最新刊期
卷 22 ， 期 5 ， 2018
- 摘要：Sea Surface Temperature (SST) is an important physical parameter in the field of marine and climate research. Passive microwave remote sensing has the advantage of completing all weather observations that disregard cloud interference, which has received increasing attention. FY-3C satellites, which carry a Microwave Radiometer Imager (MWRI) onboard, were successfully launched on December 23, 2013. Therefore, using the FY-3C MWRI to retrieve the SST is crucial. The FY-3C MWRI SST uses statistical algorithms. First, MWRI precipitation and sea ice products were used to remove the precipitation and sea ice data. Second, the MWRI brightness temperature was matched with the buoy SST using a temporal window of 0.2 h and a spatial window of 0.2°. The matchup with land within 100 km was excluded. Third, the descending and ascending statistical relationship, which was divided into four latitudes and 12 months, between the buoy SST observation and MWRI bright temperature was established. In addition, 4 × 12 × 2 regression coefficients were obtained, and corresponding regression coefficients were used to estimate the SST. The daily SST was obtained using a 0.25° × 0.25° equal latitude and longitude projections. The quality flag is set to 51 when the FY-3C MWRI SST minus a 30-year monthly mean SST is greater than 2.5 K, thereby indicating that these pixels were distributed on the edge of the land and high wind-speed region. The quality validation of the FY-3C MWRI SST after excluding the pixels with a quality flag of 51 shows that the precision of the ascending orbit SST is –0.02±1.22 K and that of the descending orbit SST is –0.15±1.28 K in comparison with the global buoy data. The precision of the ascending daily SST is 0.00±1.03 K and that of the descending daily SST is –0.09±1.08 K in comparison with the global analysis field OISST. The ascending orbit is more accurate than the descending orbit considering the non-uniform heating of the ocean surface during the day (the descending orbit). The Kuroshio Current, Gulf Stream, Western Pacific Warm Pool, and La Nina are included in the monthly SST, thereby suggesting that this SST is applicable to climatology investigation. The results of the quality validation of the FY-3C MWRI SST include the FY-3C quality control system. The SST precision is influenced by the performance, calibration, and positioning accuracy of the MWRI, precipitation and sea ice detection accuracy, land interference, and high wind speed. The improvement of the precision of the SST with a wind speed that is higher than 12 m/s is the emphasis of the next step. The buoy SST and global analysis field OISST cannot be considered a completely true value. Therefore, the triple collocation method will be utilized in the future to improve the comprehensive analysis of the error characteristics of the SST.关键词：FY-3C;microwave imager;sea surface temperature;inversion;quality validation
- 558
- |
- 126
- |
- 4
发布时间：2021-06-07 -
摘要：Sea ice concentration, which refers to the percentage of sea ice in an area, is one of the important parameters describing the characteristics of sea ice. Remote sensing monitoring of Arctic sea ice is crucial for understanding the role of the Arctic in the global climate system and in global warming. Therefore, comparing and evaluating the products of sea ice data retrieved from different satellite observations are necessary. In addition, assessing the accuracy of sea ice distribution from satellite observations is significant in studies on climate change and global warming because the extrapolation of global surface energy flows is very sensitive to the estimation of Arctic sea ice cover. To evaluate sea ice concentration from the Microwave Radiometer Imager (MWRI) onboard the FY-3B satellite, we compare the data products with the sea ice concentration from the advanced microwave scanning radiometer for earth observing system (AMSR-E) onboard the Aqua satellite. High-resolution moderate resolution imaging spectroradiometer (MODIS) data are used to validate MWRI and AMSR-E sea ice concentration. The comparison of MWRI and AMSR-E incudes daily and monthly data. MODIS L1B band 2 channel reflectance is chosen to validate the sea ice concentration of MWRI and AMSR-E. The processing procedure of MODIS mainly consists of solar zenith correction, radiometric calibration, removing bow-tie phenomenon, and map projection. The map projection is the polar stereographic projection, which is the same as the MWRI and AMSR-E products. To avoid the misjudgment of ice and water because of the cloud influence, we choose cloudfree MODIS channel 2 sub-region for ice and water recognition. Different thresholds according to the histogram of the reflectance are used to segment the ice and water pixels. The MODIS sea ice concentration in each MWRI and AMSR-E grid is calculated. The comparison results of MWRI and AMSR-E are as follows. First, MWRI and AMSR-E daily mean sea ice concentrations show a consistent change from July to September, and MWRI sea ice concentration is higher than AMSR-E each day. The monthly mean values of daily biases for July, August, and September are 8.55%, 7.67%, and 2.58%, and the monthly mean of daily standard deviations are 12.16%, 12.08%, and 10.43%, respectively. Second, from July to September, the monthly sea ice concentration