Bare permafrost refers to permafrost with almost no vegetation on the surface, which is an essential part of the ecosystem of the Tibetan Plateau. An accurate extraction of the boundaries of bare permafrost is vital for studying how it is being impacted by climate change. The accuracy of permafrost and bare land distribution maps is inadequate, and the spatial and temporal resolution is low. This is due to the challenges associated with obtaining significant amounts of data in high-altitude and alpine regions and the limitations of current mapping techniques in effectively integrating multiple factors. This study introduces a novel approach to extracting information about the distribution of bare permafrost. The approach introduced here involves amalgamating a sample extraction method, the fusion of multi-source remote sensing information, and a hierarchical classification strategy. Initially, the available multi-source permafrost data, expert knowledge, and refinement rules for training samples are integrated to produce extensive and consistent permafrost training samples. Using the random forest method, these samples are then utilized to create features and classify permafrost. Subsequently, a methodology utilizing a hierarchical classification approach in conjunction with machine learning techniques is implemented to identify an appropriate threshold for fractional vegetation cover, thereby facilitating the extraction of bare land. The bare permafrost boundary is ultimately derived through layer overlay analysis. The permafrost classification exhibits an overall accuracy of 90.79% and a Kappa coefficient of 0.806. The overall accuracies of the two stratified extractions in bare land were 97.47% and 96.99%, with Kappa coefficients of 0.954 and 0.911. The proposed approach exhibits superiority over the extant bare land and permafrost distribution maps. It is well-suited for retrieving vast bare permafrost regions and is valuable for acquiring bare permafrost distribution data across a vast expanse. It offers technical assistance in acquiring extended-term data on the distribution of exposed permafrost on the Tibetan Plateau. Furthermore, it facilitates the elucidation of the impact of climate change on exposed permafrost.

Highly accurate data on urban and rural settlement (URS) are essential for urban planning and decision-making in response to climate and environmental changes. This study developed an optimal random forest classification model for URSs based on spectral–topographic–radar polarization features using Landsat 8, NASA DEM, and Sentinel-1 SAR as the remote-sensing data sources. An optimal urban and rural settlement boundary (URSB) extraction technique based on morphological and pixel-level statistical methods was established to link discontinuous URSs and improve the accuracy of URSB extraction. An optimal random forest classification model for URSs was developed, as well as a technique to optimize URSB, using the Google Earth Engine (GEE) platform. The URSB of Xining, China, in 2020 was then extracted at a spatial resolution of 30 m, achieving an overall accuracy and Kappa coefficient of 96.21% and 0.92, respectively. Compared to using a single spectral feature, these corresponding metrics improved by 16.21% and 0.35, respectively. This research also demonstrated that the newly constructed Blue Roof Index (BRI), with enhanced blue roof features, is highly indicative of URSs and that the URSB was best extracted when the window size of the structural elements was 13 × 13. These results can be used to provide technical support for obtaining highly accurate information on URSs.