Vegetation phenology information is a key indicator for evaluating climate-vegetation interaction， land coverage， and interannual productivity changes in ecosystems. Traditional phenological monitoring methods are based on visual observation， the monitoring range is limited and requires a lot of manpower and resources. As a new monitoring method in recent years， remote sensing technology has the characteristics of large monitoring range， convenient information acquisition and saving manpower and material resources. Its application has promoted the development of vegetation phenology dynamic monitoring research. Firstly， this paper combs the process of vegetation phenology remote sensing monitoring in recent years， and clarifies the existing remote sensing phenology monitoring system； The remote sensing data sources that can be used to establish vegetation growth curve are summarized， and the application scenarios of different data sources are discussed； The existing curve noise reduction algorithms and application processes are summarized， and the causes of errors in different methods are analyzed； The main vegetation phenology extraction methods are summarized； Finally， the remaining uncertainties in remote sensing monitoring of vegetation phenology， such as data resolution， vegetation phenology stage definition， and monitoring timeliness， were discussed， and the main directions for future research on remote sensing monitoring of vegetation phenology were prospected.
LiDAR is now widely used in many fields. Massive discrete point cloud data is obtained by scanning objects， which not only include accurate three-dimensional coordinates， but also intensity information. The point cloud intensity information reflects the reflection characteristics of objects to a certain degree. Due to the influence of many factors， the original intensity data has a large variability， which needs to be corrected in order to improve its value. First， in this paper， the application of intensity in forestry remote sensing， surveying and mapping engineering， ground feature classification and Marine environment survey were reviewed in detail. Second， the main factors affecting the intensity were discussed from the perspectives of atmospheric attenuation effect， scanner characteristics， target surface parameters and data acquisition. Then， the basic theoretical basis of intensity correction was discussed， the theoretical and empirical correction methods were summarized. The intensity correction methods of intensity normalization and radiometric calibration were also briefly introduced. Finally， some urgent problems of intensity correction technology were pointed out.
Evapotranspiration （ET） is a critical component of hydrological and energy balance models and plays an important role in the groundwater monitoring and agricultural irrigation， however heterogeneous surface can lead to spatial scale errors in estimates of latent heat flux using remote sensing data. Using Sentinel data as the research data， the EFAF （Evaporative Fraction and Area Fraction） method and the temperature downscaling method were used to correct the errors of the latent heat flux， and the differences between the two methods were compared. The results show that the accuracy of the EFAF（Evaporative Fraction and Area Fraction） method and the temperature downscaling method are comparable， the coefficient of determination （R2 ） is about 0.86， the Mean Bias Error （MBE） is about 18 W/m2， the Root Mean Square Error（RMSE） is about 64 W/m2. The accuracy of both methods is higher than that of the uncorrected latent heat flux， which has a certain effect on correcting the error of latent heat flux caused by heterogeneous land surface. The distribution of latent heat flux estimated by the EFAF method at the pixel scale is consistent with the land classification data， and at the regional scale with the uncorrected latent heat flux distribution. The latent heat flux estimated by the temperature downscaling method is highly similar to the distribution of the land surface temperature at the pixel scale， and its spatial details are richer and the local features are obvious.
Terrain effect will distort the surface reflectance of remote sensing images， which in turn affects the accuracy of Leaf Area Index（LAI） estimation from reflectance data. To attenuate or eliminate the influence of topography on LAI inversion， a data set of slope reflectance and LAI was constructed as train data based on the three-dimensional radiative transfer model DART （Discrete Anisotropic Radiative Transfer）. With reflectance as the input and LAI as the output， a mountain LAI inversion model was subsequently obtained using random forest algorithm. Then the estimation of LAI in mountain area was realized by combining remote sensing image data in the study area， and the accuracy of LAI estimation was evaluated by the ground measured data. Meanwhile， a flat surface inversion model was constructed based on the DART model and random forest algorithm as a comparison to evaluate the effectiveness of the method proposed. The results indicated that the mountain LAI inversion model considering the influence of topography showed a good performance， with the determination coefficient （R2） of 0.57 and Root Mean Square Error （RMSE） of 0.77 m2/m2， which was better than that of the flat surface model （R2 = 0.46 and RMSE = 0.86 m2/m2）.The mountain inversion model based on DART model can capture the influence of slope and aspect on the surface reflectance， and the inversion results can better restore the spatial distribution of LAI in the study area. This study proves that the mountain LAI inversion method based on coupling the DART model and random forest algorithm can partly reduce the terrain effect and effectively improve the estimation accuracy of mountain LAI， which can provide reference for the remote sensing inversion research of vegetation parameters over mountainous areas.