difference between MWRI and AMSR-E (MWRI minus AMSR-E) shows a decreasing trend. The monthly biases for July, August, and September are 7.37%, 6.53%, and 1.51%, and the monthly standard deviations are 4.61%, 4.36%, and 3.64%, respectively. The validation results of MWRI and AMSR-E sea ice concentration with MODIS are as follows. First, in the region of MODIS sea ice concentration less than 95%, the sea ice concentration difference between MWRI and MODIS (MWRI minus MODIS) is 9.78%±16.90%, and the difference between AMSR-E and MODIS (AMSR-E minus MODIS) is 0.90%±13.09%. MWRI and AMSR-E tend to overestimate sea ice concentration compared with MODIS, the result of AMSR-E is closer to MODIS, and MWRI has a larger error. Second, when MODIS sea ice concentration is greater than or equal to 95%, the sea ice concentration difference between MWRI and MODIS is –5.45%±7.05%, and the difference between AMSR-E and MODIS is –7.97%±6.46%. MWRI and AMSR-E tend to underestimate sea ice concentration compared with MODIS, and the result of MWRI is slightly better.关键词：sea ice concentration;FY-3B/MWRI;Aqua/AMSR-E;comparison;validation;MODIS;ice and water recognition- 654
- |
- 187
- |
- 10
发布时间：2021-06-07
Chinese Satellites
-
摘要：A hyperspectral Light Detection And Ranging (LiDAR) can obtain high spectral properties of the observed object and provides a new method for detecting a three dimesional distribution of vegetation structure and biochemical characteristics. In this study, a data process flow and a biochemical characteristics method were proposed. Laboratory experiments were conducted on the basis of this instrument, and a vertical distribution extraction method of vegetation biochemical components was provided. First, we proposed the hyperspectral LiDAR waveform data processing method in accordance with the characteristics of the instrument. Second, an indoor Kniphofia scanning experiment was utilized, and the LiDAR point cloud data with high spectral properties were obtained. Finally, chlorophyll and carotene vertical distributions were extracted on the basis of the relationship between the vegetation index and biochemical components. Results show that the biochemical content of a red leaf at the top of vegetation is low, which is generally lower than 0.5 mg/g, and carotene is less than 0.2 mg/g. However, the biochemical component content in the middle of the green leaves was evidently high. This study showed that the instrument has a considerable application prospect in the field of quantitative remote sensing.关键词：hyperspectral LiDAR;biochemical parameter;vertical distribution;full waveform;point cloud- 557
- |
- 220
- |
- 5
发布时间：2021-06-07 -
摘要：Two-media satellite stereophotogrammetry, a combination of high-resolution satellite stereo images and two-media stereophotogrammetry, is a potential and promising method for shallow seafloor surveying, especially in areas far from the mainland. This study focuses on this method and its possible application in shallow seafloor surveying. The study has two objectives. The first is to solve the problem of poor applicability of an existing accurate positioning algorithm (i.e., refraction correction algorithm) in two-media stereophotogrammetry. The second is to explore the feasibility of shallow seafloor surveying using WorldView-2 stereo images and two-media stereophotogrammetry. First, a universal algorithm is presented for refraction correction based on the general geometry of two-media stereophotogrammetry. Two cases are considered in the algorithm, namely, the intersection and non-intersection of the corresponding two aerial straight rays of an underwater object. The positional relationship between the transitional point and other important relevant points in two-media stereophotogrammetry is also considered. Three experiments are performed to verify the feasibility of shallow seafloor surveying using two-media satellite stereophotogrammetry and the performance of the proposed refraction correction algorithm. The first experiment uses two-media stereophotogrammetry to assess whether a Digital Elevation Model (DEM) of shallow seafloor can be derived and determine how it is derived from WorldView-2 images. The second experiment assesses the accuracy of the DEM. (Note that quantitative accuracy assessment is not carried out for the Shanhu Island area because of the lack of reference data). The third experiment demonstrates the performance of the proposed refraction correction algorithm by comparing it with the results of Murase’s algorithm, which is a typical method. Two groups of WorldView-2 stereo images are used in these experiments, which show the Ganquan and Shanhu Island areas in Shansha City, Hainan Province. Results show that the DEMs of the two areas are successfully derived from the stereo images using the proposed method. For the Ganquan Island area, the DEM obtained by our refraction correction algorithm has an accuracy of 2.08 m, while the DEM obtained by Murase’s algorithm has an accuracy of 2.31 m. Previous practice suggests that the DEM accuracy of 2.08 m is satisfactory for satellite photogrammetry. These experiments show that an eligible DEM of shallow seafloor can be derived from WorldView-2 images by using two-media stereophotogrammetry, and its accuracy can be slightly improved by using our proposed refraction correction algorithm. In conclusion, the proposed refraction correction algorithm is generally applicable because it is suitable for both two-media stereophotogrammetric cases, specifically, the intersection and non-intersection of corresponding two aerial straight rays of an underwater object point. Deriving a DEM of shallow seafloor from WorldView-2 images using the method of two-media stereophotogrammetry is feasible.关键词：WorldView-2 image;two-media stereophotogrammetry;accurate positioning;refraction correction;shallow seafloor;surveying accuracy- 616
