最新刊期
卷 25 ， 期 6 ， 2021
-
摘要：The spatiotemporal distribution of water vapor has impact on precipitation state prediction, climatic disasters, and global ecological equilibrium. And the satellite-based occultation observations greatly improve the understanding of atmospheric vertical composition. This study proposed a modified water vapor retrieval algorithm.There has been information redundancy when retrieving with hyper-spectral remote sensing data of infrared occultation. The interference components will also greatly decline retrieval accuracy. Considering these problems, the Reference Forward Model (RFM) and the channels’ Jacobian were applied when choosing appropriate retrieval channels. 159 retrieval channels which include 13 micro-windows were extracted by using channel selection and sensitivity analysis. This will lessen calculative burden and improve retrieval efficiency. The hierarchical smoothing parameter, which makes the algorithm suitable for the layered detection characteristic of infrared occultation payload, was employed to modify the Rodgers’s Levenberg—Marquardt (LM) optimal estimation algorithm. Thus, smoothing coefficients at different altitudes were used when retrieving water vapor profiles based on infrared occultation data. To validate the reliability of the modified algorithm, the original and modified algorithms were used for the water vapor profiles retrieval based on the ACE-FTS observation data. The retrieval experiment based on GF5-AIUS simulation data was performed for the further validation of the modified algorithm applicability to the domestic occultation satellite payload. By comparing the overall differences and the relative differences at different altitudes of original and modified algorithms under the same number of iteration, the conclusion can be obtained.The result showed that the modified algorithm reduced the overall relative differences of retrieval from ±10% to ±6% for the ACE-FTS observation data. For the GF5-AIUS simulation data, it was from ±9% to ±5%. In addition, the relative differences was also significantly reduced in some atmospheric layers. The decrease of relative differences in different altitudes were more than 25.17% for the ACE-FTS observation data and more than 48.26% for the GF5-AIUS simulation data. The overall decrease of relative difference for the three-scene (Orbit numbers 40993, 43544, 38154) data were 34.7362%, 25.1706%, and 52.1346% for the ACE-FTS observation data. For the GF5-AIUS simulation data, they were 61.3239%, 48.2558%, and 51.9857%, respectively.In conclusion, the modified algorithm works efficiently and reliably in retrieving atmospheric water vapor profiles based on infrared occultation. Therefore, the modified algorithm provides a reference for the water vapor profile retrieval study when using infrared occultation data and is conducive to the further research in atmospheric water vapor retrieval of domestic occultation payload. It provides reliable fundamental data for water vapor for climate change prediction, global ecological balance, and other relevant fields.关键词：water vapor profiles retrieval;OEM;ACE-FTS;GF-5 AIUS;RFM- 714
- |
- 328
- |
- 1
发布时间：2021-06-28 -
摘要：The accurate extraction of quantitative information on urban vegetation coverage is of great significance for urban ecological environment assessment, urban planning, and sustainable urban development. With the development of remote sensing technology, effective means for obtaining regional and global vegetation coverage information have emerged. At present, urban vegetation coverage estimation methods based on single-sensor and single-phase remote sensing data are widely used. However, due to the complexity of urban land cover and the diversity of vegetation types, the accuracy of urban vegetation cover information extraction is compromised. In this study, we propose an urban vegetation coverage estimation method based on multi-source remote sensing data and Temporal Mixture Analysis (TMA). First, the best time series GF-1 NDVI data are obtained by using STARFM and vegetation phenomenological analysis. Second, on the basis of time series GF-1 NDVI and Landsat8 SWIR1 and SWIR2 data, TMA is used to estimate the urban vegetation coverage in Changsha City. Results show that the method based on multi-source remote sensing data and TMA can obtain highly accurate urban vegetation coverage estimates (RMSE=0.2485, SE=0.1377, MAE=0.1889). Compared with traditional methods like single-time phase spectral hybrid analysis and dimidiate pixel model, our method is more stable, and can obtain higher estimation accuracy in low, medium, and high vegetation coverage areas. This study provides an effective method for quantitative estimation of urban vegetation coverage.关键词：multi-source satellite remote sensing data;GF-1;spatiotemporal fusion;temporal mixture analysis;vegetation coverage;urban area- 1228
