Search

Abstract Extreme events with increasing frequency and intensity are significantly affecting the permafrost environment. Analysis using the ERA5-Land reanalysis data revealed that the permafrost region of the central Qinghai-Tibet Plateau (QTP) experienced the summer heat wave in 2022. Four active layer sites experienced maximum active layer thicknesses (ALT) in 2022 (mean: 207.7 cm), which was 20% higher than the mean ALT during 2000–2021 (mean: 175.9 cm). The mean annual ground temperature (MAGT) observed in 2022 was also the highest, exceeding the average of the previous years by 10%. The contribution fraction of heat wave to the seasonal thaw depth of active layer was quantified using Stefan model with ranging from 6.6% to 13.6%, and the maximum contribution fraction occurs in 2022. These findings are helpful to better understand the impact processes of extreme events on the active layer and permafrost.

Permafrost in Mongolia is located in the transition zone between high-latitude and high-altitude permafrost regions of the Northern Hemisphere, with large temperature differences and complex subsurface characteristics. In this study, the reliability of the skin temperature data from the ERA5-Land product covering Mongolia is assessed via site observations. The ERA5-Land skin temperature dataset shows a cold bias, which is more pronounced in the cold season. Following calibration based on elevation differences, significant improvements are observed at both the annual scale (92 % improvement in RMSE (root mean square error) and 98 % improvement in MBE (mean bias error)) and the seasonal scale (78 % improvement in RMSE and 82 % improvement in MBE). Additionally, the spatial variations in the surface freezing index (SFI) and surface thawing index (STI) are most pronounced in the central and northeastern Mongolia. The SFI exhibits a significant decreasing trend of 7.16 °C·d/year, while the STI shows a significant increasing trend of 4.49 °C·d/year. Furthermore, the permafrost extent in Mongolia is simulated from 1950 to 2022 using the frost number (Fn) model and the temperature on top of permafrost (TTOP) model. The validated results indicate that the accuracy of the Fn model is relatively high, with an overall accuracy of 0.9 and a Kappa coefficient of 0.47. The permafrost extent in Mongolia has declined from 734.7 × 103 km2 in the 1950 s to 480.1 × 103 km2 in the 2010 s, with a prominent decrease of 3.2 × 103 km2/decade after 1994. According to the variations in permafrost extent during past 72 years, the Hovsgol and Khentii Mountain ranges have experienced significant permafrost degradation.

Peatlands are major natural carbon pool in terrestrial ecosystems globally and are essential to a variety of fields, including global ecology, hydrology, and ecosystem services. Under the context of climate change, the management and conservation of peatlands has become a topic of international concern. Nevertheless, few studies have yet systematized the overall international dynamics of existing peatland research. In this study, based on an approach integrating bibliometrics and a literature review, we systematically analyzed peatland research from a literature perspective. Alongside traditional bibliometric analyses (e.g., number of publications, research impact, and hot areas), recent top keywords in peatland research were found, including ‘oil palm’, ‘tropical peatland’, ‘permafrost’, and so on. Furthermore, six hot topics of peatland research were identified: (1) peatland development and the impacts and degradations, (2) the history of peatland development and factors of formation, (3) chemical element contaminants in peatlands, (4) tropical peatlands, (5) peat adsorption and its humic acids, and (6) the influence of peatland conservation on the ecosystem. In addition, this review found that the adverse consequences of peatland degradation in the context of climate change merit greater attention, that peatland-mapping techniques suitable for all regions are lacking, that a unified global assessment of carbon stocks in peatlands urgently needs to be established, spanning all countries, and that a reliable system for assessing peatland-ecosystem services needs to be implemented expeditiously. In this study, we argued that enhanced integration in research will bridge knowledge gaps and facilitate the systematic synthesis of peatlands as complex systems, which is an imperative need.

