Hyperspectral imaging can capture abundant spectral information and reveal the spectral absorption properties of surface materials. Nevertheless, the trade-off in spatial resolution reduces its capacity to represent surface object textures and structures; hyperspectral super-resolution technology is a viable solution to this problem. Yet, mainstream supervised methods depend on low-scale training data and specific data distributions, restricting their generalization capability and practicality in real scenarios. Although unsupervised methods remove the reliance on training data, they still face suboptimal reconstruction quality due to the absence of reference images and inaccuracies in degradation process estimation. Furthermore, bridging the performance gap between simulated datasets and real-world applications remains challenging. To this end, we propose a semisupervised network that effectively couples unsupervised learning and supervised pre-training in multistage architecture, MCS-Net for short. The network consists of three key components: Degradation Information Estimation (DIE), Supervised Fusion Pre-training (SFP) at low-resolution, and Unsupervised Image Generation (UIG) at full-resolution. The MCS-Net first estimates deep degradation information from input image pairs using DIE. It then applies supervised learning in SFP to construct a pre-trained fusion function and its parameters from the input low-resolution data pairs and their fused outputs. Finally, the pre-trained parameters from the previous stage are used to initialize the fusion network of UIG, which is then fine-tuned under the guidance of degradation parameters estimated by DIE, enabling the network to process full-resolution images effectively. Ablation experiments validated the effectiveness of each component. Moreover, the proposed MCS-Net outperforms the existing SOTA methods across six evaluation metrics in the simulation experiments, and the experimental results on real satellite data further validate the outstanding image fusion performance of MCS-Net and its potential for practical applications.

In the context of global climate change and China’s dual-carbon strategy, this analysis examines how land-use transition is associated with land-use carbon effects in China’s resource-based cities. From the perspective of urban development stages, an analytical framework is built by linking development stage, land-use structure, and carbon source–sink structure. Using 262 resource-based cities from 2011 to 2023, we estimate land-use-related carbon emissions, carbon sequestration, and net land-use carbon effects with the carbon emission coefficient method and analyze their spatiotemporal patterns and driving factors using GeoDetector. The results show clear differences among city types. Mature cities form the largest group. Growth cities show the fastest expansion of impervious surfaces, while regenerative cities present signs of ecological recovery. This suggests that land-use transition is not simply the expansion of impervious surfaces, but a stage-dependent process of structural change. Land-use carbon effects also differ across stages. Mature cities maintain high and stable carbon-source effects. Growth cities exhibit increasing carbon-source effects, declining cities show reduced emissions but limited improvement in the carbon source–sink structure, and regenerative cities show improved carbon-sink capacity under ecological restoration. Overall, net land-use carbon effects follow a rise–decline–rebound pattern and show clear spatial heterogeneity and visually apparent clustering patterns. Population size has strong explanatory power, while interactions between socioeconomic and land-use factors further shape spatial differences. These results support stage-specific low-carbon transition strategies.