Winter warming is fast than summer warming on the Qinghai‐Tibet Plateau (QTP). However, no assessment of winter warming effects on permafrost has been attempted. Here we conducted hypothetical control experiments and used the Noah land surface model to evaluate the impacts of winter warming on the QTP permafrost. The results show that air temperature in winter (November–April) was increasing at a rate of 0.66 °C/decade during 1980s−2000s, over double that in summer (May–October). The mean annual ground temperature of permafrost increased by 0.13 °C/decade. The summer warming dominated the variations in thermal regime of permafrost before 2000. After that, the influence of winter warming on permafrost thermal regime has gradually grown and exceeded that of summer warming. Winter warming has amplified the thermal degradation of permafrost. Our findings reveal that alpine continuous permafrost on the northern QTP has experienced a prominent regional warming due to rapid winter warming since 2000.

Agricultural land extraction is an essential technical means to promote sustainable agricultural development and modernization research. Existing supervised algorithms rely on many finely annotated remote-sensing images, which is both time-consuming and expensive. One way to reduce the annotation cost approach is to migrate models trained on existing annotated data (source domain) to unannotated data (target domain). However, model generalization capability is often unsatisfactory due to the limit of the domain gap. In this work, we use an unsupervised adversarial domain adaptation method to train a neural network to close the gap between the source and target domains for unsupervised agricultural land extraction. The overall approach consists of two phases: inter-domain and intra-domain adaptation. In the inter-domain adaptation, we use a generative adversarial network (GAN) to reduce the inter-domain gap between the source domain (labeled dataset) and the target domain (unlabeled dataset). The transformer with robust long-range dependency modeling acts as the backbone of the generator. In addition, the multi-scale feature fusion (MSFF) module is designed in the generator to accommodate remote sensing datasets with different spatial resolutions. Further, we use an entropy-based approach to divide the target domain. The target domain is divided into two subdomains, easy split images and hard split images. By training against each other between the two subdomains, we reduce the intra-domain gap. Experiments results on the “DeepGlobe → LoveDA”, “GID → LoveDA” and “DeepGlobe → GID” unsupervised agricultural land extraction tasks demonstrate the effectiveness of our method and its superiority to other unsupervised domain adaptation techniques.

In recent years, convolutional neural networks (CNNs) have been widely used in hyperspectral image (HSI) classification. However, feature extraction on hyperspectral data still faces numerous challenges. Existing methods cannot extract spatial and spectral-channel contextual information in a targeted manner. In this paper, we propose an encoder–decoder network that fuses spatial attention and spectral-channel attention for HSI classification from three public HSI datasets to tackle these issues. In terms of feature information fusion, a multi-source attention mechanism including spatial and spectral-channel attention is proposed to encode the spatial and spectral multi-channels contextual information. Moreover, three fusion strategies are proposed to effectively utilize spatial and spectral-channel attention. They are direct aggregation, aggregation on feature space, and Hadamard product. In terms of network development, an encoder–decoder framework is employed for hyperspectral image classification. The encoder is a hierarchical transformer pipeline that can extract long-range context information. Both shallow local features and rich global semantic information are encoded through hierarchical feature expressions. The decoder consists of suitable upsampling, skip connection, and convolution blocks, which fuse multi-scale features efficiently. Compared with other state-of-the-art methods, our approach has greater performance in hyperspectral image classification.

Permafrost distribution maps are of importance for environmental assessment, climate system modeling, and practical engineering applications. The scarcity of forcing data and parameters often limits the uses of permafrost models over large areas. However, detailed data are often available in a few subregions through field investigations. In this study, we propose a novel approach for mapping permafrost distribution in a large and data-scarce area using an empirical model with subregion permafrost maps and satellite data as inputs. The surface frost number model (FROSTNUM) was re-inferred to include an extra soil parameter to represent the thermal and moisture conditions in soils. The optimal soil parameters were determined from the subregion maps of permafrost distribution through spatial clustering, parameter optimization, and the decision tree method. FROSTNUM was fed with satellite-derived ground surface freezing and thawing indices to map the permafrost distribution over the study area. The proposed approach was evaluated in the Gaize area on the Qinghai-Tibet Plateau, where intensive field studies have been done. The simulated permafrost distribution is consistent with a map of permafrost distribution made from borehole observations and field surveys in Gaize. Due to excellent accuracy, the approach is effective and can be used in large areas with limited data.