Vegetation dynamics in Qinghai-Tibet Plateau (QTP) have been debated in recent decades. Most studies suggest that wetter and warmer climatic conditions would release low temperature constraints and stimulate alpine vegetation growth. Other studies suggest that climate warming might inhibit vegetation growth by increasing soil moisture depletion in the southern QTP. Most of previous studies have relied on vegetation indices derived from satellite observations to retrieve large-scale vegetation changes, and the uncertainty of vegetation indices makes it difficult to accurately characterize the vegetation trends on the QTP. Here, we developed a deep learning algorithm in the Google Earth Engine (GEE) platform to accurately map the land use/cover change (LUCC) on the QTP, and then infer vegetation gain and loss and their drivers during the period 1988–2018. The vegetation on the QTP experienced rapid greening, which was dominated by transitions from bareland to alpine grassland (27.45 × 104 km2) and from alpine grassland to alpine meadow (17.43 × 104 km2) during 1988–2018. Furthermore, although human activities influence vegetation succession at the local scale, the dominant influencing factors affecting vegetation greening on the QTP are precipitation (q-statistic = 23.87 %) and temperature (q-statistic = 11.01 %). A 30-year time series analysis clarified the specific dynamics of vegetation on the QTP, thus contributing to the understanding of the response mechanisms of alpine vegetation under climate change and providing a reference for the formulating of reasonable ecological protection policies and human development strategies.

The thawing of permafrost on the Qinghai-Tibet Plateau (QTP) leads to more frequent occurrences of thaw slump (TS), which have significant impacts on local ecosystems, carbon cycles, and infrastructure development. Accurate recognition of TS would help in understanding its occurrence and evolution. Machine learning capabilities for TS recognition are still not fully exploited. We systematically evaluate the performance of machine learning models for TS recognition from unmanned aerial vehicle (UAV) and propose an ensemble learning object-based model for TS recognition (EOTSR). The EOTSR has the following advantages: 1) pioneering the introduction of spatial information to assist in recognition; 2) the misclassification of recognition models is improved by object-based technology; and 3) attempting to integrate the strengths of different machine learning models to obtain a recognition accuracy no less than that of commonly used deep learning models. The results show that object-based technology is more suitable for TS recognition than pixel-based technology. Recursive feature elimination (RFE)-based feature selection proves that texture and geometry are effective complements to TS recognition. Among the improved object-based machine learning models, support vector machine (SVM) has the highest recognition accuracy, with an overall accuracy of 93.06 %. McNemar’s test proves that EOTSR significantly improves TS recognition compared to a single model and achieves an overall accuracy of 97.32 %. The EOTSR model provides an effective recognition method for the increasingly frequent TS events in the permafrost regions of the QTP, and can produce label data for deep learning models based on satellite imagery.

Abstract Background Fractional vegetation cover (FVC) is an important basic parameter for the quantitative monitoring of the alpine grassland ecosystem on the Qinghai-Tibetan Plateau. Based on unmanned aerial vehicle (UAV) acquisition of measured data and matching it with satellite remote sensing images at the pixel scale, the proper selection of driving data and inversion algorithms can be determined and is crucial for generating high-precision alpine grassland FVC products. Methods This study presents estimations of alpine grassland FVC using optimized algorithms and multi-dimensional features. The multi-dimensional feature set (using original spectral bands, 22 vegetation indices, and topographical factors) was constructed from many sources of information, then the optimal feature subset was determined based on different feature selection algorithms as the driving data for optimized machine learning algorithms. Finally, the inversion accuracy, sensitivity to sample size, and computational efficiency of the four machine learning algorithms were evaluated. Results (1) The random forest (RF) algorithm (R2: 0.861, RMSE: 9.5%) performed the best for FVC inversion among the four machine learning algorithms driven by the four typical vegetation indices. (2) Compared with the four typical vegetation indices, using multi-dimensional feature sets as driving data obviously improved the FVC inversion accuracy of the four machine learning algorithms (R2 of the RF algorithm increased to 0.890). (3) Among the three variable selection algorithms (Boruta, sequential forward selection [SFS], and permutation importance-recursive feature elimination [PI-RFE]), the constructed PI-RFE feature selection algorithm had the best dimensionality reduction effect on the multi-dimensional feature set. (4) The hyper-parameter optimization of the machine learning algorithms and feature selection of the multi-dimensional feature set further improved FVC inversion accuracy (R2: 0.917 and RMSE: 7.9% in the optimized RF algorithm). Conclusion This study provides a highly precise, optimized algorithm with an optimal multi-dimensional feature set for FVC inversion, which is vital for the quantitative monitoring of the ecological environment of alpine grassland.

