Considering the complexity of landslide hazards, their manual investigation lacks efficiency and is time-consuming, especially in high-altitude plateau areas. Therefore, extracting landslide information using remote sensing technology has great advantages. In this study, comprehensive research was carried out on the landslide features of high-resolution remote sensing images on the Mangkam dataset. Based on the idea of feature-driven classification, the landslide extraction model of a fully convolutional spectral–topographic fusion network (FSTF-Net) based on a deep convolutional neural network of multi-source data fusion is proposed, which takes into account the topographic factor (slope and aspect) and the normalized difference vegetation index (NDVI) as multi-source data input by which to train the model. In this paper, a high-resolution remote sensing image classification method based on a fully convolutional network was used to extract the landslide information, thereby realizing the accurate extraction of the landslide and surrounding ground-object information. With Mangkam County in the southeast of the Qinghai–Tibet Plateau China as the study area, the proposed method was evaluated based on the high-precision digital elevation model (DEM) generated from stereoscopic images of Resources Satellite-3 and multi-source high-resolution remote sensing image data (Beijing-2, Worldview-3, and SuperView-1). Results show that our method had a landslide detection precision of 0.85 and an overall classification accuracy of 0.89. Compared with the latest DeepLab_v3+, our model increases the landslide detection precision by 5%. Thus, the proposed FSTF-Net model has high reliability and robustness.

Class imbalance is a key issue for the application of deep learning for remote sensing image classification because a model generated by imbalanced samples training has low classification accuracy for minority classes. In this study, an accurate classification approach using the multistage sampling method and deep neural networks was proposed to classify imbalanced data. We first balance samples by multistage sampling to obtain the training sets. Then, a state-of-the-art model is adopted by combining the advantages of atrous spatial pyramid pooling (ASPP) and Encoder-Decoder for pixel-wise classification, which are two different types of fully convolutional networks (FCNs) that can obtain contextual information of multiple levels in the Encoder stage. The details and spatial dimensions of targets are restored using such information during the Decoder stage. We employ four deep learning-based classification algorithms (basic FCN, FCN-8S, ASPP, and Encoder-Decoder with ASPP of our approach) on multistage training sets (original, MUS1, and MUS2) of WorldView-3 images in southeastern Qinghai-Tibet Plateau and GF-2 images in northeastern Beijing for comparison. The experiments show that, compared with existing sets (original, MUS1, and identical) and existing method (cost weighting), the MUS2 training set of multistage sampling significantly enhance the classification performance for minority classes. Our approach shows distinct advantages for imbalanced data.

Industrial heat sources have made a great contribution to Chinese economic development. However, it has also been found that emissions from industrial heat sources are the main contribution to regional air pollution. Therefore, the detection of industrial heat sources and the expression of related information is becoming important. In this paper, the detection of industrial heat sources was used to express industrial information, thus that the accuracy of the detection of industrial thermal anomalies could be improved and the problems of noise and missing parameters addressed. A model for expressing industrial information based on object-oriented industrial heat sources and using multi-source thermal anomaly data in China was, therefore, proposed. It was a new real-time, objective, and real way to describe the production operation status of industrial heat sources on a large-scale area. First, 4340 working industrial heat sources in mainland China were detected by applying an adaptive k-means algorithm to ACF (NPP VIIRS 375-m active fire/hotspot data) data from the period 19 January 2012 to 31 December 2020. Secondly, several features of working industrial heat sources were extracted from NPP VIIRS 375-m active fire/hotspot data (ACF), VIIRS Nightfire data (VNF), and the Fires product based on Landsat-8 AIRCAS (L8F) data. Areas containing working industrial heat sources were then identified based on these different types of fire data. Light, land-surface temperature, and CO₂ and N₂O emissions data related to the working industrial heat sources were also extracted. The results show that feature parameters extracted from the multi-source thermal anomaly data mostly have a good positive correlation with the other parameters.

Meteorological hazards threaten sustainable development by affecting human safety, economic stability, and food security. Climate change increases extreme weather frequency, underscoring the urgency for comprehensive evaluation frameworks. However, existing frameworks rarely integrate multiple impact dimensions, limiting their practical utility. To address this gap, our core objective is to develop two novel index series, a single-hazard composite impact index (SHCI) and a multi-hazard composite impact index (MHCI), employing entropy weighting to integrate demographic and economic factors, enabling a more holistic assessment of meteorological hazard impacts in China. Analysis of 2004–2021 data on drought, rainstorm and flood (RF), hail and lightning (HL), typhoon, and low-temperature freezing (LTF) revealed decreases in the national MHCI and SHCI. Key results include the following: (1) the relative MHCI decreased by 74.8%, exceeding 61.21% of absolute MHCI; (2) nationally, 2010, 2013, and 2016 had high MHCI values, and Sichuan has the most extreme hazard years (three) among all the provinces; and (3) provincially, Ningxia has the highest absolute and relative MHCI, while SHCIs varied spatially. These findings provide specific references for climate adaptation planning and the optimization of hazard risk reduction strategies. The methodology offers a versatile framework for multi-hazard risk assessment in nations experiencing climatic and demographic transitions.