- |
- 541
- |
- 9
发布时间：2021-06-28
Chinese Satellites
-
摘要：Light Use Efficiency (LUE) characterizes the efficiency with which vegetation converts intercepted or absorbed solar radiation into organic dry matter through photosynthesis. LUE is a key parameter to estimate vegetation productivity, especially in the widely used methods of global or regional Gross Primary Productivity (GPP) and Net Primary Productivity (NPP) estimation with remote sensing data, such as MODIS vegetation productivity algorithm (MOD17), the Carnegie-Ames-Stanford Approach (CASA), the Vegetation Photosynthesis Model (VPM) and the Eddy Covariance-Light Use Efficiency approach (EC-LUE). In these methods, LUE was estimated by multiplying temperature, water and other stress factors with maximum LUE at optimal environment conditions. Due to the combined effects of vegetation type and climate, LUE shows significant spatial heterogeneity and seasonal variation. The uncertainty of LUE estimation is an important fact for the low accuracy of subsequent productivity models. It is necessary and significant to improve the accuracy of LUE estimation.Based on global Fluxnet site data and MODIS LAI (Leaf Area Index)/fPAR (Fraction of Photosynthetic Active Radiation) products, this paper compared five existing LUE estimating methods (MOD17, CASA, Dry Matter Productivity algorithm of Global Monitoring for Environment and Security (GME DMP), VPM, EC-LUE) first. And in order to reflect the difference of LUE between the sunlit and shaded leaves, we took the Clearness Index (CI, the fraction of solar incident radiation on the surface of the earth to the extraterrestrial radiation at the top of the atmosphere) into account and established two LUE estimation models by stepwise linear regression method and parameter optimization method respectively. In the parameter optimization method, we first determined the maximum LUE of the sunlit and shaded leaves for each type of vegetation by an optimization algorithm developed at the University of Arizona (SCE-UA), and then we estimated the actual LUE by adjusting the maximum LUE with CI, temperature and water stress factors.The comparison of five existing LUE estimation methods shows that the R2 between estimated LUE and fluxnet LUE is within 0.028—0.282, and the RMSE range is 0.545—0.681 gC·MJ-1. The MOD17 method has the lowest RMSE at 0.545 gC·MJ-1, followed by EC-LUE with 0.579 gC·MJ-1. While R2 is highest for EC-LUE (R2=0.282), and followed by CASA (R2=0.185), which is related to the factor of evaporation fraction (EF) adopted by both methods. The correlation coefficient between EF and LUE is higher than other factors. On the whole, EC-LUE performs best (R2=0.282, RMSE=0.579 gC·MJ-1) among these five methods.The validation results show that the inclusion of Clearness Index can improve the accuracy of LUE estimation, and the RMSE by the stepwise linear regression method and parameter optimization method are both less than 0.5 gC·MJ-1. Although the stepwise linear regression method lacks the mechanism, the estimation accuracy of LUE (R2=0.461, RMSE=0.403 gC·MJ-1) is higher due to the more selected factors. The parameterization method has a slightly lower accuracy (R2=0.306 , RMSE=0.489 gC·MJ-1) due to fewer factors and a relatively fixed model form. The LUE estimation models established in this paper can be used for the estimation of regional or global LUE and vegetation productivity.关键词：light use efficiency;stepwise linear regression;parameter optimization;remote sensing;clearness index;gross primary production- 985
- |
- 401
- |
- 3
发布时间：2021-06-28 -
摘要：Pan-sharpening is a task in the field of remote sensing data fusion, in which multispectral (MS) images with rich spectral information but low spatial resolution and panchromatic (PAN) images with rich spatial details but only grey information are fused to yield images with high spatial and spectral resolution. Traditional Component Substitution (CS) methods replace a particular component of the MS image transformation with a PAN image, and then inversely transforms it to obtain the final fused image. The traditional MultiResolution Analysis (MRA) methods first extract spatial structures from the PAN image by using MRA transforms, and then the extracted spatial structure information is injected into the up-sampled MS images to obtain the fused image. The whole fusing process of the CS and MRA methods can be described as linear functions. However, the performance of such linear models are limited by their linearity, which often has spectral distortion. In recent years, many advanced nonlinear deep learning models have been proposed. However, those existing deep learning fusion models are relatively simple and pose difficultly in learn in-depth features. To overcome the shortcomings of the current models, we propose a deep recursive