Oil palm is a major economic crop and the area of land converted to oil palm cultivation in the tropics has expanded rapidly. Oil palm has become the world's largest source of vegetable oil and it provides tremendous regional economic benefits. However， the expansion of oil palm cultivation has led to the loss of forests， arable land， and peatland， which has caused severe ecological and environmental problems. Application of 3S （RS， GIS， GNSS） technology is useful for the collection， analysis， and management of spatial information， and is essential for both optimizations of the spatial distribution of land use and sustainable development. This paper analyzes the progress of 3S technology application in oil palm research on the basis of a literature review and scientometric analysis. The factors affecting the precision of oil palm mapping are also discussed. We established that papers describing 3S technology application in oil palm research are based primarily on the study of land cover change， and that scientific institutions and researchers in Malaysia， the United States， China， Indonesia， and the United Kingdom are the major contributors. Currently， the application of 3S technology in oil palm research includes oil palm mapping， oil palm land change monitoring， oil palm tree counting， tree age estimation， aboveground biomass and carbon storage estimation， suitability analysis， yield estimation， pest and disease monitoring， and plantation management. The accuracy of mapping is not correlated significantly with the year of publication of specific literature but is correlated with RS data sources and classification methods. The use of 3S technology in oil palm research is currently dominated by RS， which has been used in diverse fields of oil palm research. GIS technology is used mainly for oil palm land change mapping， suitability analysis， plantation management， and pest and disease monitoring， while GNSS is used largely as an additional tool in pest and disease monitoring and plantation management.
Mineral resources are important production materials for human survival and development， and the monitoring of mine environment is crucial for mineral resources exploitation and protection. Due to the advantages including large-scale， multi temporal and comprehensive， remote sensing technology has become the main means of mine monitoring. Aiming to the requirements of mine development and utilization， geological disasters， ecological environment monitoring and quality evaluation， we systematically summarized data sources， methods and models used in remote sensing monitoring of mine environment. Especially， data sources adopted in remote sensing monitoring of mine have tended to diversify and involve in all aspects of mine monitoring. Along with the rapid development of cloud computing platform and artificial intelligence technology， methods such as big data analysis and deep learning have gradually played an important role in remote sensing monitoring of mine environment， while multi-source data fusion， intelligent extraction of features， three-dimensional deformation monitoring and quantitative inversion are the main problems and challenges.
Accurate extraction of ground elevation and vegetation canopy height is of great significance for the study of topography， ecology and so on. The new generation of Global Ecosystem Dynamics Investigation （GEDI） launched in December 2018 provides an unprecedented opportunity for accurate extraction of ground elevation and vegetation canopy height over large areas. The purpose of this paper is to verify the accuracies of ground elevation and canopy height extracted by GEDI using airborne LiDAR data， and to explore the influence of geographic positioning error， terrain slope， aspect， vegetation cover， azimuth， acquisition time， beam type and vegetation type on the estimation accuracy. The results show that the estimation accuracies of ground elevation and canopy height can be significantly improved by correcting the geolocation error of GEDI data. The main factor that affects the extraction accuracy of canopy height is vegetation cover， followed by slope； while the extraction accuracy of ground elevation is significantly affected by the aspect and slope. Additionally， the results also indicated that the estimation accuracy is high when the vegetation cover is more than 25%， and the accuracies of ground elevation and canopy height are the highest in gentle slope area with slope 0~5°. Overall， this study will provide a basis for the screening and application of GEDI data.
Backpack Laser Scanning （BLS） is a potential tool in forest resource survey， but shows much uncertainty for the extraction accuracy of single-tree volume and forest stand volume in complex topographic circumstances. Using BLS point cloud data from the Gaofeng Forest Farm in Guangxi Province， this study implemented the estimation of single-tree volume and sample plot volume by random forest approach. First， individual tree segmentation was conducted using the BLS point cloud data， 8 characteristic parameters were extracted including Diameter at Breast Height （DBH）， Tree Height （Htree）， Crown Diameter （CD）， Crown Area （CA）， Crown Volume （CV）， Canopy Cover （CC）， Gap Fraction （GF）， and Leaf Area Index （LAI）， and 56 stratification height indicators were calculated （height percentage， cumulative height percentage， coefficient of variation， canopy undulation rate， etc.）. Then， an individual treee volume estimation model was developed using the random forest technique， and the prediction accuracy of various parameter combinations was investigated. The results showed that： （1） modeling with only 8 characteristic parameters of an individual tree structure indicated an estimated accuracy of R2=0.83、RMSE=0.097 m3； （2） modeling estimation accuracy was improved with the addition of the layered height index： R2=0.87、RMSE=0.087 m3； （3） the Boruta algorithm for variable screening reduced the input parameters from 64 to 52， with little difference in estimation accuracy： R2=0.87， RMSE=0.087 m3； （4） the estimation accuracy of sample plot volume was R2=0.97， RMSE=0.703 m3·ha-1. The results suggested the application potential to use the BLS point cloud for individual tree volume estimation and the sample volume by random forest algorithm.