- |
- 87
- |
- 2
发布时间：2021-06-07 -
摘要：Remote sensing image scene classification aims to tag remote sensing images with semantic categories according to the content of the image and is important in disaster monitoring, environmental detection, and urban planning. Scene classification results can provide valuable information about object recognition and image retrieval and can effectively improve the performance of image interpretation. The general process of remote sensing image scene classification mainly consists of feature extraction and scene classification based on image features. Given that the design of classifiers is relatively mature, this work focuses on feature extraction strategy. The influence of various strategies on the performance of scene classification is short of unified evaluation, which limits its development. The effect of various feature extraction strategies on the performance of high-resolution remote sensing image scene classification is evaluated in this study. In the second section of this paper, existing feature extraction strategies are divided into two categories: (1) hand-designed and (2) data-driven feature extraction. Hand-designed features, such as Color Histograms (CH) and Scale Invariant Feature Transform (SIFT), provide the primary description of images and are presented in the early period. Further abstract description of the images is introduced by coding of hand-designed features, such as Bag of Visual Words (BoVW) and has higher classification accuracy than hand-designed features. However, these feature extraction strategies generally suffer from poor generalization capability due to specific requirements for designing. Furthermore, hand-designed features require significant domain knowledge. By contrast, data-driven features can directly learn powerful features from a large number of sample images and are generally divided into shallow and deep learning features. Shallow learning feature extraction mainly involves Principal Component Analysis (PCA), Independent Component Analysis (ICA), and sparse coding algorithms. Typical deep learning feature extraction strategies include stacked autoencoder (SAE), Deep Belief Network (DBN), and Convolutional Neural Network (CNN). Compared with deep learning models, shallow learning models can be regarded as a neural network with a single hidden layer and thus cannot capture high-level semantic features. The superiority of deep learning features is obvious when dealing with complex scene classification. Furthermore, CNN-based features exhibit improved performance compared with SAE- and DBN-based features because the one-dimensional structure of SAE and DBN destroys the spatial information of images. In the third section of this paper, 29 feature descriptors are quantitatively compared in UC Merced, AID, and NWPU RESISC-45 datasets and eight combinations of feature descriptors are quantitatively compared in the NWPU RESISC-45 dataset. The effect of different feature extraction strategies on the performance of scene classification and the complexity of each dataset are evaluated through quantitative comparison. The experimental results are as follows. (1) The classification accuracy and stability of hand-designed features is poor, however the efficiency of most features is satisfactory and can attain better performance by combining with other types of features. (2) Among all feature extraction strategies, the coding of hand-designed features possesses moderate levels of classification accuracy, efficiency, and stability. (3) The classification accuracy and stability of data-driven features are best, but most of them have low efficiency. (4) AlexNet, a deep learning model with few layers, exhibits the best comprehensive performance and is suitable for occasions that require high classification accuracy, efficiency, and stability. (5) Some scene classes belonging to land use type are easy to be confused because of similar landmark buildings or sites. Moreover, some scene classes belonging to land cover type are easy to be confused because of their similar geomorphologic features. (6) The recently proposed NWPU RESISC-45 dataset is more complex than the other datasets and is more challenging for scene classification algorithms. Finally, the summary and conclusion of this paper are presented, and the discussion of future development is provided. On the one hand, combining prior knowledge introduced by hand-designed features with the CNN model may be one of the future development directions. On the other hand, introducing Generative Adversarial Networks (GAN) into CNN training may be a research hotspot in the future. In addition, remote sensing parameters, such as NDVI and NDWI, and multi-spectral information can be integrated with current feature extraction strategies for practical applications.关键词：high-resolution;scene classification;feature extraction strategy;hand-designed features;data driven features;deep learning- 655