residual network that is specifically designed for the pan-sharpening task.Considering that the low-resolution input image and the high-resolution output image have high similarity, learning the relationship between input and output is highly redundant and difficult. If the sparse residual features between input and output are learned directly, then the network convergence can be significantly improved. Thus, the residual learning introduces the network structure, in which the introduced residuals include global residuals and local residuals. Such a structure is conducive to learning and not prone to overfitting. Moreover, the residual network can solve the problem of deep network gradient disappearance and gradient explosion well. Recursive network improves accuracy by increasing the number of network layers without increasing weight parameters. Specifically, as we use the residual network globally, recursive learning is introduced into residual learning by constructing recursive blocks structure, whereas multiple local residual units are stacked together in the recursive block. Through such an end-to-end network design, a better image fusion effect is obtained.Given that no ideal fusion result has been used as a label, we made a data set according to Wald’s protocol using the original MS as the ideal fused image, downsampling and then upsampling the MS as the MS of the network input, and the downsampled PAN as the PAN of the network input. To comprehensively analyze our experimental results, we performed a large number of simulation experiments and real experiments on the 4-band GaoFen-1 data and 8-band WorldView-2 data with abundant feature types. We then generalized them to 4-band GeoEye data and 8-band WorldView-3 data. Experimental results are compared with traditional methods and the existing deep learning methods. The subjective visual analysis and objective evaluation indicators show that the proposed method reduces the spectral distortion phenomenon of traditional methods and preserves the spectrum of an image better than the existing deep learning method does.The deep network designed in this paper has learned more in-depth and more luxurious image features and has achieved better fusion effects than existing methods. It uses a residual network to solve profound network gradient disappearance, gradient explosion, and degradation problems. In addition, the weight parameters are reduced by the design of the recurrent recursive block, and the network speed is improved. The generalization experiment shows that our network has a good generalization ability.关键词：remote sensing image fusion;space spectrum fusion;deep learning;convolutional neural network;residual network;recursive network- 812
- |
- 420
- |
- 4
发布时间：2021-06-28 -
摘要：Owing to vectorized capsule neurons and dynamic routing algorithm, capsule network possesses stronger feature representation capability than traditional convolutional neural networks. In the field of remote sensing, hyperspectral image (HSI) classification methods based on capsule network have obtained better classification results than traditional deep learning models. Aiming at the problems in the capsule classification models, such as shallow network layers and insufficient utilization of spatial–spectral information, this paper constructs a new deep capsule network for HSI classification. The designed network utilizes convolutional capsule layer, residual connection, and three dimensional convolutional capsule layer to further improve classification accuracy.The proposed method includes a convolutional layer, four capsule residual blocks, a class capsule layer, and a reconstruction network. First, the proposed method takes HSI data cubes as the input directly to retain the spatial–spectral details in the HSI. Then, the deep features in the input data are extracted layer by layer with the capsule residual blocks. Finally, the three dimensional convolutional capsule layer is introduced to make full use of the spatial–spectral information, so as to improve the classification accuracy.Three public HSI data sets including University of Pavia, Indian Pines and Salinas, as well as a large-scale hyperspectral data set Chikusei, are selected for the experiments. The results demonstrate that the proposed method outperforms the existing deep learning classification methods. Compared with the shallow capsule network model and the deep three dimensional convolutional network with residual structure, the proposed method improves the overall classification accuracy by 1%—3%, 2%—5%, 1%—2%, and 2%—4% on four different data sets. In addition, hyperparameters such as learning rate, capsule neurons dimension, network structure, and spatial neighborhood have been analyzed in detail. The effectiveness of three dimensional convolutional capsule layer has also been proven by conducting ablation studies.Compared with the existing HSI classification model based on capsule network, the proposed method has the following advantages. (1) The high-dimensional cubes are used for input data without any dimension reduction process, thus enabling the model to make full use of spatial–spectral details. (2) A deep network is constructed utilizing convolutional capsule layers and residual connection, thus allowing the model to extract more robust and abstract deep features. (3) The introduction of three dimensional convolutional capsule layer can make full use of spatial–spectral information in HSI to further improve the classification accuracy.