High Asia is one of the regions with the most concentrated distribution of glacial lakes in the world, and the use of remote sensing technology to carry out glacial lake research in this region is of great significance for global change analysis and natural disaster assessment. This paper refers to a large number of domestic and foreign research literature and reports, comprehensively reviews the development process of data sources and information extraction methods for remote sensing data extraction of high Asian glacial lakes, and further analyzes the spatial and temporal changes of glacial lakes and their responses to global changes. The current research progress and main achievements of the research on the high Asian glacial lake in China and abroad are analyzed in detail. Finally, combined with the latest development of remote sensing mechanism, image processing technology and remote sensing data source, the development trend of high-spatial-resolution remote sensing in glacial lake related research fields is predicted.
Water is a very important resource, and it is an important material basis for the survival and development of all human beings and organisms. Water extraction can result to easily understand the general situation of existing water resources, thus being conducive to the rational planning and management of water resources and having a significant impact on human life and social activities. Traditional artificial methods are time-consuming and laborious, and therefore satellite remote sensing data is now used to extract water parameters such as water position, area, shape and river width, which has become an effective method and means to obtain water parameters quickly. On the basis of extensive literature research, this paper illustrates the basic ideas of water extraction of remote sensing image and its development course as well as the basic method and current situation of water extraction performed by experts, and makes a comprehensive review and analysis of the advantages and disadvantages of various methods so as to explain the problems of water extraction and research prospect, make the readers understand the situation of this field and provide some ideas for the study in this field.
The equality of ecological environment has been severely affected by black-odor water bodies, and hence strengthening the treatment of black-odor water bodies is an important task for aquatic environment management. Macro-monitoring of black-odor water bodies is the prerequisite for governance, and remote sensing technology has huge advantages in the field of macro-monitoring. There have been very insufficient studies on black and odorous water bodies. This paper systematically summarizes the current status of identification and recognition of black-odor water bodies, mainly analyzes the optical characteristics of black-odor water bodies from the three identification characteristics of reflection spectrum, watercolor, and inherent optical quantity, summarizes the recognition algorithms and the problems of these algorithms, which include the low versatility of the algorithm, the inaccurate reflectance caused by the problem of atmospheric correction features, and the overlapping features of different types of water recognition features. The future development trends are predicted: ① mining recognition characteristics; ② performing classification of reflection spectrum; ③ applying of machine learning algorithms.
At present, the remote sensing identification of urban black-odor water bodies is in the preliminary stage of algorithm; due to the influence of water depth, shadow and other factors, the accuracy is low in practical applications, and there is little research on the long-term dynamic monitoring of black-odor water bodies. In this study, the Jiujiang District of Wuhu was chosen as a research area to analyze the causes and apparent characteristics of black-odor water bodies. For single-band threshold method, band difference method, normalized index method and slope index method, threshold correction was performed based on GF-2 images, the accuracy was evaluated, combined with the visual interpretation of the black-odor water bodies for dynamic monitoring at the same time. The results are as follows: ① The occurrence of black and odor in the water body is usually accompanied by features such as color abnormality, river siltation, and secondary environmental problems. ②The band difference method has the best recognition effect in the single recognition algorithm, and the total accuracy is 87.5%. ③ The high spatial resolution feature of GF-2 improves the efficiency and accuracy of visual interpretation, which can effectively reduce the interference of water depth and building shadows on its remote sensing recognition; compared with the use of a single algorithm, it further improves the recognition accuracy and reliability of dynamic monitoring. ④ The four GF-2 images from 2014 to 2020 were used to extract the areas of black-odor water bodies in the main urban area of Jiujiang District, which are 0.313 km2, 0.152 km2, 0.069 km2, and 0.008 km2 respectively. The results show that the black and odor phenomenon in the water body of Jiujiang District has been gradually improved, but the black and odor phenomenon in the water system of Shenshan Park is still serious.