Accurate monitoring of plantation resources is a prerequisite for improving the cultivation quality of plantation forests， the level of sustainable operation and management， and accurately estimating the effect of increasing carbon storage of plantation forests. The use of airborne lidar can obtain high-precision forest canopy structure information. However， the dynamic monitoring of plantation resources can only be effectively realized on the basis of obtaining multi-phase lidar point clouds in the same forest area and accurately matching them. Aiming at the characteristics of single tree species， regular arrangement， and lack of necessary texture features， this study created a highly robust multi-phase airborne lidar plantation point cloud matching algorithm based on the relative spatial relationship of the trees： First， Use ground points for rough registration， and perform individual tree segmentation on the two-phase point cloud to obtain tree position and height information， and extract individual tree matching features according to the relative relationship between the horizontal and vertical directions of the trees； secondly， establish a suitable similarity function， Combined with individual tree matching features to construct a weighted bipartite graph model， and use the Kuhn-Munkres algorithm to obtain the corresponding relationship between the two periods of plantation； finally， use singular value decomposition to solve the optimal transformation matrix to complete the registration. Tests were conducted in the typical coastal plantation research area of Jiangsu Province （the main tree species are poplar and metasequoia）. The results show that the matching algorithm created has a good registration effect in the typical sample plots of Metasequoia and poplar. Among them， the metasequoia plots （RMSE=42.5 cm， after registration） The result of registration is better than poplar plots （RMSE=58.8 cm，after registration）. This algorithm can effectively improve the matching accuracy and efficiency of multi-phase airborne lidar plantation point clouds， and provides a technical prerequisite for the dynamic monitoring of plantation individual tree （such as felling， growth， etc.）.
Leaf Area Index （LAI） is an important index for crop growth monitoring and yield estimation. Accurate and efficient LAI retrieval plays an important role in the macroscopic management of farmland economy. This study explored the potential of combining UAV LiDAR and hyperspectral data to retrieve maize leaf area index， studied the effects of different sampling size， height threshold and point density of LiDAR data on LAI inversion accuracy， and determined the optimal values of the three parameters. In this study， LiDAR variables and vegetation indices were extracted from resampled LiDAR data and hyperspectral imagery respectively. Then， based on Partial Least Squares Regression （PLSR） and Random Forest （RF） regression， LiDAR variables， vegetation indices， combined LiDAR variables and vegetation indices were used to construct prediction models， and the optimal prediction model for LAI inversion of maize was determined. The results show that the optimal sampling size， height threshold and point density of maize LAI inversion are 5.5 m， 0.55 m and 18 points/m2 respectively. We found that the highest point density （420 points/m2） did not obtain the optimal LAI inversion accuracy of maize. Therefore， it is not reliable to improve the inversion accuracy of LAI by increasing point density alone. The LAI inversion accuracies based on LiDAR variables （PLSR： R2 = 0.874， RMSE = 0.317； RF： R2 = 0.942， RMSE = 0.222） were higher than those based on vegetation indices （PLSR： R2 = 0.741， RMSE = 0.454； RF： R2 = 0.861， RMSE = 0.338）， and the inversion accuracies of the prediction model constructed using combination variable （PLSR： R2=0.885， RMSE=0.304； RF： R2=0.950， RMSE=0.203） were better than using single variable， in which the random forest regression model established by using combined LiDAR variables and vegetation indices is the best prediction model. Therefore， the fusion of the two data sources has a certain potential in improving the accuracy of vegetation LAI retrieval.