关键词：hyperspectral image classification;deep capsule network;three dimensional convolutional capsule;three dimensional convolutional routing;deep learning- 772
- |
- 320
- |
- 4
发布时间：2021-06-28 -
摘要：Pan-sharpening (also known as remote sensing image fusion) aims to generate Multi-Spectral (MS) images with high spatial resolution and high spectral resolution by fusing high spatial resolution panchromatic (PAN) images and high spectral resolution MS images with low spatial resolution. Traditional pan-sharpening methods mainly include the component substitution method, the multiresolution analysis method, and the model-based optimization method. These fusion methods involve linear models, which are difficult to use in achieving the appropriate trade-off between spatial improvement and spectral preservation. In addition, they often introduce spectral or spatial distortion. Recently, many fusion methods based on deep learning have been proposed. However, their network depth is relatively shallow, and detailed information is inevitably lost during feature transfer. Hence, we propose a deep residual network with dense convolution for pan-sharpening.As the network becomes deep, the features of different levels become complementary to one other. However, most fusion methods based on deep learning ignore making full use of the information of each convolution layer. The densely connected convolutional network allows the features of all previous layers to be used as input for each layer in one densely connected block. To fully utilize the features learned from all convolution layers, we establish the multiple densely convolutional blocks to reuse features. Moreover, the information flow is accelerated by the transition layer between every two blocks. These maximize the use of features and extract rich features. Given the great correlation between deep features and shallow features, residual learning is used to supervise the densely convolutional structure to learn the difference between them, that is, residual features. Thus, residual learning combines shallow features and residual features to obtain further advanced information from MS and PAN images, which prepares for obtaining fused images with high spatial and spectral resolution.To evaluate the effectiveness of the proposed method, we conduct simulated and real-image experiments on the 4-band GaoFen-1 data and 8-band WorldView-2 data with multiple land types. The trained network is generalized well to WorldView-3 images without pre-training. The visual and the quantitative assessment results show that the high-resolution fused images obtained by using the proposed method are superior to the results produced by the traditional and deep learning methods. The proposed approach achieves high spectral fidelity and enhances spatial details by reusing features.The proposed method makes comprehensive use of the advantages of densely convolutional blocks and residual learning. In the feature extraction stage, different levels of features are connected in series through the densely convolutional blocks. This characteristic makes the transmission of features and gradients effective in alleviating the gradient disappearance problem and provides rich spatial and spectral feature for fusion results. In the feature fusion stage, residual learning is used to learn the difference between deep features and shallow features, that is, residual feature. Hence, the convergence speed of the network is accelerated. The experiment result shows that our network has good fusion and generalization abilities.关键词：pan-sharpening;remote sensing image fusion;deep learning;densely connected convolutional network;densely convolutional blocks- 901
- |
- 387
- |
- 3
发布时间：2021-06-28
Technologies and Methodologies
-
