The microbial community is one of the key indicators for evaluating the health of alpine marsh wetlands, and understanding the composition and health of alpine wetland communities provides a scientific rationale for conservation and restoration efforts. Taking the alpine marsh wetlands in Qilian Mountain National Park, Qinghai Province, as the research object, 27 soil samples (0–30 cm depth) were collected in July 2024 from three types of wetlands: non-degraded (ND), low-level degraded (LD), and heavily degraded (HD). Using high-throughput sequencing, PICRUSt2 functional prediction, nonmetric multidimensional scaling (NMDS), and redundancy analysis (RDA), we analyzed the bacterial community structure and functional characteristics as well as the soil physicochemical properties across different degradation levels and soil depths. Pearson correlation analysis and RDA were used to identify key soil indicators influencing microbial community characteristics. The results showed that (1) compared to ND, the relative abundance of Acidobacteriota increased from 12.3% to 23.7%, and that of Pseudomonadota increased from 28.5% to 35.1% in HD wetlands. Meanwhile, the Shannon index rose from 5.31 in ND to 6.52 in HD, indicating significantly increased microbial community diversity and complexity with wetland degradation (p < 0.05). (2) Vertically, the six major primary metabolic functions gradually weakened with increasing soil depth in all three types of wetlands, the relative abundance of Proteobacteria decreased from 0 to 30 cm, and the α-diversity indices of soil bacteria also declined with depth. (3) Compared to ND, LD and HD showed significantly lower soil moisture content, organic matter, and total organic carbon (p < 0.05), while total potassium and pH increased significantly (p < 0.05). With increasing depth, total nitrogen significantly decreased across all degradation types (p < 0.05). Bacterial diversity, as measured by the Shannon and Simpson indices, showed a significant correlation with several soil properties (moisture, organic matter, total nitrogen, total potassium, cation exchange capacity, and total organic carbon; p < 0.05). Furthermore, pH emerged as a primary environmental driver shaping microbial community structure across different soil depths. These findings offer technical guidance and a theoretical framework for comprehending the degradation and restoration dynamics of alpine marsh wetland ecosystems in the Qilian Mountains.

Fractional Vegetation Cover (FVC) is a crucial indicator describing vegetation conditions and provides essential data for ecosystem health assessments. However, due to the low and sparse vegetation in alpine meadows, it is challenging to obtain pure vegetation pixels from Sentinel-2 imagery, resulting in errors in the FVC estimation using traditional pixel dichotomy models. This study integrated Sentinel-2 imagery with unmanned aerial vehicle (UAV) data and utilized the pixel dichotomy model together with four machine learning algorithms, namely Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), and Deep Neural Network (DNN), to estimate FVC in an alpine meadow region. First, FVC was preliminarily estimated using the pixel dichotomy model combined with nine vegetation indices applied to Sentinel-2 imagery. The performance of these estimates was evaluated against reference FVC values derived from centimeter-level UAV data. Subsequently, four machine learning models were employed for an accurate FVC inversion, using the estimated FVC values and UAV-derived reference FVC as inputs, following feature importance ranking and model parameter optimization. The results showed that: (1) Machine learning algorithms based on Sentinel-2 and UAV imagery effectively improved the accuracy of FVC estimation in alpine meadows. The DNN-based FVC estimation performed best, with a coefficient of determination of 0.82 and a root mean square error (RMSE) of 0.09. (2) In vegetation coverage estimation based on the pixel dichotomy model, different vegetation indices demonstrated varying performances across areas with different FVC levels. The GNDVI-based FVC achieved a higher accuracy (RMSE = 0.08) in high-vegetation coverage areas (FVC > 0.7), while the NIRv-based FVC and the SR-based FVC performed better (RMSE = 0.10) in low-vegetation coverage areas (FVC < 0.4). The method provided in this study can significantly enhance FVC estimation accuracy with limited fieldwork, contributing to alpine meadow monitoring on the Qinghai–Tibet Plateau.