FY series satellites can provide important data support for remote sensing monitoring of the atmosphere， land， and ocean on a global scale. As optical satellite images are inevitably affected by cloud coverage， obtaining accurate cloud masks through cloud detection is the key to the processing and application of FY series satellite images. Most of the existing cloud detection methods use simple and efficient threshold methods， however， the optimal threshold in the traditional threshold method is difficult to determine in the absence of a large number of cloud and clear sky labels due to differences in sensor spectral response and radiance differences between different underlying surfaces. Therefore， a Threshold Adapted Cloud Detection （TACD） method is proposed in this paper， which has taken the band characteristics and underlying surfaces differences into consideration comprehensively， then sets up multi-channel threshold tests consisting of reflectance and reflectance combination test， brightness temperature test， brightness temperature difference test and cirrus cloud test under different scenarios， and establish global Optical-LiDAR cloud detection dataset to achieve iteratively optimize thresholds in TACD algorithm， and finally perform cloud detection based on the optimal thresholds. We take FY-3D MERSI-II images as an example to establish a high-precision global cloud detection sample dataset collocated with CALIOP cloud layer data， compare the cloud detection results of the proposed TACD method with the official cloud mask products. The evaluation results show that the accuracy of the cloud masks produced by TACD is significantly improved compared with the official masks， in which the mIoU is increased from 80.35% to 84.09% and the recall can reach 92.67%. In conclusion， TACD has great potential for application.
A method for Living Vegetation Volume （LVV） calculation based on laser-scanned point cloud data was proposed， by taking into consideration the differences of leaf density and spatial structure among various trees. The procedure is as follows： firstly， according to the main direction similarity and the axial distribution density of the local point cloud of the tree， the branches （non-photosynthetic components） and leaves are separated， and the leaf point cloud is extracted； then the voxel model of leaves is created. The Graham algorithm was utilized to determine the boundary of the layered tree canopy and the laser contact frequency， the Leaf Area Density （LAD） of the canopy was obtained based on the Voxel-based Canopy Profiling （VCP） method； finally， the living vegetation volume of the tree is accumulated by adding up the layered prism volume multiplied by the leaf area density. We used a Riegl VZ-400 ground-based laser scanner to obtain point cloud data of 13 trees of different shapes and species. The results show that the living vegetation volume calculated by our algorithm is 36.69% of that of the platform method and 47.80% of that of the convex hull method. Compared with traditional methods， this approach focuses on the statistics of the leaf point cloud of photosynthesis components distribution of the tree canopy by taking into account leaf area density， and the living vegetation volume estimation is more in line with the actual situation of the tree.
Hainan province is a golden place to develop tropical characteristic and efficient agriculture. It is of great significance to analyze the change of multiple cropping index with high spatial and temporal resolution. Based on Sentinel-2 data， maximum value composite and Savitzky-Golay filtering and smoothing were used to reconstruct NDVI time series curve. The second difference method was used to calculate the multiple cropping index of cultivated land in Hainan province from 2016 to 2020， and the spatial-temporal evolution characteristics of the multiple cropping index were analyzed. The results showed that the overall accuracy of multiple cropping index extraction in Hainan was 91.94% and the Kappa coefficient was 0.88， verified by the ground survey data in 2020. The multiple cropping index of hainan cultivated land increased from 1.53 in 2016 to 1.66 in 2020， an increase of 0.13. From 2016 to 2020， the single-season planting area increased by 6.10 percent， the two-season planting area decreased by 2.65 percent， the three-season planting area increased by 5.10 percent， and the fallow or abandoned farmland decreased by 5.60 percent. The multiple cropping index of all cities and counties in Hainan province is in the range of 1.28—1.96. The multiple cropping index of Haikou city， Sanya City， Dongfang City， Lingao County increases， while the multiple cropping index of Qionghai City， Wanning City and Qiongzhong County decreases. The results can provide data and decision-making support for agricultural departments in Hainan to adjust fallow and reclamation policies reasonably and implement sustainable development strategy of tropical efficient agriculture.