The waterline method is an effective way to obtain a large area of silty tidal flat terrain. Accurate extraction of the waterline is the key to the construction of the tidal flat digital elevation model (DEM). Affected by the tidal conditions and the surface of the tidal flat, the waterlines in different satellite images have large differences in spectrum and texture. It is difficult to extract waterlines accurately using a single method. In this paper, a tidal flat DEM construction method based on multi-algorithm waterline extraction is proposed. The waterlines are classified into four categories according to the tide conditions and are extracted by edge detection, threshold segmentation, object-based segmentation, and improved watershed algorithm respectively. Then, combined with the instantaneous tide level of waterlines, the tidal flat DEM is constructed. In this paper, the method was verified using Laizhou Bay as the research area. There was a high correlation between the inversion elevation and the measured elevation, R 2=0.86, and relative error was between 0.31~0.78 m. It is shown that the method in this paper can effectively obtain the approximate topography of the tidal flat.
The extraction of urban built-up areas plays an important role in urban development planning. To find out the method of extracting remote sensing image urban built-up area based on convolutional neural network which can balance efficiency and recognition accuracy, the authors started with the principle of neural network structure and compared as well as analyzed the internal structure of multiple semantic segmentation networks. The semantic segmentation network was trained separately and the results were comparatively studied. The experimental result shows that the ShelfNet-50 network could ensure high recognition accuracy while training speed, achieved 77% foreground segmentation accuracy while training time was only 14 hours, and the result of ShelfNet-50 network prediction was also highly consistent with the corresponding remote sensing image data. The experiment confirms that ShelfNet-50 network can be applied to high-resolution remote sensing image urban built-up area extraction problems.
Urban green spaces are important ecological resources in cities; therefore, quantitative assessment of these green space resources as well as establishment of monitoring system at multiple scales is urgently required for assisting natural resources management and eco-city construction. The objectives of this study are to summarize major methods used to assess and monitor two typical urban green space resources, i.e., vegetation and water bodies, in terms of quantity, quality, and ecosystem service value, and to discuss advantage and disadvantage of these methods. Some results have been obtained: ① Although traditional sampling methods can obtain quantitative information for urban vegetation, fragmentation and patch of urban vegetation has limited scaling such information to larger scales; ② Satellite remote sensing (RS), which can provide information such as spatial distribution, area, vegetation classification, and water quality, is an effective method to assess and monitor urban green spaces; nonetheless, detailed information, such as biomass and water volume, requires high spatial resolution (e.g., < 5 m) RS data as well as corresponding methods to process the data; ③ Unmanned aerial vehicle (UAV) can provide land surface information at high spatial resolution (e.g., < 5 cm); however, UAV has limitations, such as limited data coverage and challenged data processing; ④ Lots of studies focus on the relationship between urban green spaces and urban heat islands, but the mechanism, i.e., how much energy is consumed by evapotranspiration and its impact on cooling effect, is less focused, which is likely due to a relatively low spatial resolution of available thermal infrared RS data. In summary, there are still lots of challenges in assessing and monitoring nature resources, including urban green spaces.
Acquisition of surface features of the mining area is greatly helpful to safe mining operation and management. In this paper, the authors propose an object-oriented combined with deep-learning classification method to extract surface features of the mining area based on unmanned aerial vehicle (UAV) images. Firstly, images are segmented by object-oriented method with manual correction to make annotation data set for deep learning models. Secondly, prepared training image data set is used to train 3 deep learning models (FCN-32s, FCN-8s and U-Net) and obtain 3 trained deep learning models respectively. Thirdly, classification accuracy is improved, and 2 integrate algorithms, which are majority voting algorithm and scoring algorithm based on these deep learning models, are proposed. The experimental results show that, compared with the single object-oriented classification method, the proposed methods have higher surface feature extraction accuracy and higher Kappa coefficient, from which the scoring integrate model has the best recognition effect. The overall accuracy of feature extraction on the testing image data set is 94.55%, which is 5.96 percentage points higher than the single object-oriented classification method, with the Kappa coefficient being 0.819 1.
In close range oblique photography, there are problems of structural adhesion and distortion of 3D model caused by dead angle of aerial photography. In this study, based on a large number of close range images, the technology uses image processing techniques such as feature line matching, point cloud matching, texture mapping to fill in the blind area generated by oblique photography, so as to further optimize the structure and texture of the corresponding near earth model, realize the fine reconstruction, and solve the problem that it is difficult for the oblique photography model to browse near the ground. Comparative experiments show that the proposed method can improve the effect of close range correction to a certain extent and improve the visualization effect of oblique photography digital 3D imaging, thus providing reference for obtaining high-quality and high-precision 3D real scene model.
