Geographical Detector (OPGD) and the Geospatial Convergent Cross Mapping (GCCM) model to investigate the driving mechanisms and corresponding optimization pathways of urban land expansion. The results show that: (1) From 2000 to 2020, the total urban land area in Guangdong Province continued to expand, increasing by 7.53%. The Pearl River Delta region experienced the most significant growth, exhibiting a spatial pattern with the Pearl River Delta region as the center, where urban land expansion gradually weakened towards the periphery. (2) The OPGD results reveal that the intensity change of urban land expansion is a result of multi-factor interactions. The explanatory power of two-factor interactions (64.2%-85.7%) is significantly higher than that of single factors. (3) The GCCM results indicate a significant bidirectional causal relationship between per capita GDP and urban land expansion intensity during the study period, suggesting a mutual driving effect. The model also identifies per capita GDP as the strongest predictor of urban land expansion intensity. Furthermore, analyses of causality among urban land expansion intensity, population urbanization, road network density, and the Remote Sensing-based Ecological Index (RSEI) show that urban land expansion drives population urbanization, while transportation networks and the ecological environment are also causal factors promoting urban land expansion. Therefore, in the future development process, how to simultaneously improve the quality of urbanization and promote urban-rural integration has become a key issue.

This study focuses on 57 county-level administrative units in Guangdong Province, using flow network analysis and system coordination coupling degree research methods to calculate the coupling degree between the transportation and economic systems, and explore the coordination relationship between factor flows and economic development. Research has found that: (1) The internal county-level units surrounded by Huizhou, Heyuan and Jieyang, as well as the internal county-level units of Zhanjiang City, form two relatively independent high-intensity traffic flow subgroups, which have a strong radiation effect on the traffic flow of surrounding county-level units. (2) There are 15 county-level units with coordinated traffic flow and economic development, 33 county-level units on the brink of imbalance or with slight imbalance, and 8 county-level units with severe imbalance. (3) The coupling coordination degree of traffic flow and economic development exhibits spatial heterogeneity. The coupling coordination degree of county-level units in the Pearl River Delta is relatively good, with economic development and traffic flow development promoting each other; while the coupling coordination degree of county-level units in eastern, western, and northern Guangdong is relatively low, and there is a certain degree of disorder between economic development and transportation system development. The article proposes differentiated transportation optimization strategies and suggestions to improve the long-term guarantee mechanism for county-level finance, thereby promoting high-quality economic development at the county level.