Land use suitability evaluation of the Hainan coastal zone is of important significance to realize the rational allocation of land resources. In this paper， we use the Land Use Suitability Evaluation Model （LandUSEM） to determine the spatial distribution and changes of land use suitability of the coastal zone in the study area， selecting suitable factors and using land use data from 2000， 2010 and 2020.The research shows that： （1） The annual average value of land use suitability in the coastal zone of Hainan Island is relatively high and the overall suitability is good. From 2000 to 2020， the overall land use suitability in the study area increases first and then decreases. （2） Among the four areas divided into the coastal zone of Hainan Island according to the latest development plan， the comprehensive land use suitability of southern group is the best， while which of the northern group is the worst.（3）There is a coupling between land use types and land use suitability. The suitability changes in offshore areas are relatively significantly， where the artificial surface has expanded significantly. It is necessary to limit the expansion of construction land of cities and towns in the coastal zone.
Carrying out coastal erosion vulnerability assessment of Hainan Island is of great significance to the protection of ecological resources and disaster prevention in coastal areas. In this paper， the coastal vulnerability index EI （Exposure Index） of the Integrated Valuation of Ecosystem Services and Trade-offs （InVEST） model is used to evaluate the coastal erosion vulnerability of Hainan Island. Then the evaluation system of coastal characteristics-coastal dynamics-economic and social indicators was established for the typical study area. Suitable evaluation factors for coastal zone characteristics were selected including coastal erosion rate， coastal type， coastal habitat， etc. The vulnerability index was quantified using the integrated index method. Finally， the vulnerability of the coast of Hainan Island under different scenarios， as well as the coastal erosion rate and erosion vulnerability class of the key areas were obtained. The study shows that： （1） The spatial distribution of erosion vulnerability on Hainan Island is low in the east and high in the west， with the highest vulnerability in the southwestern cities and counties， the lowest vulnerability in the southeastern cities and counties， and moderate vulnerability in the remaining areas. The coastal vulnerability in the habitat-free scenario is much higher than in the habitat-protected scenario. （2） The sandy shore of the east and west coast of Haikou in the typical study area is subject to more erosion from 2016 to 2020， with the most places exceeding 20 m/a. Coastal erosion vulnerability is high in the main urban areas of Haikou such as Longhua and Meilan District， followed by the west coast and east coast， respectively， and the lowest vulnerability in the Dongzhai Port area. （3） The study found that the coast under the protection of mangroves and other habitats can be effectively protected with very low vulnerability， while the degraded sandy shoreline shows high vulnerability， so it is necessary to protect coastal habitats and prevent coastal sediment loss.
Hainan Island is an important wet tropical treasure land in my country. Affected by factors such as natural geographical environment and conditions， frequent droughts and severe droughts and floods on the island have caused significant economic losses to agricultural production and people's lives. Based on MODIS data， this paper calculates the normalized vegetation index NDVI and land surface temperature LST， and then builds the vegetation water supply index model VSWI， and analyzes the characteristics of the drought in Hainan Island from 2004 to 2020. The conclusions are as follows： （1） During the period from 2004 to 2020， the overall drought in Hainan Island was more severe in 2004， 2005， 2010 and 2015， and the drought in Hainan Island in 2005 was more severe. The most serious drought area distribution can reach 56.76% of the total area， among which the area proportions of severe drought， moderate drought and mild drought are 7.37%， 20.75% and 28.64% respectively. （2） The change of the VSWI index in Hainan Island showed a unimodal trend of decreasing at first and then increasing. It showed a downward trend from January to May. The drought continued to increase with time. The drought reached its peak in April and May， and from June to December. Due to climate， the drought eased slightly. In May 2005， the drought was the most serious. 84.27% of the entire region was in various degrees of drought. The severely affected areas were concentrated in the southwestern region， and Danzhou City was the most severely affected. The area with extreme drought reached 35.57%， and the area without drought was only 2.31%. （3） Affected by geographical factors and climatic factors， the changes of VSWI values ??in different land use types tend to be consistent within the year. Forest land and grassland are less affected by drought， and cultivated land and urban areas are more affected by drought due to sparse vegetation. in April. （4） There are obvious seasonal differences in the spatial distribution of VSWI in Hainan Island from 2004 to 2020. In Hainan Island， winter drought and spring drought are the main ones， and summer drought and autumn drought also occur from time to time. The drought in each city and county has obvious regional differences and Seasonal difference， the coast is heavier than the inland， the surrounding is heavier than the middle， the south is heavier than the north， and the west is heavier than the south. （5） Hainan Island VSWI is closely related to rainfall and temperature factors. The correlation between rainfall and vegetation water supply index VSWI is the highest， and rainfall factors account for a large proportion of the area. Therefore， the drought in Hainan Island is mainly affected by meteorological factors. The research results can provide a reference for the drought warning in Hainan Island.