Land surface temperature is a key parameter in the study of the balance of water and heart between land surface and atmosphere. Obtainment of land surface temperature under all-weather conditions is very important. Although thermal infrared remote sensing technology can retrieve land surface temperature with high spatial resolution and full space coverage in cloud-free sky, the missing data in cloudy sky limit the all-weather applications of land surface temperature in some areas. This study develops two methods for reconstructing missing land surface temperature in cloudy skies. One of the methods is a space-time matched interpolation method helped with dataset of lands surface temperature assimilation. The other method is by data interpolating empirical orthogonal function (DINEOF), which is already popular in reconstruction of sea surface parameters but is rarely used in reconstruction of land surface parameters. The two methods are evaluated by both remotely sensed data and ground measured data in 2017, and the results demonstrate that both of them are adaptable in all seasons and all over China. The accuracies of two methods are very close and located between 2.5 and 3.5 K in cloudy conditions in four seasons in China. This study aims to give some useful references in the study of obtainment of land surface temperature under all-weather conditions.
Image segmentation is a key step in object-oriented analysis of high resolution images and plays an important role in information extraction accuracy. In order to improve the segmentation performance of object-oriented algorithms for high-resolution remote sensing images, this paper proposes a segmentation method (PCSLIC-MW) to improve the superpixel and marker watershed, including feature fusion, superpixel initial segmentation, and control marker watershed segmentation. In the phase of superpixel segmentation, a new distance measure calculation rule is proposed, which combines color space, spatial position information and phase consistency texture feature. And then the gray value of each patch is calculated after superpixel segmentation, image reconstruction after segmentation, and morphological extension technology is used to extract local minimum (H-minima) so as to control the number of segmentation regions. The over-cutting produced by the traditional mathematical morphologic watershed segmentation algorithm is optimized and improved. The reconstructed image is conducted by Gaussian filter, and then the control marker watershed algorithm is used to re-segment the reconstructed image. For experiment, ZY3-02 satellite image and airborne aerial image are adopted to verify the proposed method, the precision and recall rate are used to evaluate the segmentation accuracy, and the results are compared with those of other segmentation methods to prove the segmentation effectiveness of the proposed method.
Cloud, snow and fog are important factors affecting the quality of optics remote sensing images, and hence researchers should detect the range of cloud, snow, fog in remote sensing images and remove unwanted images so as to improve the utilization of remote sensing images. In this paper, the authors studied the method based on Random Forest to detect cloud, snow, fog and tried to reduce the error detection rate by means of adding a “second detection”. Experiments show that this method has high detection accuracy and efficiency.
Cloud and cloud shadow detection is an important part in the production of Landsat images. In recent years, deep learning has greatly improved the accuracy of cloud detection in Landsat images. However, deep convolutional neural network model training relies on a large scale of labeled images, and it is necessary to manually label each pixel as clearness, cloud or cloud shadow. Manually labeling is rather costly and time-consuming, which is not conducive to train practical models. Inspired by weakly supervised learning, this paper proposes a new deep learning method for cloud and cloud shadow detection. Firstly, conventional cloud detection algorithm CFMask is used to detect cloud and cloud shadow in Landsat images. Then, the results are used to replace the manually labeled images to train the deep convolutional neural network model for cloud detection. Finally, the model is used to detect the cloud and its shadow in new images. Experimental results show that the overall accuracy of the proposed method is 85.55%, which is better than that of CFMask and indicates that it is feasible to train the deep network model to detect cloud and cloud shadow without manually labeled data.
After observing a large number of aerial images, it is found that the effect is not ideal and the contrast is still not high. In this paper, through the study of the dark channel prior defogging algorithm, the process of fog image degradation is analyzed, and an aerial image defogging effect optimization method based on the dark channel prior is proposed. When the original image is uneven, the method of enhancing the contrast of atmospheric transmittance layer is used to improve the quality of the output image. In addition, for all the input images with fog, an image processing method of automatic contrast or automatic color enhancement is used to enhance the brightness of the output image. The optimization algorithm uses the objective image quality evaluation method without reference to evaluating the image effect before and after optimization. The analytical results show that, on the basis of ensuring the operation time, the optimized algorithm makes the output defog image more clear and meets the requirements of UAV aerial image data quality control.