Using microwave remote sensing to invert forest parameters requires clear canopy scattering characteristics, which can be intuitively investigated through scattering measurements. However, there are very few ground-based measurements on forest branches, needles, and canopies. In this study, a quantitative analysis of the canopy branches, needles, and ground contribution of Masson pine scenes in C-, X-, and Ku-bands was conducted based on a microwave anechoic chamber measurement platform. Four canopy scenes with different densities by defoliation in the vertical direction were constructed, and the backscattering data for each scene were collected in the C-, X-, and Ku-bands across eight incidence angles and eight azimuth angles, respectively. The results show that in the vertical observation direction, the backscattering energy of the C- and X-bands was predominantly contributed by the ground, whereas the Ku-band signal exhibited higher sensitivity to the canopy structure. The backscattering energy of the scene was influenced by the incident angle, particularly in the cross-polarization, where backscattering energy increased with larger incident angles. The scene’s backscattering energy was influenced by the scattering and extinction of canopy branches and needles, as well as by ground scattering, resulting in a complex relationship with canopy density. In addition, applying orientation correction to the polarization scattering matrix can mitigate the impact of the incident angle and reduce the decomposition energy errors in the Freeman–Durden model. In order to ensure the reliability of forest parameter inversion based on SAR data, a greater emphasis should be placed on physical models that account for signal scattering and the extinction process, rather than relying on empirical models.

Earthquakes can alter the structure of ecosystems rapidly. This study aimed to assess the impact of earthquakes on plateau ecosystems using the 2021 Maduo 7.4-magnitude earthquake as a case study. A plateau ecosystem assessment index (PEAI) was proposed based on the 3S-Integrated Valuation of Ecosystem Services and Trade-offs model (3S-InVEST) and the analytic hierarchy process (AHP). It was constructed using information on land cover, vegetation coverage, landscape pattern, habitat quality, and ecosystem service value to comprehensively evaluate the impact on plateau ecosystems. The findings revealed that: (1) After the earthquake, the overall PEAI decreased from 5.47 to 5.09, which was a decline of 6.94%. The value of ecosystem services decreased by RMB 5.5814 billion, and the degree of impact displayed a spatial gradient pattern that diminished with seismic intensity. (2) The area with intensity level X was the most affected and that with intensity level VII was the least affected, with a decline of 40.76% and 1.64% in PEAI, respectively. The spatial heterogeneity analysis revealed that land cover changes were the most drastic in Ngoring Lake, Gyaring Lake, and the river area in the north of the VII seismic intensity zone. (3) A big change in land cover was witnessed after the earthquake, whose direction was from lake wetland to sandy land, marshy wetland, and grassland, and from riverine wetland to marshy wetland and grassland. Among the various types of wetlands, marshy wetlands experienced the most significant decrease, followed by lake wetlands and riverine wetlands, with a reduction of 286.28 km2, 61.66 km2, and 30.82 km2, respectively. (4) The vegetation coverage showed structural degradation, with medium- and high-coverage areas decreasing by 6.03% and 3.16%, respectively. However, the ecosystem service value of grasslands increased against the trend, revealing their ecological resilience. The study identified land cover change, degradation of ecosystem service value, and vegetation decline as the core driving factors behind a decrease in PEAI. This assessment system provides a quantitative basis for prioritizing post-earthquake ecological restoration efforts on the plateau.

The abundant satellite data have enabled the study of the dynamics of forest logging and its corresponding carbon balance with remote sensing. Change detection techniques with moderate-resolution imagery have been widely developed. Yet the signal processing or machine learning methods are sample-dependent, lacking an understanding of spectral signals of forest growth and logging cycles, which is necessary to distinguish logging from other types of disturbance, and mechanism models addressing post-logging tree changes are too complex for parameter inversion. We therefore proposed an efficient physical-based model for spectral simulation of annual forest logging by coupling forest dynamic model ZELIG and the stochastic radiative transfer (SRT) model. The forest logging simulation was conducted and validated by Abies forest field data before and after logging in Wangqing County, Northeastern China (R² = 0.85, RMSE = 10.82 t/ha). The spectral changes in Abies forest stands with annual growth and varying logging intensities were simulated by the novel model. The annual Landsat-8 and Gaofen-1 fusion multispectral imagery of the study area from 2013 to 2016 was furtherly used to extract annual sequence spectral data of 350 forest plots and perform inversion of the annual difference in above-ground biomass (dAGB). With the inversion method combining the look-up table of the ZELIG-SRT model and the random forest regression, the retrieved dAGB of the 350 plots indicated consistency with the measured data on the whole (R² = 0.71, RMSE = 13.32 t/ha). The novel physical-based approach for AGB monitoring is more efficient than previous 3D computer models and less dependent on field samples than data-driven models. This study provides a theoretical basis for understanding the remote sensing response mechanism of forest logging and a methodological basis for improving forest logging monitoring algorithms.