Xisha Islands locate in the tropical zone which frequently suffers from cloud cover. Optical systems are vulnerable to bad weather which results in data gaps or low data quality， resulting in difficulties in tropical surface monitoring. To solve this problem， a study on analyzing Xisha vegetation was conducted based on a low-altitude platform. Multi-spectral images were obtained via the DJI Phantom 4 UAV and four vegetation indices from five spectral bands were derived， including the Normalized Difference Vegetation Index （NDVI）， Grassland Chlorophyll Index （GCI）， Green Normalized Difference Vegetation Index （GNDVI）， and Normalized Green-red Difference Index （NGRDI） to analyze the vegetation growth in North island during May 2020. Combined with key meteorological parameters and Worldview2 optical images， the vegetation growth changes between 2020 and 2018 as well as their potential attribution were analyzed. Results showed that the average NDVI， GCI， GNDVI and NGRDI in North Island were 0.30， 0.84， 0.26 and 0.05 in May 2020， reflecting a low vegetation coverage and health status， which was consistent with the ground monitoring results. In 2020， the index difference between artificially managed and natural vegetated region increased from -23%—15% in 2018 to 15%—40%， indicating that the growth of natural vegetation is significantly worse than that of artificially managed vegetation in 2020 which demonstrates strong environmental stress. Meteorological data in this region showed that from April to May 2020， the average daily temperature and wind speed increased while the cumulative precipitation decreased compared with the same period of previous years， leading to increased soil water deficit. This may be the main reason for the deterioration of vegetation growth. These results demonstrated that DJI Phantom 4 images could effectively and quantitatively reflect the vegetation growth which will greatly support the ecological environmental monitoring over tropical islands.
Mangrove wetlands play an important role in protecting coastal ecological environment and resisting natural disasters such as tides and typhoons. Based on Hausdorff dimension， stability index and aggregation dimension of Fractal theory， the change characteristics of mangrove landscape’s spatial structure in Dongzhaigang National Nature Reserve of Hainan from 1987 to 2017 were analyzed by using mangrove remote sensing thematic information products. ①During 1987—2017， the total area of mangroves in Dongzhai Harbor showed an overall increasing trend. In 2017， the area of the mangroves reached 18.05 km2. The main expansion modes of mangrove landscape were marginal expansion and internal connection. ②During 1987—2003， the complexity and stability of mangrove landscape distribution showed a downward trend with strong dynamic， and the landscape structure was in an unstable state. During 2003—2013， the complexity and stability of landscape of mangroves gradually increased steadily， mangrove patches expanded and the landscape structure reached a steady state. After 2013， the complexity and stability of landscape of mangroves decreased slightly. ③Expansion of mangroves in Dongzhai Harbor is aggregate-centered and the aggregation dimension is 1.47. The aggregation centers have radiation effects to the growth of surrounding mangroves. The aggregated distribution area of mangroves in Dongzhai Harbor has reached a stable state of landscape development. Under the circumstances of maintaining and fostering properly， its aesthetic and ecological function can be developed moderately. While for those mangrove patches that are still expanding， sustainable development can be achieved through artificial afforesting and fostering.