The successful extraction of tobacco single plant automation is of great significance to the realization of tobacco agricultural information, but there are still great difficulties in tobacco fine extraction. Therefore, a tobacco extraction method based on Fuzzy superpixels (FS) algorithm is proposed. Firstly, vegetation coverage area in UAV image is obtained by green space extraction method; secondly, super-pixel segmentation of image is carried out by using FS algorithm, and the mean value, brightness, shape index, aspect ratio, custom vegetation index and other features of super-pixel are counted; finally, the number of tobacco plants is extracted and counted by calculating the feature threshold of super-pixel. Three UAV images were selected as the experimental data. The experimental results show that the overall accuracy of this method is 84.28%, 89.05% and 82.97% respectively. This method can be used for automatic extraction of tobacco plant and can provide effective reference for later calculation of tobacco yield.
This paper proposes an algorithm for automatic registration of remote sensing images based on grid index, aiming at tackling the problems of a small number of registration point pairs and a large number of mismatches captured by the SIFT algorithm in the process of remote sensing image registration. First, SIFT algorithm is used to extract feature points and feature vectors, and matching is made by Euclidean distance; secondly, a grid index is established to eliminate part of the mismatched point pairs, thereby improving the accuracy of the random sampling consensus algorithm; finally, geometric polynomials are used to achieve accurate registration of remote sensing images. The experimental results show that the algorithm has higher accuracy of matching point pairs than the traditional block algorithm in remote sensing images, and takes into account the differences in registration scenes of different remote sensing images.
Lakes in the Tibetan Plateau constitute one of the most important natural factors in the plateau ecological environment. So, it is an urgent task to investigate and monitor lakes in the Tibetan Plateau. Because of the unique backscatter characteristics of water body in the image, the extraction and analysis of the lake based on SAR image has become a research hotspot. In order to further eliminate the interference of surface features and improve the classification accuracy, this paper proposes a high-precision lake extraction FR-MorphACWE (Faster Region-based Convolution Neural Network-MorphACWE) model of SAR image. The Interferometric Wide Swath (IW SLC) of the European Space Agency's sentinel-1A interference wide-band mode is used as the main data source, and the sentinel-2a multispectral image level-1c product is used as the reference data source. This model combines the morphological analysis advantages of Faster R-CNN target detection algorithm and the contour extraction advantages of MorphACWE model. The classification experiments were carried out from extraction of comprehensive interference multi-lake. The target detection algorithm was applied to eliminate non - lake surface disturbance. On such a basis, the active contour model was used to extract the lake boundary, and the morphological characteristics and radar reflection characteristics of plateau lakes were fully utilized to achieve high-precision extraction of plateau lakes from the south of Naqu City to the north of Xigaze City in Tibet. The experimental results show that the accuracy of the algorithm can reach 99.71% and the accuracy and recall rate are higher than 98% in the situation of multi-lake interference.
Carbonate content in soil is an important basis for soil classification and fertility evaluation. Based on an analysis of calcium carbonate content, the authors chose 78 soil samples from Loess Plateau of Shaanxi Province as the research objects. The visible near infrared hyperspectral reflectance (350~2 500 nm) data of soil samples were obtained by hyperspectral imager. Three mathematical transformations, i.e., first-order differentiation, second-order differentiation and continuum removal, were carried out on the original spectral curve, and correlation analysis was used. The method and the continuous projection algorithm were used to select the sensitive band respectively, and the Stochastic Forest regression was used to establish the estimation model of soil calcium carbonate. According to the results obtained, the spectral curve characteristics of Huangmian soil are almost the same, there are obvious absorption characteristics at 1 440 nm, 1 900 nm, 2 200 nm and so on, and the calcium carbonate content and spectral reflectance show a positive correlation trend; the accuracy of random forest estimation model based on the second-order differential and continuous projection algorithm is the highest, the validation set R 2 is 0.82, and the PRD value is 2.37.