摘要：River discharge is the basis for water resources and ecological protection, and it is an essential part of the hydrological cycle. However, some large areas in the world lack hydrological data. How to obtain river hydrological data conveniently and accurately remains a hot topic in the prediction of ungauged basins, especially in small and medium rivers. Solving the problem of data acquisition in areas lacking traditional hydrological data is beneficial for the management of water resources and rapid prediction of water disasters. Remote sensing technology, which is a non-contact data acquisition method, has been applied in many fields. This technology breaks the limits of space in acquiring river discharge, especially in areas where obtaining basic data is difficult, and has become a preferred method of data acquisition.Therefore, this study tries to take advantage of low-altitude remote sensing in obtaining terrain data. The study combines high-density terrain data with classical Manning-Strickler formula to estimate the river discharge. This study is based on representative rivers in the Junggar Basin, which is the second largest inland basin in China. The Kazanying section, Bortonggu section, Anji Sea section, Daheiyanzi section, and Rocha River section are selected as study sections. Combined with the Manning-Strickler formula, Digital Orthophoto Map (DOM) and Digital Surface Model (DSM) image data obtained by an Unmanned Aerial Vehicle (UAV) are used to calculate the theoretical river discharge. The Manning-Strickler formula needs four parameters to calculate discharge: cross-section area, hydraulic radius, hydraulic gradient, and roughness. Cross-section area and hydraulic radius are extracted in cross-section, which is combined with DSM and measured data; hydraulic radius reflects the slope of the river, which is calculated with DSM and DOM in Arc GIS; roughness, an empirical value that measures the level of obstruction of water flow in the river bed and embankment, is obtained by using DOM. On the basis of the values of these parameters, we calculated the discharge in Kazanying section, Bortonggu section, Anji Sea section, Daheiyanzi section, and Rocha River section. Their calculated values are 28.73 m3/s, 46.29 m3/s, 104.84 m3/s, 19.77 m3/s, and 6.83 m3/s, respectively. To verify the effectiveness and accuracy of the low-altitude remote sensing method in river flow calculation, we used traditional measurements to record the measured values of river discharge. This study selected 20% of the allowable error as the standard of Relative Accuracy (RA) and used the measured value to evaluate the calculated value. Root Mean Squared Error (RMSE) and Mean Percentage Error (MPE) are important methods of evaluating accuracy, which are used as criteria for evaluating overall reliability.According to the established evaluation method, the results in the five sections show that the average error is 10.74%, the maximum value is 28.48%，the minimum value is 1.43%, the RMSE is 4.82 m3/s, and the MPE is 0.065. The reason why the maximum relative error occurs in the Bortonggu section is because the value of roughness is too small. The results of the accuracy analysis indicate that the method used in this study is reasonable and has a well applicability in the study area. Moreover, the results prove that the classic Manning-Strickler formula can be combined with low-altitude remote sensing data.To resolve the problem of river discharge monitoring in ungauged areas, this study developed a new method that integrates classic river discharge algorithms with low-altitude remote sensing. The advantages of UAV are fully exploited in this study, and the use of UAV complements the research gap in the acquisition of small and medium-sized river discharge through remote sensing data. Research results have significant values for the application of flow simulation methods in ungauged regions and also provide a new solution for rapid and convenient collection of hydrological information. This study has unique advantages for water resources management and water disaster monitoring in key areas.关键词：UAV;Manning-Strickler formula;low-altitude remote sensing;river discharge calculation;accuracy analysis;ungauged basins- 829
- |
- 313
- |
- 5
发布时间：2021-06-28 -
摘要：The ICESat-2 satellite, which operates through the technology of multi-beam single-photon-counting, provides a new opportunity for the mapping of global forest structures. Although previous studies based on the airborne simulator of ICESat-2 have shown that it has great potential to estimate forest structure parameters using single-photon-counting data, the performance of ICESat-2 needs to be examined due to the major differences between airborne system and ICESat-2. The National Aeronautics and Space Administration has released nine types of ICESat-2 data products, one of which is the vegetation canopy height and surface elevation data (ATL08). Therefore, the purpose of this paper is to evaluate the performance of ATL08 on the estimation of forest height by referencing to small-footprint LiDAR data.The basis for the accurate estimation of forest height using photon-counting LiDAR data is the correct classification of photons, that is, identification of noise photons, ground photons, and vegetation photons. The ATL08 