The carbon budget of terrestrial ecosystems is an important indicator of global carbon cycle research and an important parameter of climate change. Based on the terrestrial ecosystem flux observation and remote sensing satellite observation data， machine learning methods are applied for carbon budget estimation. In this study， random forest algorithm is established to automatically learn features from training data and differences in time series dependencies， and carbon related parameters （Gross Primary Production， GPP； Net Ecosystem Production， NEP） could be estimated. Finally， standard indicators are selected to objectively evaluate the model using the validation data set. The result analysis shows that compared with MODIS GPP products， this method has greatly improved the estimation accuracy. Among them， the prediction result of deciduous broad-leaved forest is the best， the decision coefficient R2 is 0.82， and the root mean square error is 1.93 gCm-2 d-1.It is also significantly better than traditional light energy utilization model products in other vegetation types. The NEP machine learning model established based on the same method has also obtained good estimation results. The correlation between the output results of the deciduous broad-leaved forest model prediction model and the NEP obtained by the flux tower is 0.70 and RMSE=1.75 g C m-2 d-1. The difference in accuracy between GPP and NEP models indicates that when machine learning modeling is performed， the selection of independent variables in the training data set still needs to consider theoretical model. In order to quickly estimate the carbon budget of the terrestrial ecosystem， a remote sensing monitoring platform is established. The platform uses the GEE （Google Earth Engine） big data platform as the data storage and computing backend， and Django， HTML， CSS， JavaScript， etc. as the front-end， in order to quick calculation， real-time visualization and other functions. Based on the platform and algorithm， the global （60° N—60° S） GPP results obtained from 2002 to 2016 show that there are obvious spatial differences in the global average GPP， and the significant increase is mainly concentrated in eastern Asia and forested areas in North America. Research shows that remote sensing monitoring of carbon budget parameters based on machine learning and big data platforms can quickly provide regional and global-scale carbon storage and the results are consistent with true ground observations. The obtained estimation results avoid the complicated parameter setting of the physiological process model， and reduce the uncertainty of regional and global large-scale carbon budget monitor.
The traditional application mode of remote sensing satellites is complicated and complex， and unable to meet the demand for real-time remote sensing services that users are increasingly concerned about. To equip satellites with intelligent brains can reduce the data transmission bandwidth on the one hand， and improve the time-effectiveness of data acquisition on the other hand. Therefore， on-board intelligent processing has become an essential choice for the development of remote sensing satellites. However， it is difficult to debug on-board processing in orbit， and the existing ground test systems for remote sensing satellite on-board processing platforms are all formed temporarily for different satellite payloads during the satellite laboratory testing. They lack versatility and have not formed an integrated device， resulting in low efficiency of existing ground test platforms for on-board intelligent processing. Especially in the face of the new demand for intelligent on-board processing at present， there is a lack of an on-board processing ground simulation system with high performance， low power consumption and full process. Aiming at the new characteristics of automation and intelligent of remote sensing data processing development， this article proposes a set of ground simulation system for remote sensing image on-board processing based on the combination of FPGA and GPU. This system can realize the 0 to 1 level data pre-processing of multiple payloads in ground simulation and realize the accelerated recognition of intelligent remote sensing images on the basis of pre-processing. The key difficulties are as followed： the balance between the high computational complexity of remote sensing image intelligent processing algorithms and the limited computational power of embedded computers； the balance between the solidification of AI-specific algorithm and hardware acceleration in remote sensing image processing field； the balance between the testing requirements of different satellite platforms and the generality of system architecture. This article illustrates the approach of simulation platform design， builds a basic prototype and verifies it. The test results show that the system can better complete the whole ground testing process for typical algorithms of on-board intelligent processing， and all hardware can be directly assembled on-board with high completeness. It has a certain reference value for optimizing and guiding the operation management system of satellite ground simulation system.
For the multi-source and multi-scale remote sensing image registration method，there are too many pairs of error matching， and the low registration accuracy. In order to further improve the accuracy of remote sensing image registration， a remote sensing image registration method based on mutual information was proposed. Firstly， the SIFT algorithm is used for feature point extraction， and the FLANN is used to complete the rough matching. Secondly， establish a 4×4 neighborhood around the initial matching point， calculate the mutual information value between the matching points， eliminate the matching points with a smaller mutual information value， seek the optimal transformation matrix after filtering and optimize. Finally， the registered image with the largest mutual information value with the reference image is output as the best registration result. The experimental results show that this method can effectively eliminate mismatched points and improve registration accuracy compared with SIFT algorithm. Conclusion for traditional image registration method， registration mismatches have many pairs of points. The paper presents a new remote sensing image registration method. The experimental results show that This method can be applied to multi-source and multi-scale remote sensing image registration， and can effectively improve the registration accuracy.