In order to understand the regional reliability of the Fengyun-3C(FY-3C) satellite snow products, the authors used the snow cover data of 118 meteorological stations in the Tibetan Plateau from October 1, 2018 to April 30, 2019 to evaluate the snow cover (MULSS_SNC) and snow water equivalent (MWRIX_SWE) products. The results show that, for snow cover pixels of MULSS_SNC and MWRIX_SWE, the accuracy rate is 87.18% and 72.32% respectively, the recall rate is 66.67% and 49.63% respectively, the false rate is 12.81% and 27.68% respectively, and the missing rate is 33.33% and 50.37% respectively. In terms of mixed pixels or pixels with snow depth less than 0.5 cm, both MULSS_SNC and MWRIX_SWE tend to identify with no snow, and the missing rate of snow depth less than 1cm is up to 60%. When the snow depth of MULSS_SNC is more than 2cm, the recall rate can reach 89.09%. However, for MWRIX_SWE, only when the snow depth is more than 5cm can the snow recall rate reach 63.37%. The snow depth in the Tibetan Plateau from MWRIX_SWE has a large error with ground observations, and there is no linear positive correlation, so it is not recommended to use it directly.
The spatial-temporal characteristics of spring drought are very important for decision-making and many agricultural applications. In this study, the spatial-temporal analysis of vegetation drought in Qinghai-Tibet region from 1995 to 2010 was carried out by using the vegetation state index of NOAA. According to the characteristics of VCI as a drought index, a variety of methods were used, which included frequency analysis, trend analysis and man Kendall experiment. The results show that the Qinghai-Tibet Region is less affected by monsoon, the frequency of drought in Hengduan Mountain and Qilian Mountain is relatively low, and most of the droughts are light and medium drought. According to the analysis, the trend of drought in this area is not unidirectional and can be divided into two stages. Before 2000, the VCI index was relatively high, and the volatility was relatively large; after 2000, the VCI index was relatively low, and relatively stable.
Drought is a kind of natural disaster with great influence, heavy disaster and long recovery period. As Guangxi is a large agricultural region, it is of great significance to analyze and forecast the drought situation in Guangxi for disaster prevention and mitigation. In this study, the authors analyzed the rainfall in Guangxi from 1998 to 2019, and introduced the standardized precipitation index (SPI) SPI drought index to verify the applicability of tropical rainfall measurement mission (TRMM) data in Guangxi. In addition, the evolution of drought in Guangxi in the past 22 years was studied, and the trend of drought change in Guangxi was predicted. The results are as follows: ① TRMM 3B43 rainfall data have a high correlation with the measured data of ground stations, which proves that TRMM3b43 rainfall data are suitable for drought monitoring in Guangxi. ② Drought and flood disasters occur frequently in Guangxi, with a large range of flood events every 6 years and serious drought events every 2~3 years. ③ The rainfall in Guangxi is the largest in summer and the smallest in winter, and the overall rainfall pattern is “more in the east and less in the west”. ④ It is estimated that there would be no major drought and flood events in Guangxi in 2020; nevertheless, some cities would have mild floods and mild droughts.
The Wuxia section of the Three Gorges reservoir area is an area where landslides and dangerous rock collapses easily and frequently occur, which seriously endangers the safety of the Yangtze River channel. The oblique aerial photography technology can accurately describe the side texture information of the observed objects and provide basic data for the early identification of hidden dangers of geological hazards. In this paper, based on the elaboration and analysis of oblique aerial photography technology, the authors constructed three-dimensional models based on the acquired oblique aerial images, airborne position and orientation system (POS) data and ground control points, and established a geological hazard risk assessment model based on factors such as geology, geomorphology and hydrology. The development and distribution of new geological disasters in the work area were clarified, the main factors controlling the occurrence of geological disasters in each area were figured out, and the geological disasters susceptibility characteristics and hidden dangers in the area were grasped, thus the application potential of oblique aerial photography technology in geological survey was demonstrated.
The main driving factors of runoff change are climate change and land use. In order to accurately predict the runoff change trend in the upper reaches of Hanjiang River in the future, the authors predicted runoff changes under the two land-use change modes based on SWAT model, weather generator BCC /RCG-WG and Ca-Markov model. The setting of land-use scenarios was based on the prediction results of Ca-Markov model. The results show that, in the future, the runoff in the upper reaches of Hanjiang River will show an obvious upward trend, and the increase rate of runoff under the scenario of increasing forest land is less than that in the natural ecological scenario, which may be related to the increase of ecological water demand caused by the increase of forest land. Therefore, the protection of water resources in the upper reaches of Hanjiang River is not feasible. Researchers should follow the natural growth law of vegetation, continue the current ecological protection policy, try to reduce man-made pollution and improve the ideological awareness of water conservation.