- |
- 340
- |
- 19
发布时间：2021-06-07 -
摘要：Based on the characteristics of road images, a road extraction method for high-resolution optical remote sensing images with multiple feature constraints is proposed in this paper. First, the road image features are analyzed. Second, the feasibility of extracting road images with different feature constraints is determined to emphasize the necessity of building a road feature detection model. This study provides specific methods and steps. First, an improved line segment secondary extraction model is presented based on road geometric characteristics. The model is composed of the aggregation constraint and line segment dynamic fitting. Meanwhile, a candidate road seed set is constructed based on the geometric relationship between the extracted maximum length of 100 lines and known road width. Second, road structure information is inputted. The road radiation characteristics of the candidate road seed set are determined through an entire evaluation, which is based on improved template matching in road value formula, and the overall matching step concrete is identified. Third, when the candidate road seed set cannot meet the requirements of radiation characteristics, road context features are used to evaluate the seeds. Therefore, a light vehicle detection model that includes vehicle image geometric parameter analysis, morphological processing, edge detection, vehicle closed radiation information analysis, and other steps is established. The detection result of light vehicles will be used as the one of the road’s context features. Fourth, the candidate seed points are verified according to the road topological feature. In this process, the road topological analysis model is constructed. The model uses the matching–tracking model and the contextual features to verify the road inspection points. The study proposes an improvement in the matching–tracking model and matching measure. Fifth, in road post-processing, the seed points of the extraction road are optimized, the false road seed points and the low precision road seed points are eliminated, and the road line fitting optimization method is proposed. This work starts by experimenting the secondary extraction method of the high-resolution remote sensing line segment. Results show that the secondary extraction line segment is relatively complete and that defining the geometric features of the road is easy. Then, an experiment of light vehicle extraction is performed by using the remotely sensed image to verify the feasibility of the proposed method. Furthermore, the results can be used to reflect motor vehicle road context characteristics. Different experiments are performed in different scenes and resolutions of high-resolution remote sensing images. In comparison with the mature software ERDAS and ECognition, the experimental results show that this method is relatively automated, efficient, and considerably suitable in performing road extraction in complex scenes with diversified road types and geometrical spectral noise.关键词：information extraction;road seed point;multiple feature;constraint;road extraction;high resolution;optical- 629
- |
- 103
- |
- 6
发布时间：2021-06-07 -
摘要：After a disaster, the same object with different spectra (such as suspended material, wave, and water depth, each of them having more than one spectrum) leads to incomplete extraction of water body information. Different objects with the same spectrum, such as shadow, asphalt pavement, and dense vegetation, decrease the extraction precision of water body information. To address this problem, the extraction method of water body information based on its characteristics after a disaster is presented in this study. First, the characteristics (i.e., spectra, geometry, texture, and spatial relationship) of water body information are analyzed on remote sensing images to construct a knowledge decision tree. Second, object-oriented segmentation is performed on remote sensing images to obtain object elements and calculate the characteristic parameters. Third, water body information after a disaster is extracted with the support of the knowledge decision tree. Finally, the water body information is post-processed. A validation experiment was performed using the high-spatial resolution remote sensing images of “5.12 Wenchuan Earthquake.” The location of the extracted water is accurate and the boundary is clear. The accuracies of the producer and the user are 0.85 and 0.94, respectively. Results showed that the method can effectively extract water body information after a disaster even when the background is complex.关键词：characteristic knowledge;decision tree;Wenchuan earthquake;water body information extraction after disaster;accuracy evaluation- 828
- |
- 134
- |
- 8
发布时间：2021-06-07 -