Gully erosion is the major driver of land degradation and the unique landforms on the Loess Plateau. It is of practical significance to assess the applicability of extracting gully from satellite images with different resolutions and explore automatic gully extraction method. Google image （0.5 m resolution） and GF-1 images （2 m and 8 m resolution） were used to extract gullies automatically with object-based image analysis and random forest in Zhongduo tableland located in the southeastern Loess Plateau. Gully morphological parameters of 30 gullies extracted from three satellite images were compared to those from UAV data （0.14 m resolution）. The results were as follows： （1） The importance of image feature variables used for gully extraction is sorted as follows： spectral feature > texture feature > geometric feature. （2） The user accuracy and producer accuracy of gully extraction based on 0.5 m and 2 m resolution images were higher than 90%， while the user accuracy and producer accuracy reduced to 85% when 8 m resolution image was used. （3） The errors of gully length and width extracted from 0.5 m and 2 m resolution images were about 5% and 13%. The average error of extracted gully length， area and width from 8 m resolution image were 18.82%， 27.62% and 18.93%， respectively. （4） A model was put forward for improving the accuracy of gully length and gully area extracted from GF-1 image with 8 m resolution， based on the gully parameters extracted from 0.5 m resolution image， i.e.， L=1.22L'- 0.28， R2=0.896 and A=1.44A'+ 31.56， R2=0.916.
The existing methods are difficult to accurately estimate the volume of landslides， to solve this problem， the artificial intelligence algorithm is introduced， and transfer learning and differential algorithms coupled landslide volume estimation by low-altitude photogrammetry is proposed. Firstly， high-precision three-dimensional dense point clouds are derived from low-altitude UAV stereo images by using SfM and SGM dense matching algorithms， and ground point clouds are separated from the dense point clouds by combining visible light vegetation index and bilateral filtering algorithm. Then， a deep neural network for data interpolation is constructed to map the nonlinear relationship between two-dimensional coordinates and elevation information， and the elevation value can be predicted based on the transfer learning of parameter sharing and adaptive optimization， and the digital surface model of landslide area can be reconstructed. Finally， the volume of landslide is estimated based on the elevation difference of the digital surface model before and after the landslide in the target area and the differential algorithm. The experimental results show that the average relative error of the proposed method is approximately equal to 2.7%. Compared with the common methods， the proposed method can significantly improve the accuracy of landslide volume estimation， and is suitable for landslide volume estimation under different terrain.
Analyzing the morphological characteristics and influencing factors of urban evolution is essential for urban ecological governance and territory control. However， the current research lacked the long-term identification of the evolution process of urban morphological characteristics in northern China and ethnic minority areas. In this study， remote sensing data， land use/cover dataset， UISA and UGS fraction dataset were used to analyze the spatial and temporal characteristics of urban expansion during 1949 to 2018 and urban land cover change since 2000 in Hohhot. On this basis， the influencing factors of urban evolution was revealed， and the role of urban planning in the process of urban evolution was discussed. The results showed that the urban area of Hohhot has increased by 67.62 folds under the expansion speed ranging from deceleration to acceleration from 1949 to 2018， featured by the spatial morphology of infill—loop spreading—uniaxial spreading—biaxial spreading. Since 2000， the proportion of impervious surface in Hohhot showed a trend of “rising first and then decreasing”， while the proportion of urban green space showed a trend of “fluctuating increase”. The central city green space allocation has reached the green sustainable development configuration. China’s Western Development Strategy， the overall urban planning， and other related planning policies， as well as economic factors played important roles in urban evolution in Hohhot. The results can provide an important reference for the sustainable urban development of Hohhot and the northern China.