Satellite remote sensing technology has been used in China for more than ten years to obtain annual national mine geological environment remote sensing monitoring data. How to make full use of existing remote sensing monitoring results to carry out large-scale mining geological environmental assessment zoning is worthy of discussion and research. Therefore, the Zhungeer Coalfield was taken as the research area, and the artificial interpolation and hierarchical weighted improved gray correlation method was used to evaluate and analyze the geological environment of the 16 open-pit coal mines in the area, and the evaluation level divisions based on the mines were obtained. The evaluation conclusion is that the 3 mines of Huilong, Mengxiang and Liangjiaqi are severely affected by the environment, and the 5 mines of Tianciyuan, Weijiamao, Tingziyan, Liuhuliang and Heidaigou have relatively high environmental impacts. In the severe areas, 8 mines of Harwusu, Jinzhengtai, Zhaofu, Yongsheng, Cui’ergezui, Hongran, Huafu and Zhengren are general environmental impact areas, and other non-mine areas are the non-influenced areas. The evaluation results can reflect the geological environment status of the Zhungeer coalfield relatively objectively. Therefore, the gray correlation evaluation method of mine geological environment could be popularized and applied in a large area or even in whole China.
In order to scientifically evaluate the ecological quality of the land and effectively identify the main controlling factors of the land ecology, the authors established a remote sensing evaluation model based on ideal points in Guang’an which served as a research area, evaluated the ecological quality of the land in Guang’an in 2000, 2005, 2010 and 2015, and analyzed the main controlling factors. A kilometer grid was used as the evaluation unit. The evaluation index system was constructed based on the fourteen evaluation criteria in the four criterion layers, i.e., ecological background, ecological structure, ecological benefits and ecological stress. The evaluation index system was constructed by applying Delphi method and entropy weight method. The weight value of each evaluation index and the ideal point values were calculated by using the ideal point model, and the ideal point level was divided. The principal factor analysis method was used to obtain the main control factors of each year, and then the relationship between the spatial distribution of the ideal point level and the environmental impact factor was performed. Through research, the authors obtained the overall upward trend of land ecological quality in Guang’an City. It is shown that the proportion of land ecological quality at various levels of area and spatial distribution and the proportion of forest land area and temperature factor are the most important main control factors, and the proportion of woodland and the temperature are positively related to the land ecological quality. After analysis, the suggestions on land ecological quality supervision in Guang’an City are put forward. which can provide references for land ecological quality supervision in other areas.
Recently, with the continuous development of domestic satellites, the quantity and the quality of remote sensing data have been improved observably, the needs of “Belt and Road initiative” in remote sensing geological investigation can be satisfied. In this paper, the authors carried out interpretation of geology and minerals in the southern part of Iran which is an area abundant of chromite based on GF-1 satellite data. The authors studied regional geological background and metallogenic condition and proposed the favorable areas for mineralization. Through the study of interpretation, it is found that the layered ultrabasic rocks in the northwest of Sorkband complex have a good prospect of chromite exploration, and it also verifies the good application effect of domestic GF-1 data in the overseas geological and mineral survey. This study is aimed at supporting the establishing of the system for domestic geological survey and making suggestions for mining enterprise.
The cultural landscape of the Zhangjiakou Ming Great Wall is susceptible to surface deformation, making the systematic conservation of cultural landscape in this corridor quite challenging. In order to fix the methodology and application gaps of Great Wall monitoring (large-scale linear heritage) systematically, the authors applied the SBAS-InSAR technology to the time-series deformation surveillance in this pilot case study. In the procedures of InSAR data processing, an external weather model (GACOS) was firstly used to reduce the atmospheric artifacts on interferograms; moreover, a 40 m Gauss and the Goldstein filters were sequentially applied for the phase noise suppression relevant to the natural landscape. In total 67 Sentinel-1 SAR images including 33 ascending and 34 descending data acquired from May 2017 to July 2018 were collected for the line of sight (LOS) deformation calculation using the SBAS-InSAR approach. The derived deformation rates were then projected onto vertical direction for the further analysis. Afterwards, motion rate profiles of ascending and descending datasets from a typical mountain and a flat area were selected for cross-validation, resulting in the maximum and averaged root mean square errors of 9.3 mm/a and 4.0 mm/a, respectively. With considering the significance level, the result demonstrates that 79.5% of the Great Wall corridor (85.1 km totally observed) is relatively stable (with deformation rates in the range of -10 mm/a to 10 mm/a) while remaining 20.5 % shows significant motions (the maximum subsidence rate up to -64.5 mm/a) using the 10 mm/a as the threshold. This pilot study implied the applicability of the applied SBAS-InSAR approach to the synoptic deformation monitoring of large-scale linear heritage sites.