摘要：Snow depth is a general input variable in many models of agriculture, hydrology, climate, and ecology. This study adopts the Sentinel-1 C-band of the European Space Agency using the two-pass method of differential interferometry to conduct an inversion study of the snow depth distribution in typical areas of Bayanbulak Basin in the Middle Tianshan Mountains of Xinjiang, China. Based on Sentinel-1 SAR image, the image of day October 31, 2016 is selected as the master image and the image of day December 18, 2016 is used as the slave image to form the image pair. After the interferogram is formed, the orbit phases, terrain, ground effect, and noise effect are removed. The phase unwrapping of the remaining phase aims to obtain the distribution of snow depth with the spatial resolution of 13.89 m on day December 18, 2016 by relying on the relationship between snow depth and snow phase in the typical Bayanbulak region. The study demonstrates the following: (1) After proper preprocessing of Sentinel-1 data, snow depth distribution inversion is achieved by utilizing the InSAR-based two-pass method. However, owing to the difference of image-pair coherence and snow accumulation conditions, a relatively accurate inversion result of snow depth is available when the snow depth is larger than 10 cm (R=0.65, RMSE=4.52 cm, and average relative error is 22.42%). The estimated snow depth is slightly lower than the actual depth. When the snow depth is less than 10 cm, the inversion result is not accurate: it is larger than the actual depth, and the average relative error is higher than 34.52%. (2) The inversion accuracy of snow depth is also significantly influenced by the height and actual snow depth. Moreover, the inversion result of snow depth is influenced by coherence losses. This study demonstrates that the InSAR method is more promising in obtaining and estimating snow depth compared with optical technology and passive microwave remote sensing.关键词：snow depth;Sentinel-1;D-InSAR;error analysis;coherence- 632
- |
- 239
- |
- 4
发布时间：2021-06-07
Technology and Methodology
-
摘要：With the rapid development of the economy, the demand for mineral resources is growing, and the contradiction between supply and demand is increasing. The shortage of mineral resources has become one of the important factors that restrict national economic development. Therefore, research on how to efficiently and accurately explore mineral resources is a critical endeavor. Remote sensing mineralization alteration information extraction is an important application of remote sensing technology in geological exploration, which is of utmost significance to mineral exploration and evaluation. Owing to the influence of vegetation, cloud, and snow, alteration information from remote sensing mineralization is often superimposed with the complex geological background and exists only in the form of a weak signal in the background of the remote sensing image. Research on effective remote sensing mineralization alteration information extraction methods can provide the basis for the study of regional metallogenic prognosis and speed up the evaluation of mineral resources exploration, which helps promote the healthy and stable development of the local mining economy. To improve the accuracy of remote sensing mineralization alteration information extraction method, a remote sensing mineralization alteration information extraction method based on Principal Component Analysis (PCA) and Support Vector Machine (SVM) optimized by cuckoo algorithm is proposed in this study. First, the mineralization alteration information in the remote sensing image of the study area is enhanced by band ratio method, and the ratio images are obtained. Then PCA is applied to the ratio images of the study area. The hydroxyl principal components and iron staining principal components are selected, after which the training samples are extracted. Subsequently, the training samples are trained by SVM, while cuckoo algorithm is used to find the optimal kernel parameter and penalty factor of SVM. Thus, the optimal SVM model is determined. Finally, the optimal SVM model is used to accomplish the extraction of remote sensing mineralization alteration information in the study area. Wulonggou area of Qinghai Province, which is rich in mineral resources, is selected as the study area where the hydroxyl alteration information and iron alteration information are extracted. A detailed comparison among the proposed method and four methods proposed recently, namely, the PCA method, the method based on spectral angle mapper and SVM, the method based on particle swarm optimization and SVM, and the method based on band ratio, PCA, and SVM optimized by particle swarm optimization in terms of extraction effect and matching rate, is given in this paper. Experimental results show that by using the proposed method, the extracted information can comprehensively reflect the remote sensing mineralization alteration information of the study area. Moreover, the matching degree of hydroxyl alteration information and iron alteration information are 86.5% and 69.2%, respectively. Meanwhile, compared with the four methods, the proposed method can obtain the highest matching degree with the best extraction effect. The proposed remote sensing mineralization alteration information extraction method based on PCA and SVM optimized by cuckoo algorithm is an effective method that provides a new idea for mineral exploration and metallogenic prediction.关键词：remote sensing;mineralization alteration information extraction;principal component analysis;support vector machine;cuckoo algorithm;band ratio method- 552
- |
- 78
- |
- 3