only records photon classification and forest structure parameters, and the geographical coordinates of each photon are recorded in the Terrain Elevation (ATL03) product. Therefore, ATL03 and ATL08 need to be connected first according to the organization of two products. Taking the data of small-footprint LiDAR as reference, the ATL08 products were evaluated from two aspects. First, the classification results of photons are evaluated by referencing to the profiles from the Digital Terrain Model (DTM) and the Digital Surface Model (DSM) derived using the small-footprint LiDAR data. Then, percentile height metrics of each 100-m-long segment in ATL08 product were evaluated by referencing to corresponding metrics calculated using the point cloud data of small-footprint LiDAR.This study was carried out in two test sites: Snyder County in Pennsylvania, United States and Ketapang in West Kalimandin, Indonesia. These sites represented temperate forest and tropical rainforest, respectively. In the temperate forest, the classification accuracy of noise photons, ground photons, and canopy photons in ATL08 products were acceptable. The estimation accuracy of maximum canopy height was slightly better than that of mean canopy height with R²=0.61 against 0.54 and RMSE=10.71% against 16.78%. In the tropical rainforest with relatively low canopy cover, the ground photons in ATL08 could be correctly identified. However, in the forest with high canopy cover, the number of photons near the ground is inadequate. In addition, the identified ground photons in ATL08 were not enough to fit the terrain under the forest, which caused obvious errors in the estimation of forest height. The results of the quantitative analysis of the error of mean canopy height between canopy cover and terrain show that with the increase of canopy cover, the error of canopy height calculated by ATL08 will increase. In the tropical rainforest, the error of the mean canopy height will increase as the slope increases. When the slope is 0°—10°, 10°—20°, and 20°—30°, the error is 5.7 m, 6.6 m, and 9.3 m, respectively.Ie can be concluded based on the results of this study that ATL08 product could be used to estimate the height of temperate forests. However, in dense tropical rainforests, due to the limited penetration capabilities of LiDAR and difficulties in the correct identification of ground photons, the existing ATL08 data will be difficult to use for forest height estimation.关键词：ICESat-2;ATLAS;ATL08;forest structure parameters- 1631
- |
- 635
- |
- 5
发布时间：2021-06-28 -
摘要：Crop residue is the remaining stems, leaves, and fruit pods in the field after crop harvest. Crop residue plays an important role in the farmland ecosystem. Remote sensing technology has advantages in time and space, and it has become the main method to estimate Crop Residue Cover (CRC). Using remote sensing technology to estimate CRC can obtain information about ground CRC quickly in a large scale, which is of great significance to the promotion of conservation tillage. On the basis of a Sentinel-1 SAR image and a Sentinel-2 optical image, radar index and optical remote sensing index were constructed, respectively. The autumn and spring field sample data in 2018 and 2019 in Lishu County, Jilin Province were combined. The correlation of the remote sensing index and the maize residue cover was explored, and the method of soil texture zoning modeling was adopted to reduce the influence of surface background factors on the estimation of CRC. To further improve the estimation accuracy of maize residue cover, the radar index and optical remote sensing index were combined. Moreover, the optimal subset regression and soil texture zoning were used to establish the maize residue cover estimation model, and the estimation mapping of maize residue cover in the study area was then completed. Results show that: soil texture zoning modeling can effectively solve the problem of soil heterogeneity, thus improving the accuracy of inversion. The performance of each remote sensing index in autumn high coverage period in 2018 is better than that in spring low coverage period in 2019. The STI and NDTI index have strong stability and the best performance in optical remote sensing index. R2 is 0.701 and 0.697, respectively; whereas in the radar index, the correlation between- 1061
- |
- 311
- |
- 5
发布时间：2021-06-28 -