In the context of global warming, the study of the long-term spatial change characteristics of the boreal forest cover not only is important for global climate change and sustainable development research but also can provide the support for the further research on the response of the boreal forest changes to climate change. The data sources were Landsat TM/OLI images with 2 temporal series in summer season from 1985 and 2015, respectively. The Krasnoyarsk region in Russia was selected as the typical research area of the boreal forest in Siberia. The forest cover in 1985 and 2015 was classified based on the decision tree method and verification with random sample points of GF-2 satellite images, and the classification accuracy was 94.53%. The information of the dynamic spatial distribution of forest cover was quantified through latitude zones with 2° interval in the range of N51°~69° and the spatial overlay analysis for the dynamic forest cover maps of the two periods. The results show that, in the past 30 years, the boreal forest cover in Siberia changed significantly, and the overall forest cover changed from 75.42% in 1985 to 80.53% in 2015, increasing by 5.11 percentage points. Simultaneously, the changes of forest land area were different with each latitude zones: the highest change rate occurred in the latitude zone N65°~67°, followed by the latitude zone N67°~69° and the lowest was in N57°~59°. Overall, the forest cover increased in all latitude zones, the most significant increase was in N63°~67°; the change of forest cover was relatively stable in N57°~63° and the increase of forest cover decreased with the latitude zone in N51°~57°.
Dajiu lake wetland is a rare subtropical alpine wetland in Central China. The wetland has experienced several periods of large-scale development since the founding of the People’s Republic of China, which has led to the serious destruction of the wetland. The “Dajiu Lake wetland protection and restoration and park construction project” implemented in 2005 has made the wetland function recovered to a certain extent. To understand the land use changes in Dajiuhu wetland, the authors identified nine land use types of Dajiu lake wetland based on field investigation and previous work. The high-resolution remote sensing images acquired in 2005, 2011 and 2017 and UAV images in 2018 were used to visually interpret the land use types. The dynamic change and type conversion of land use in three periods were examined and the driving forces were explored. The results show that, from 2005 to 2011, new lakes (84.41 hm2) were added, and the most decreased area was farmland, which mainly transformed into xerophytic meadow and wet herbaceous swamp. From 2011 to 2017, a new type of mesophytic meadow (80.07 hm2) was added, which was mainly transformed from wet peat swamp, wet herbaceous swamp and xerophytic meadow. Most of the reduction was in farmland, which was mainly converted to xerophytic meadow. In a word, during the research period, the wetland types and areas of Dajiu Lake were increasing, the wetland landscape was restored to a certain extent, and the wetland ecological environment was improved. The analysis of driving forces shows that the establishment of Wetland Nature Reserve and a series of effective wetland ecological restoration projects are the main driving forces of land use change in Dajiu Lake wetland. The results of this study can provide reference and suggestions for wetland restoration and protection.
As an important water resource, soil water has a significant impact on the distribution and growth of vegetation by its temporal and spatial distribution and dynamic changes. The Mongolian Plateau is a typical arid-semi-arid climate zone, and it is also a major component of the temperate grasslands of the Eurasian continent. Changes in soil water content caused by climate change will undoubtedly have a direct impact on the health and stability of the grassland ecosystem. Clarifying the soil moisture of the Mongolian Plateau as well as the temporal and spatial characteristics and their response to climate change helps provide scientific support for the formulation of ecological protection related policies. Based on GLDAS-Noah soil moisture data, the authors used linear regression analysis, correlation analysis, and Mann-Kendall (MK) test methods to analyze the temporal and spatial patterns, changing trends, and mutation characteristics of soil moisture at different depths from 1982 to 2018. Combined with CRU temperature and precipitation data, the authors explored the response of soil moisture to changes in meteorological factors. The results are as follows: ① The annual average soil moisture of the Mongolian Plateau is generally in a spatial distribution pattern of “high in the northeast and low in the southwest”, and there are obvious high-value areas, transitional zones and low-value areas. ② In the past 37 years, the soil moisture of 0~10 cm (SM1) in the Mongolian Plateau has shown an insignificant upward trend, with a rate of change of 0.002 m 3/m3/10 a. The results of MK showed that a sudden change occurred around 2012; soil moisture of 10~40 cm (SM2), the downward trend was more significant, the rate of change was -0.005 m3/m3/10 a, and its sudden change occurred around 1996. ③ The correlation analysis based on the pixel scale shows that the soil moisture in different seasons has a significant positive correlation with precipitation on the whole, and has a significant negative correlation with temperature.