发布时间：2021-06-07 - 摘要：Beijing–Tianjin–Hebei and its surrounding areas are some of the most PM2.5-polluted regions in China. Satellite remote sensing technology is an advanced and important means for monitoring the change in spatio-temporal distribution of large-range PM2.5. In this study, we conduct AOT retrieval and validation analysis based on the Dark Target (DT) Method by utilizing FY-3B/MERSI and AQUA/MODIS satellite data in this region. The weather and ground observation data are brought in to retrieve the regional PM2.5 concentration using the GWR model, and cross-verification assessment for remote sensing and retrieval results is conducted. Through comprehensive comparison, this study analyzes the capability of MERSI and MODIS in monitoring aerosol and PM2.5. Finally, a preliminary exploration analysis on the monthly temporal and spatial changing status of PM2.5 concentration is conducted in the first quarter of 2017 in Beijing–Tianjin–Hebei and its surrounding areas by utilizing MERSI data. Results show that the remote sensing monitoring capability of FY-3B/MERSI is slightly better than that of AQUA/MODIS. TheR2 between AOT and PM2.5 dataset with a resolution of 1 km retrieved from MERSI and that from the ground station observation results are 0.76 and 0.79, respectively. The root-mean-square errors are 0.26 and 28 μg/m3, respectively, while the mean absolute errors are 0.16 and 15 μg/m3, respectively. The results can basically meet the demand of fine PM2.5 monitoring in Beijing–Tianjin–Hebei and its surrounding areas. The remote sensing monitoring results of monthly PM2.5 concentration in the first quarter of 2017 in Beijing–Tianjin–Hebei and its surrounding areas show that the spatial pattern of PM2.5 is closely related to the terrain and landscape, with the high concentration zone mainly lying along the Taihang mountains in flakes. From the view of temporal change, a decreasing trend is noted, with March seeing a plunge in concentration value compared with the first two months. The findings suggest that FY-3B/MERSI remote sensing retrieval results can provide effective reference for environmental quality monitoring and environmental management effectiveness evaluation. This study is significant for the development of domestic satellite application in atmospheric environmental remote sensing sector.关键词：Beijing-Tianjin-Hebei and its surrounding areas;FY-3B/MERSI;AOT;PM2.5;remote sensing
- 560
- |
- 114
- |
- 10
发布时间：2021-06-07 -
摘要：Forest species diversity, as a key component of biodiversity, plays an irreplaceable role in maintaining ecological balance, processes, and services. In recent years, forest tree species diversity is facing a serious threat with intensifying human activities and influence of climate change. The status and trends of forest tree species diversity must be dynamically monitored to develop effective forest biodiversity conservation approaches. In this study, an airborne light detection and ranging (LiDAR) (>4 points/m2) and hyperspectral (PHI-3 sensor with spatial resolution of 1 m) data combined with 37 field sample data are used to detect tree species variation in the structural and spectral properties in the Shennongjia Forest Nature Reserve of China. First, we use the morphological crown control method based on a watershed algorithm to isolate individual tree crowns by using LiDAR. We select optimal structural indices from nine commonly used structural indices derived using LiDAR based on the theory of structural and spectral variation hypothesis. Meanwhile, we select optimal vegetation indices from 16 VIs based on the hyperspectral data by conducting a correlation analysis with the field samples. Second, a self-adaptive fuzzy C-means clustering algorithm is applied to map the species diversity (i.e., richness and Shannon-Wiener index) in the study area for each 30 m×30 m window at the individual tree crown scale. Result indicates that the individual tree isolation by using the watershed algorithm can obtain a high accuracy (R2=0.88, RMSE=13.17, P<0.001). The 95th quintile height, canopy cover, and vegetation permeability are the optimal structural indices, and their correlation with the Shannon–Wiener index are from 0.39 to 0.42 (R2=0.39—0.42, P<0.01). The correlation among RI, OSAVI, narrow band NDVI, SR, Vogelmann index 1, PRI, and field inventory Shannon–Wiener index are relatively high based on the airborne hyperspectral data (R2=0.37—0.45, P<0.01). Finally, we use the selected three structural and six vegetation indices to predict the optimal clustering numbers (species richness) and the Shannon–Wiener index by using the self-adaptive fuzzy C-means clustering algorithm. The result shows that the maximum tree species that can be predicted is 20. The prediction accuracy of species richness isR2=0.69, RMSE=3.11, and the Shannon–Wiener index is R2=0.70, RMSE=0.32. This method shows the potential of LiDAR combined with hyperspectral data in mapping species diversity of a subtropical forest. Moreover, it could provide an effective method for analyzing the current situation and its changing trend of forest biodiversity at a regional scale.关键词：forest diversity;species richness;lidar;hyperspectral;individual tree crown isolation;clustering- 649
- |
- 294
- |
- 7
发布时间：2021-06-07 -