Land degradation under the influence of global warming and ecological environmental destruction due to poor land management is the main challenge facing the Mongolian Plateau, and its future ecological risk change trends and drivers are also unclear. Therefore, to address the context relevant to this challenge, planning based on measured information from land use patterns is required. Based on land use and land cover (LULC), this study evaluates the landscape ecological risk (LER) of the Selenga River Basin by calculating the landscape pattern index. The spatiotemporal pattern and influencing factors of landscape ecological risk in the Selenga River Basin from 1990 to 2040 were analyzed. According to the results of LULC analysis, forest and grassland were the primary land use types in the Selenga River Basin. The built area, forest, and cropland showed an increasing trend, while the grassland area showed a fluctuating decreasing trend. From 1990 to 2010, the comprehensive land use dynamic degree showed a trend of rising first and then falling, specifically from 0.13% in 1990 to 0.29% in 2010, and will drop to 0.06% by 2040, indicating that the range of land use change is becoming more and more stable. The landscape ecological risk assessment shows a distribution pattern of “low at the edge and high in the middle”. The landscape ecological risk index (LER) first increases and then decreases, with the peak value in 2010 (0.085). By calculating the spatial aggregation of LER and the partial correlation with climate, we found that the Moran’s I index showed an “anti-V”-shaped change trend from 1990 to 2040, and the average landscape ecological risk presents positive spatial correlation, primarily with high-value aggregation, and peaked in 2010. Precipitation had a negative correlation with landscape ecological risk controlling for temperature, while there was a positive relationship between temperature and landscape ecological risk under the influence of controlling precipitation. This study provides a scientific basis for LULC planning in the Selenga River Basin, and is of great significance for maintaining the ecological security of the Mongolian Plateau.

High-precision canopy chlorophyll content (CCC) inversion for marsh vegetation is of great significance for marsh protection and restoration. However, it is difficult to collect the CCC measured data for marsh vegetation that matches the pixel scale of remote sensing image. This article proposes a new method based on unmanned aerial vehicle (UAV) multispectral images to obtain multiscale marsh vegetation CCC sample data. A random forest (RF) regression algorithm was used to evaluate the application performance of GF-1 wide field view (WFV), Landsat-8 Operational Land Imager (OLI), and Sentinel-2 multispectral instrument (MSI) satellite remote sensing data in marsh vegetation CCC inversion. In addition, parameter optimization of the RF regression model was used to construct an optimization algorithm suitable for marsh vegetation, and the importance of input variables was quantitatively evaluated. The results showed that the UAV multispectral images assisted in the acquisition of marsh vegetation CCC sample data, as the method expanded the number of CCC samples while quantifying the CCC sample data collection accuracy [R2 ≥ 0.86, root mean square error (RMSE) ≤ 6.98 SPAD], which improved the CCC inversion accuracy compared with traditional sampling methods. Extracting pure vegetation pixels through binary classification reduces the uncertainty of the UAV-scale CCC inversion results. Parameter optimization of the RF regression model further improves the CCC inversion accuracy at GF-1 WFV, Landsat-8 OLI, and Sentinel-2 MSI scales. Among the three satellite remote sensing data, Sentinel-2 MSI achieved the highest CCC inversion accuracy for marsh vegetation (R2 = 0.79, RMSE = 10.96 SPAD) due to the inclusion of red-edge bands that are more sensitive to vegetation properties. Red-edge Chlorophyll Index (Clred-edge) and Green Chlorophyll Index (Clgreen) have the highest influence on the CCC inversion accuracy among input variables.