摘要：In order to obtain better phase delay estimation, improve the reliability and the practical operability of GNSS-MR(GNSS Multipath Reflectometry) soil moisture inversion, and also to simplify the complex process of the satellite selection, a multi-system multi-satellite GNSS-MR soil moisture inversion algorithm based on the robust estimation was proposed in view of the poor reliability and operability of the single system single satellite GNSS-MR soil moisture inversion and the least squares estimation of no robustness. In this algorithm, the spatial difference of multipath environment and the periodic characteristics of multipath were taken into account to screen the SNR (Signal to Noise Ratio) observations. Then, the phase delay combination representing the change trend of soil moisture was obtained by using the robust estimation based on IGGIII (Weight Function III Developed by Institute of Geodesy and Geophysics) weight function. Compared with multi-system multi- satellite combination (scheme 1) and the single-satellite combination (scheme 3), the experimental results showed that the multi-system multi-satellite combination (scheme 2) and the single-satellite combination (scheme 4) based on the robust estimation achieved higher modeling accuracy, which were benefited from the positive performance of the robust estimation. The correlation coefficients between the estimated phase delays and the measured soil moisture were 0.97 and 0.95, respectively, and the root mean square error of the soil moisture fitting residual were 0.010 and 0.012, respectively. At the same time, scheme 2 and scheme 4 also achieved higher soil moisture prediction accuracy, with the correlation coefficient between the predicted soil moisture and the measured soil moisture being 0.92 and 0.91, respectively, and the root mean square error of the soil moisture forecast residuals being 0.016 and 0.023, respectively. In addition, compared with scheme 4, scheme 2 not only adopted the robust estimation, but also adopted the multi-system multi-satellite combination, which contributed to better modeling effect and higher modeling accuracy. Moreover, because it could avoid the complex process of the satellite selection, scheme 2 owned better performance in GNSS-MR soil moisture inversion.关键词：multi-system multi-satellite GNSS-MR;soil moisture inversion;signal to noise ratio;delayed phase;robust estimation- 658
- |
- 218
- |
- 2
发布时间：2021-06-28 -
摘要：The dynamic variation of the ecological environment, as the material basis of human breeding, directly affects people’s quality of life. The terrestrial ecosystem continuously evolves with the developing human society. The rapid and accurate extraction of land-cover change information is conducive to ecological environment protection, the scientific management of natural resources, and the maintaining the human-earth relationship.The objective of this paper is to determine how to construct the dynamic ratio method with the unified threshold to quantitatively measure the change direction, the changed region, and the change type of wetland.On the basis of remote sensing ecological index (i.e., MNDWI, NDVI, and NDSI), the change detection method was proposed to extract the max-difference in the growing season (2018) and the inter-annual dynamic change rate, respectively. The Landsat images data in 2006 and 2018 were used. The main goal is the annual significant assessment of ecological change in land-cover, the determination of the ecological change direction of land-cover, and the analysis of the transformation types of land cover. Principal component analysis was used to extract the first principal component of the monthly max-difference and inter-annual dynamic change rate. Finally, 5626 samples were collected by using visual interpretation. Accuracy assessment was subsequently conducted on the result of the annual and inter-annual change detection.(1) The dynamic ratio method is suitable for the inter-annual change detection in the study area and has higher change detection accuracy and greater stability than single-index (or band) change detection. The overall accuracy of the method is 93.8%, and the Kappa coefficient is 0.876. (2) The monthly max-difference of the ecological indexes (i.e., MNDWI, NDVI, and NDSI) was used to quantitatively measure the annual ecological variation in the study area. Finally, the first principal component retained 83.04% change information. Moreover, land vegetation and water have a positive contribution to the surface ecological change in the study area. The normalized difference soil index contributed negative effects to land ecological change. The dynamic change detection method does not need to set the dynamic threshold, and is more suitable for the efficient updating of land-cover types on a global scale.The quick extracting method of change information with single-phase remote sensing image between two different times enriches remote sensing monitoring technology. Such enhancement is salient especially in the evaluation of land surface ecological factors. It also provides updated data with support for natural resource inventory at the global scale.关键词：the change detection;the Dynamic Ratio method;remote sensing land cover;Songnen plain;Northeast China- 697
- |
- 169
- |
- 0
发布时间：2021-06-28
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