摘要：The aboveground biomass of grasslands is an important measurement index for grassland ecosystems and an important basis for the optimal use of grassland resources. In addition, identification of aboveground biomass can be used in monitoring the balance between grassland forage supply and livestock demand. To estimate the aboveground biomass of natural grasslands and determine the variation trend rapidly, accurately, and effectively, we selected the natural rangeland in the northern hillside of Tianshan Mountain as a typical study area and analyzed the spatio-temporal change characteristic of its aboveground biomass. With their rapid development, Unmanned Aerial Vehicles (UAVs) have been extensively used in remote sensing because of their convenient operation, lower cost, and shorter revisit cycle compared with satellites. In addition, the lightweight sensors of UAV allow low-altitude remote sensing, which could capture high-spatial, high-spectral resolution images. We conducted a survey of the different grassland types and vegetation varieties in shady and sunny slopes of the rangeland. We used a multi-rotor UAV equipped with Micro-MCA12 Snap to obtain high-resolution multispectral images and collected field survey data. We established a relational model based on the correlation between the aboveground biomass and Vegetation Indexes (VIs) by regression analysis. Results showed poor correlations between the aboveground biomass and VIs, but these correlations improved remarkably after considering the terrain factors. The effectiveness of the VIs varied in different grassland types and vegetation fractions. Accuracy analysis showed large differences in the fitting accuracy of the different slope aspects and small differences in the effectiveness of the same slope aspect. In sum, the highest effectiveness between the Ratio Vegetation Index (RVI) and the aboveground biomass was obtained in the southern and northern slopes, with an estimation precision of more than 75%. The main conclusions are the following. (1) Different grassland types and vegetation fractions led to the poor correlations between the aboveground biomass in the entire area and VIs. (2) The RVI value in sunny slope was higher than that in shady slope, whereas the aboveground biomass in sunny slope was lower than that in shady slope. Grassland degradation resulted from sustained drought and high temperature. (3) This study proved indirectly the relative insensitivity of heavily vegetated areas. Therefore, the findings of this study coincided well with the actual situation. This research provided a reference for the monitoring of grassland ecosystems and reasonable utilization of grassland resources.关键词：natural grassland;biomass;multispectral images;shady-slope and sunny-slope;estimation models;Unmanned Aerial Vehicles (UAV)- 633
- |
- 180
- |
- 16
发布时间：2021-06-07 - 摘要：Land Surface Temperature (LST) is an important parameter in land surface physical processes in regional and global scales. LST plays an crucial role in the interaction and energy exchanges between the atmosphere and land. LST has been widely used in weather forecasting, ocean circulation, drought monitoring, and energy balance. At present, the spatial resolution of LST products is relatively low, which can no longer meet the demand for monitoring urban heat island and estimating regional evapotranspiration. Landsat satellites offer numerous high-spatial-resolution data of land surface. However, an operational Landsat8 LST product is unavailable, thereby limiting the use of the data. In this study, we developed a physical single-channel algorithm to retrieve LST from Landsat 8 TIRS database, which can be used for retrieving Landsat LST with long time series. A physical single channel algorithm was developed to retrieve the LST from Landsat 8 TIRS data. First, ASTER Global Emissivity Database and vegetation cover method were used to calculate the land surface emissivity. Then, MERRA reanalysis data and fast radiative transfer model RTTOV 11.3 were utilized for atmospheric correction of Landsat8 thermal infrared images. The validation results were divided into two parts: (1) validation using the simulated data and (2) validation based on the ground measured data. First, simulation data calculated by TIGR atmospheric profile and MODTRAN were used to validate the accuracy of the algorithm, and the in situ LSTs between 2013 and 2015 acquired from the HiWATER experiment were used to evaluate Landsat LST. The transmittance and atmospheric upward radiance simulated by RTTOV are close to those of MODTRAN. The mean bias of transmittance between RTTOV and MODTRAN is approximately 0.01, and that for atmospheric upward radiance is approximately 0.04 W/(m2·sr·μm). Compared with the NDVI threshold method, the proposed method can both reflect the dynamic change trend of land surface emissivity over vegetations and reflect the variation among different soil types. The validation results show that the overall deviation for both PSC algorithm and JMS method is within ±0.2 K. Moreover, the RMSE of the PSC algorithm is around 2.2 K, whereas that for JMS algorithm is 2.4 K. LST with high spatial resolution and precision could be obtained using the proposed method for monitoring urban heat island and estimating regional evapotranspiration.关键词：Landsat 8;ASTER GED;RTTOV;physical single-channel algorithm;LSE;LST
- 752
- |
- 346
- |
- 12
发布时间：2021-06-07 -