The Mongolian Plateau is located in the permafrost transitional zone between high-altitudinal and high-latitudinal permafrost regions in the Northern Hemisphere. Current knowledge of the thermal state and changes in the permafrost on the Mongolian Plateau is limited. This study adopted an improved calculation method of the Mongolian Plateau air freezing and thawing index using the monthly air temperature reanalysis dataset from the Climate Research Unit (CRU). The spatial and temporal variation characteristics from 1901 to 2019 were further assessed by the Mann–Kendall (M–K) test and spatial interpolation methods. The results indicate that the spatial distributions of the freezing and thawing index show clear latitudinal zonality. Over the study period, the air freezing index decreased by 4.1°C·d/yr, and the air thawing index increased by 2.3°C·d/yr. The change point in the air thawing index appeared in 1995 (p < 0.05) based on the M–K method, in contrast to the so-called hiatus in global warming. Our results reveal rapid warming on the Mongolian Plateau, especially in the permafrost region, and are useful for studying permafrost changes on the Mongolian Plateau.

The majority of the Qinghai-Tibet Plateau (QTP) and Mongolia are underlain by permafrost. We have examined trends in air temperature and associated freezing/thawing index by using a non-parametric statistical method for the QTP and Mongolia from 1961 to 2011. The annual air temperature and associated freezing/thawing index exhibit similar patterns, suggesting similar warming trend in the two regions. The annual warming trends of air temperature are 0.33 °C/decade and 0.37 °C/decade in the QTP and Mongolia, respectively. The freezing index show significantly decreasing trends with −56.7 °C·days/decade and −57.5 °C·days/decade, while the thawing index present obvious increasing trends of 68.2 °C·days/decade and 68.3 °C·days/decade in the QTP and Mongolia, respectively. We find that the variations of air temperature and freezing/thawing index exhibit prominent spatial heterogeneity, and the warming trends is attributed to different seasonal warming. The warming trends in the QTP are dominated by winter warming, it is coincide with previous studies. Contrary to the QTP, autumn warming mainly accounts for the warming trends in the Mongolia. In addition, a winter cooling trend is observed in the Mongolia during the last two decades. These findings will be helpful to better understand the spatial heterogeneity of permafrost changes.

The increase in temperatures and changing precipitation patterns resulting from climate change are accelerating the occurrence and development of landslides in cold regions, especially in permafrost environments. Although the boundary regions between permafrost and seasonally frozen ground are very sensitive to climate warming, slope failures and their kinematics remain barely characterized or understood in these regions. Here, we apply multisource remote sensing and field investigation to study the activity and kinematics of two adjacent landslides (hereafter referred to as “twin landslides”) along the Datong River in the Qilian Mountains of the Qinghai-Tibet Plateau. After failure, there is no obvious change in the area corresponding to the twin landslides. Based on InSAR measurements derived from ALOS PALSAR-1 and -2, we observe significant downslope movements of up to 15 mm/day within the twin landslides and up to 5 mm/day in their surrounding slopes. We show that the downslope movements exhibit distinct seasonality; during the late thaw and early freeze season, a mean velocity of about 4 mm/day is observed, while during the late freeze and early thaw season the downslope velocity is nearly inactive. The pronounced seasonality of downslope movements during both pre- and post-failure stages suggest that the occurrence and development of the twin landslide are strongly influenced by freeze–thaw processes. Based on meteorological data, we infer that the occurrence of twin landslides are related to extensive precipitation and warm winters. Based on risk assessment, InSAR measurements, and field investigation, we infer that new slope failure or collapse may occur in the near future, which will probably block the Datong River and cause catastrophic disasters. Our study provides new insight into the failure mechanisms of slopes at the boundaries of permafrost and seasonally frozen ground.

