Search

Abstract Coastal areas have the most obvious ocean–land interaction and experience the most frequent human activities. As the development of coastal areas has a high degree of spatiotemporal variability, local governments bear direct responsibility for marine governance, yet accurately evaluating and analyzing local governments’ marine governance efficiency in coastal areas is challenging. This study constructs a spatiotemporal coupling coordination model to comprehensively evaluate local governments’ marine governance efficiency in six coastal cities in Liaoning Province from 2004 to 2019. A complex system was necessary to obtain the development level, discrete degree, and development speed of each subsystem. The construction of the evaluation index system was the foundation, and the construction of the spatiotemporal weight matrix was the key. The results show that overall, the local governments’ marine governance efficiency level is generally increasing, and the agglomeration effect is obvious. The efficiency of each cities’ economic, ecological, and social governance subsystem is in a process of continuous and dynamic change. The coupling and coordination degrees of the six governance systems have continuously improved and the spatial and temporal differences have decreased; each city shows different coupling and coordination degrees in each subsystem. Regarding the factors affecting comprehensive marine management, Liaoning’s coastal areas fail to attract foreign tourists; the discharge and treatment of industrial wastewater restricts ecological governance; and the reduction of fisheries hinders the social governance system’s efficiency. The results contribute to the understanding of costal cities’ marine governance and promote the sustainable development of coastal areas.

Data mining student information helps to understand the compulsory education population size, structure, distribution and flow characteristics. Here, the spatial evolution, regional disparity, and emerging spatiotemporal distribution patterns are analyzed based on statistical data of primary and secondary education in Liaoning Province from 2010 to 2020. Demographic and spatial analysis methods, population size and structural changes were used in the assessments. The scale of primary and secondary school students in Liaoning Province declined, the gender imbalance was alleviated, and the proportion of migrant students increased. The distribution of students in primary and secondary schools is clearly unbalanced, and an increasingly evident central direction of flow from the countryside to the city exists. The overall difference in distribution of primary and secondary school students in Liaoning Province exhibits a trend of narrowing and then expanding, which is largely due to the uneven distribution among the non-agricultural municipal district, agriculture-related municipal district, and agriculture-related county. The emerging hot spot analysis patterns were dominated by consecutive and intensifying cold spots, spatiotemporal persistent and intensifying hot spots. Overall, these educational statistics and spatial analysis results provide important insights into population and educational geography.

The long-term effects of administrative division adjustments on economic development in Dalian City, China, is presently unclear. Therefore, we employed a geographically and temporally weighted regression model to construct an index assessment system using the annual number of newly registered enterprises as the dependent variable. We empirically studied the effects of administrative division adjustments of townships in Dalian City from 2010 to 2020. The results showed that: (i) adjustments in townships administrative divisions negatively affected regional enterprise concentrations and economic development in general, whereas industrial and transportation factors demonstrated positive effects to variable degrees, and the effects of policy, location, and population factors were not significant. (ii) Over time, the effects of townships administrative division adjustments changed from positive to negative, and the negative effects displayed an increasing trend, revealing that (iii) the effect of administrative division adjustments in Dalian City was positive and negative over time. This further reveals the objective problem that administrative division adjustment in recent years was effective over the short term, but not over the long term. (iv) During the study period, administrative division adjustments positively affected regional enterprise concentrations and economic development solely during a short time frame and in a small geographical area. However, its negative effects increased extensively, which infers that maintaining a relatively stable administrative division is more beneficial for promoting economic development in Dalian City.

Vision Transformers (ViTs) have demonstrated exceptional accuracy in classifying remote sensing images (RSIs). However, existing knowledge distillation (KD) methods for transferring representations from a large ViT to a more compact Convolutional Neural Network (CNN) have proven ineffective. This limitation significantly hampers the remarkable generalization capability of ViTs during deployment due to their substantial size. Contrary to common beliefs, we argue that domain discrepancies along with the RSI inherent natures constrain the effectiveness and efficiency of cross-modal knowledge transfer. Consequently, we propose a novel Variance Consistency Learning (VCL) strategy to enhance the efficiency of the cross-modal KD process, implemented through a plug-and-plug module within a ViTteachingCNN pipeline. We evaluated our student model, termed VCL-Net, on three RSI datasets. The results reveal that VCL-Net exhibits superior accuracy and a more compact size compared to 33 other state-of-the-art methods published in the past three years. Specifically, VCL-Net surpasses other KD-based methods with a maximum improvement in accuracy of 22% across different datasets. Furthermore, the visualization analysis of model activations reveals that VCL-Net has learned long-range dependencies of features from the ViT teacher. Moreover, the ablation experiments suggest that our method has reduced the time costs in the KD process by at least 75%. Therefore, our study offers a more effective and efficient approach for cross-modal knowledge transfer when addressing domain discrepancies. [ABSTRACT FROM AUTHOR]