摘要：China is a country with the largest plantation area in the world. Timely, quantitatively, and accurately acquiring planted forest stand characteristics is the key to monitoring and managing forest plantation, which contributes to the global carbon cycle. This study was conducted at a north subtropical coastal plain forest plantation to estimate forest stand characteristics using light detection and ranging (LiDAR) point clouds combined with 55 field-measured data. First, Canopy Height Distribution (CHD) and Foliage Profile (FP) were extracted and used to obtain Weibull parameters by fitting a Weibull density function to the profiles (Suite 1). Second, Height-Related Metrics (HRM) and Density-Related Metrics (DRM) (Suite 2) were directly extracted from the LiDAR point clouds. Finally, multiple regression models were constructed to predict the stand characteristics (i.e., stand density, diameter of breast height (DBH), basal area, Lorey’s tree height, volume, and aboveground biomass) on the basis of the field-measured data and two suites of metrics. Results showed that (1) the accuracies of the forest stand characteristics that use point cloud (Suite 2) and canopy vertical profile metrics simultaneously (Suite 1) were relatively higher than that only using point cloud metrics (Suite 2) ( ΔAdjusted R2=0—0.13, ΔrRMSE=0.08%—3.65%). (2) For all the forest stand characteristic estimations, the models of Lorey’s tree height (Adjusted R2=0.85, rRMSE=7.66%) and volume (Adjusted R2=0.84, rRMSE=14.27%) demonstrated the highest accuracies, followed by the aboveground biomass (Adjusted R2=0.78, rRMSE=14.15%), basal area (Adjusted R2=0.73, rRMSE=14.70%), and mean DBH (Adjusted R2=0.64, rRMSE=15.05%). Stand density exhibited the lowest accuracies (Adjusted R2=0.58, rRMSE=26.16%). (3) The Weibull function was appropriate for describing the vertical stand canopy structure of planted forests. The Weibull metrics can be favorable in effectively improving the estimation accuracy of the forest stand characteristics. This study provided an effective approach in estimating the forest stand characteristics that use the LiDAR point clouds by combining vertical profile and point cloud metrics. This study also proved that the vertical profile metrics can help in increasing the accuracy of the predictive models and provide the mechanism for interpreting the forest stand vertical structure.关键词：lidar;canopy vertical structure profiles;Weibull;foliage profile;point clouds metrics- 664
- |
- 134
- |
- 17
发布时间：2021-06-07 - 摘要：Snow Depth (SD) measurements are important to hydrology, climatology, and agriculture. Global navigation satellite system multipath reflectometry (GNSS-MR) technology is a relatively new and powerful method for sensing SD. Thus far, many algorithms have been proposed to detect SD from the Signal-to-Noise Ratio (SNR) data of different satellites’ reflectometry signals. However, the relevant studies based on GLONASS satellites have less consideration for the development of existing GNSS-MR algorithms. This study aims to verify and analyze the suitability and reliability of applying GNSS-MR model for SD detection from GLONASS data. Theoretical analysis and formula derivation are conducted to systematically and quantitatively determine the SD detection. By analyzing the characteristics of the ground-based GNSS SNR data caused by multipath, this study introduces the detection model and the basic theory of GNSS-R technology based on SNR data to detect SD data. The SNR data of GLONASS L1 reflectometry signals can be obtained from the CORS (Continuously Oporating Reference Stations) network provided by the International GNSS Service (IGS). The residual sequence of this SNR serves as key knowledge for the SD detection in GNSS-R model. The SD is estimated through the spectral analysis using a Lomb-Scargle algorithm. To verify the above theory, the GLONASS-MR model is used to detect SD over the Yellowknife Henderson area in Canada. The observed period is from November 2015 to June 2016, 243 days in total. Then, the GLONASS products are compared with in situ measurements of the National Climate Data Center. Tesults show the following: (1) The accuracy of the detected SD based on GLONASS-MR can reach centimeter level, and the RMSE is 3.3 cm. The detected results likewise reveal good spatial coincidence, with a correlation coefficient of 0.969. (2) The different snow depths can directly affect the amplitude frequency of GLONASS SNR data and the vertical reflection distance. (3) For the same satellite, a linear relation is discovered between the peak values of the amplitude frequency and the detected SD. (4) Under the same conditions, SD detected using more GLONASS satellites can have less bias than that from a GLONASS satellite. This study aims to add new SNR data to the GNSS-MR algorithm to improve SD detection when other satellites’ data cannot be obtained. Validation using both simulated SD data and in situ measurements indicates that the proposed method is effective and can be used for real-time and continuous SD monitoring. Therefore, this study provides references for the future development of GNSS-R technology in China.关键词：GNSS-MR;GLONASS;SNR;snow depth detection;Lomb-Scargle
- 573
- |
- 188
- |
- 5
发布时间：2021-06-07
Remote Sensing Applications
- Postal code：100190
- Tel：010-58887052 Email：nrsb@aircas.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备20021838号-9
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0