The accurate classification of wetland vegetation is essential for rapid assessment and management. The Honghe National Nature Reserve (HNNR), located in Northeast China, was studied. The multi-scale remote sensing data of a new generation of Chinese high-spatial-resolution earth observation satellites Gaofen-1 (GF-1), Gaofen-2 (GF-2), Ziyuan-3 (ZY-3), and international earth observation satellites Sentinel-2A and Landsat 8 OLI were selected as sources. Based on the DeepLabV3 Plus deep learning model, 12 intelligent marsh vegetation classification models were constructed. We quantitatively analyzed the applicability and identification ability of DeepLabV3 Plus for classifying complex marsh vegetation. We discuss the differences in accuracy of marsh vegetation classification with different remote sensing data sets. The spatial resolution of remote sensing data sets ranges from 30 m to 0.8 m, and spectral bands range from blue bands (450 nm) to shortwave infrared bands (2280 nm). The specific conclusions of this study are as follows: (1) The DeepLabV3 Plus model better identified marsh vegetation, but there were significant differences in the classification accuracy of 12 DeepLabV3 Plus intelligent identification models. (2) Under the same conditions of the spectral bands (four Blue ~ NIR bands), the accuracy of deep-water marsh vegetation classification gradually increased as spatial resolution improved. For shallow-water marsh vegetation, when the accuracy of vegetation classification increased to a certain level, the classification accuracy decreased with the improvement of spatial resolution, which indicated that high-resolution images reduced pixel mixing to a certain extent, but for some vegetation types, the internal spectral difference increased, which made classification more difficult. (3) The increase of spectral bands improved the classification of marsh vegetation, while the classification accuracy of models with spectral indices was better than that of models only including spectral bands. (4) The accuracy of marsh vegetation classification was greatly improved by combining spectral bands and spectral indices. (5) The classification of the five sensor satellite images had statistical differences between models with different spatial resolutions and models with different spectral ranges.

Discriminating marsh vegetation is critical for the rapid assessment and management of wetlands. The study area, Honghe National Nature Reserve (HNNR), a typical freshwater wetland, is located in Northeast China. This study optimized the parameters (mtry and ntrees) of an object-based random forest (RF) algorithm to improve the applicability of marsh vegetation classification. Multidimensional datasets were used as the input variables for model training, then variable selection was performed on the variables to eliminate redundancy, which improved classification efficiency and overall accuracy. Finally, the performance of a new generation of Chinese high-spatial-resolution Gaofen-1 (GF-1) and Ziyuan-3 (ZY-3) satellite images for marsh vegetation classification was evaluated using the improved object-based RF algorithm with accuracy assessment. The specific conclusions of this study are as follows: (1) Optimized object-based RF classifications consistently produced more than 70.26% overall accuracy for all scenarios of GF-1 and ZY-3 at the 95% confidence interval. The performance of ZY-3 imagery applied to marsh vegetation mapping is lower than that of GF-1 imagery due to the coarse spatial resolution. (2) Parameter optimization of the object-based RF algorithm effectively improved the stability and classification accuracy of the algorithm. After parameter adjustment, scenario 3 for GF-1 data had the highest classification accuracy of 84% (ZY-3 is 74.72%) at the 95% confidence interval. (3) The introduction of multidimensional datasets improved the overall accuracy of marsh vegetation mapping, but with many redundant variables. Using three variable selection algorithms to remove redundant variables from the multidimensional datasets effectively improved the classification efficiency and overall accuracy. The recursive feature elimination (RFE)-based variable selection algorithm had the best performance. (4) Optical spectral bands, spectral indices, mean value of green and NIR bands in textural information, DEM, TWI, compactness, max difference, and shape index are valuable variables for marsh vegetation mapping. (5) GF-1 and ZY-3 images had higher classification accuracy for forest, cropland, shrubs, and open water.