The ecological barrier effect of Qinling Mountain on aerosol optical depth (AOD) and the relationship between aerosol and terrain were studied by spatial analysis and Kernel Density Estimation using AOD retrievals obtained from the Terra-MODIS Collection 6.1 Level-2 aerosol product at 3 km spatial resolution from January 2002 to December 2017. The results showed an obvious effect of the ecological barrier effect of the Qinling Mountain on atmospheric aerosol. An annual average AOD value at the northern foot of the Qinling Mountain was higher than that at the southern foot of the Qinling Mountain in the past 16 years. High-values of AOD were in Guanzhong urban agglomeration at the northern foot of Qinling Mountain. By comparing and analyzing the changes of AOD along different longitudes of Qinling Mountain, obvious differences were observed in AOD between the south and the north of Qinling Mountain in terms of extreme, mean and special values, which showed the characteristics of “north-south differentiation” and further illustrates the obvious barrier effect of Qinling Mountain on AOD. AOD over the Qinling Mountain showed a significant elevation stratification effect, i.e., a logarithmic downward trend with the increase of elevation. Under the elevation of 2 000 m(mutation point), AOD showed a significant accelerated downward trend with a rate of 0.001/1 000 m. Over the elevation of 2 000 m, AOD showed a significant uniform downward trend with a rate close to zero. The results also showed that the distribution of AOD varied greatly in different terrain of the Qinling Mountain. The AOD agglomeration centers in the plain areas were located at 330~420 m above mean sea level (a.s.l.) , where 79% of AOD were concentrated between 0.35 and 0.71, and high-value AOD (AOD =0.7) agglomeration centers were located between 330 and 340 m. The AOD agglomeration centers in the low-mountains areas were located at 900~1 000 m a.s.l., and 79% of AOD were concentrated between 0.15 and 0.32. The AOD agglomeration centers in middle-mountains areas were located at 1 000~1 400 m a.s.l., and about 60% of AOD were concentrated between 0.19 and 0.3. AOD presented a foggy distribution under plain, low-mountain and middle-mountain areas. AOD in high-mountain areas presented a sporadic point distribution without concentration centers. On the northern slope of Qinling Mountain, the elevation distribution of AOD concentration center was 500 m, while that of the southern slope was about 1 100~1 200 m. The AOD concentration center value (AOD=0.6) on the northern slope was higher than that on the southern slope (AOD=0.22). The AOD decreased logarithmically with the elevation of the northern slope of Qinling Mountain, but the decreasing trend of the southern slope was not obvious.
Land surface temperature (LST) retrieved from remote sensing plays an important role in climatology, hydrology, ecology and other fields, and microwave detection has the wide range and all-weather advantages. It is of great significance to verify the reliability of LST products from domestic satellite on a large scale. Based on the microwave LST product of Fengyun 3C combined with ground surface temperature observed from 97 meteorological stations in Hunan Province, the authors explored the accuracy of microwave inversion and its influencing factors. The results show that the mean absolute error, the root mean squared error, the coefficient of determination, the relative error between LST product and observed data is 6.54℃, 8.88℃, 0.57 and 0.29% respectively, the accuracy of ascending (nighttime) and the south is better than that of descending (daytime) and the north, and the worst consistency is around Dongting Lake. The LST product is of high precision in low temperature but with general underestimation, the accuracy is linearly correlated with the average temperature of each site, and in most cases it is comparable with MODIS products. The precision of LST product increases with the altitude, and varies with seasons, the time series fluctuation of ground temperature can be accurately captured at the sites with strong consistency. According to the analysis results, the inversion accuracy and applicability of LST product could be improved by modifying the retrieval algorithm in the future.