As one of the geological disasters causing huge economic losses and casualties， landslides have attracted more and more attention from society. In order to accurately identify landslide disasters in mountainous woodland areas， the Leijiashan landslide， which occurred on July 6， 2020 in Panping Village， Nanbei Town， Shimen County， Changde City， Hunan Province， was taken as the research object. Different fusion methods such as Principal Component Analysis （PCA）， Gram-Schmidt （GS） and Nearest-Neighbor Diffusion （NNDiffuse） are used to fuse the images of Sentinel-1A Interferometric Wide Swath （IW） Ground Range Detected （GRD） image after non-decibelization and decibelization with Sentinel-2A MSI2A image. Through the quality evaluation of the fused image， the PCA fusion method effect of the VV polarization image of Sentinel-1A after decibelization and Sentinel-2A image is the optimal， that is， the optimal fusion image is PCA-VV-DB. The Support Vector Machine （SVM） method was used to identify the landslide of the optimal fusion image （PCA-VV-DB） and the original optical image Sentinel-2A， respectively. Finally， the Sentinel-2A landslide visual interpretation results were used as the inspection standard to evaluate and compare the accuracy of SVM landslide identification results. At the same time， the Shaziba landslide in Mazhe Village， Tunbao Township， Enshi City， Hubei Province， on July 21， 2020， was used as a case to verify the feasibility of this scheme. The results show that compared with the single use of optical image for landslide recognition in the study area， the accuracy of landslide recognition using the optimal fusion image is increased from 95.24% to 96.65%， and the quality of landslide extraction also increased from 87.18% to 91.84%. The leakage recognition and excessive recognition of landslides are reduced， and the research scheme is popularized. It shows that the fusion of optical image and Synthetic Aperture Radar （SAR） image can improve the accuracy of landslide recognition in mountainous woodland areas， and provide valuable information for landslide risk assessment， disaster emergency investigation and disaster recovery and reconstruction.
To address the problem of poor registration accuracy caused by large nonlinear radiometric differences and local geometric distortions between multisource high-resolution images， this paper proposes a multisource high-resolution image registration method based on edge features. Our method first constructs the nonlinear scale space for the input images by anisotropic diffusion filters， on the basis of which the extended phase congruency maximum moments are calculated for each scale to obtain rich edge features， and extracts stable feature points using a FAST detector based on a blocking strategy. Secondly， a Main Orientated Index Map （MOIM） was generated using multiscale Multi-Orientation Log-Gabor filters and combined with Gaussian weighting to construct a robust feature descriptor. Finally， the corresponding points are obtained using the Bhattacharyya distance and Fast Sample Consensus （FSC） method. Multiple sets of multisource high-resolution images are selected for experiments， and the results show that proposed method can effectively overcome nonlinear radiometric differences and local geometric distortions between multisource high-resolution images， with better registration results than other related methods， and an average registration accuracy of better than 1 pixel.
The Fraction of absorbed Photosynthetically Active Radiation （FPAR） is a key parameter for carbon balance and climate change in wetland ecosystems， which directly reflects the growth and development of wetland vegetation. The empirical statistical method based on vegetation indexes is simple and efficient， and which has been widely used in the simulation of FPAR of grassland， forest and crop vegetation， but it is rarely used in wetlands. There is a lack of systematic research on the adaptability of different vegetation indexes to wetland FPAR estimation. In this paper， 14 common vegetation indexes are compared， and the optimal vegetation index is selected to invert the FPAR of the wetland in the Zoige Plateau during the growing season. The results indicate that the MSAVI index dynamically considers soil information， and can better adapt to the estimation of wetland vegetation FPAR among the common vegetation indexes， and its error and R2 are better than other vegetation indexes. The FPAR value of the Zoige Plateau wetland in the growing season is between 0.22 and 0.8， and the overall distribution is relatively uniform. The average FPAR of peat wetland， wet meadow and marsh wetland are 0.46， 0.63 and 0.58 respectively. During the growing season， the FPAR of different types of wetlands on the Zoige Plateau showed a trend of first increasing and then decreasing with time.
The cloud top spectrum-microphysical information of the typhoon precipitation cloud system is closely related to the change of precipitation intensity， and it is an important parameter for the quantitative inversion of precipitation from satellite remote sensing. Taking seven typical super typhoons and five non-super typhoon in the Northwest Pacific from 2001 to 2012 as examples， Based on the fusion data of rain radar and visible light infrared scanner onboard tropical rain measurement satellites， Random Forest （RF） model of typhoon precipitation retrieval is established by using the precipitation cloud top spectrum-microphysical parameters. The cross-validation of RF model shows that the correlation coefficient between the retrieved precipitation intensity and the observed precipitation intensity is 0.773， and the root mean square error is 0.299 mm/h， indicating that the model has a high precipitation inversion accuracy. Among all the cloud top spectral-microphysical parameters input in the random forest construction process， the 3.7 μm cloud top brightness temperature contributes the most to the variance of the RF model， exhibiting the highest importance. As cloud system is mature and its cloud top is high at the typhoon center， 10.8 μm Infrared channel brightness temperature， cloud optical thickness， and cloud water content contribute little to the variance of the model and are of low importance.