The quantitative relationship between FPAR（Fraction of Absorbed Photosynthetically Active Radiation）and vegetation indices has certain reference value for improving FPAR inversion accuracy and guiding production practice. Based on the three-dimensional radiative transfer model LESS， a module named LESS1D （formally released with LESS though www.lessrt.org） with advantages of simplicity of 1D model and accuracy of 3D model is proposed. Based on this model， the influences of vegetation canopy， coverage and other factors on the relationship between FPARgreen and 6 vegetation indices were explored in random homogeneous scenes and 3D heterogeneous scenes. The results showed that in homogeneous scenarios， NDVI， SAVI and EVI fit FPARgreen best in homogeneous scenarios， while NDVI and RVI fit FPARgreen best in heterogeneous scenarios. In heterogeneous scenes， the fitting accuracy of FPARgreen and vegetation index under different crown shapes is cylindrical > ellipsoidal > conical； When the vegetation coverage is low， the fitting accuracy of vegetation indices to FPARgreen is poor； As the solar zenith angle increases， the relationship between RVI and FPARgreen changes from linear to exponential. Canopy volume and canopy geometry are the key factors affecting the size of FPARgreen with different crown shapes， while leaf aggregation， vegetation coverage and vegetation index type are the relevant factors affecting the saturation effect of vegetation index.
The anisotropic land surface reflectance is characterized by Bi-directional Reflectance Distribution Function （BRDF）， which is the basis of quantitative optical remote sensing. The inversion of BRDF relies on multi-angular observations. Due to the limited observations from satellites， aerocrafts and goniometers， it is very critical to design a feasible sparse angular sampling to achieve high-quality BRDF inversion. In this study， based on RossThick-LiSparse Reciprocal （RTLSR） kernel-driven model， we designed the optimal angular sampling by using the PROSAIL model simulated reflectance and observations from POLDER and in situ by employing the angular information content to quantify the information which the observation geometry can contribute to the inversion. Firstly， the information content and BRDF inversion accuracy with different angle combinations are calculated. The relationship between them is then obtained， and the angular information threshold for high-precision inversion is -3.5. Secondly， the optimal observation plane and the least angles required in BRDF inversion were found out by analyzing the angular information content of combinations in each observation plane. It shows that the optimal plane is the principal plane， and the minimum number angles is 5 while the recommended number is 6 and 7. Thirdly， the optimal angle combinations under different solar zenith angles are found as relatively regularly distributed in the forward and backward scattering， and there should be two angles around the hot spot within ±10°. The validation finally proves that the optimal angle combinations are suitable for most land surface cover types except the snow/ice case and especially good for sparse vegetation， with RRMSE （Relative Root Mean Squared Error） of 0.14 in red band， and 0.046 in Nir band. The results of this study are useful for multi-angular satellite sensor design， multi-angular reflectance observation experiment and angular weights assigning in BRDF inversion.
Rice is one of the main grain crops in China， and rice yield is related to people's well-being. Timely and accurate acquisition of rice planting area information and its spatial distribution is of great significance for regional agricultural development planning and yield assessment. To solve the problems of rice mixing easily with other crops and optical data being susceptible to cloud and rain weather， taking the Sanjiang Plain in northeast China as an example， a complete rice phenological growth curve was constructed by using time-series water index SDWI and vegetation index NDVI， respectively， based on sentinel-1 microwave data and Sentinel-2 optical data. The spectral differences of four important growth stages of rice were analyzed， including transplanting stage， tillering stage， heading stage and maturity stage， and the planting area of rice in different phenological stages was extracted by threshold segmentation and combination of data of different stages， and compared with the traditional method based on single optical data. The results show that the proposed method can accurately extract the planting area of rice in several key growth stages in Sanjiang Plain， and is superior to the method using optical data alone. At the same time， the overall accuracy of rice area extraction from single growth period images such as transplanting period images can also reach 87.08%. With the completeness of growth period data， the overall accuracy of rice area extraction based on the whole growth period is also as high as 91.88%， and the Kappa coefficient is 0.834， which can meet the requirements of practical application. Therefore， the rice planting area extraction method combined with multi-source data can accurately and efficiently extract the rice planting area in different phenological periods in Sanjiang Plain， and provide a basis for short-term agricultural situation investigation and management and regional agricultural sustainable development.
To better predict the impact of grazing activities on changes in the spatial and temporal patterns of alpine grassland ecosystems on the Tibetan Plateau. A grazing experiment was conducted in an Alpine Steppe Meadow Site for Cold-season Pasture （ASMWP）， an Alpine Steppe Meadow Site for Warm-season Pasture （ASMSP）， and an Alpine Meadow Site for Warm-season Pasture （AMSP） on the Tibetan Plateau in July 2008. The Normalized Difference Vegetation Index （NDVI）， green NDVI （GNDVI）， Soil-Adjusted Vegetation Index （SAVI）， Aboveground Biomass （AGB） and Gross Primary Production （GPP） were obtained during growing seasons in 2012~2015 to compare the responses of plant production to grazing between warm-season grazing and cold-season grazing， and between different types of grasslands. Grazing significantly decreased the NDVI by 22.27%， SAVI by 23.50%， AGB by 17.28% and GPP by 22.48% at the ASMWP site， but did not significantly change the NDVI， SAVI， AGB and GPP at the ASMSP and AMSP sites across all the four growing seasons in 2012~2015. Grazing marginally significantly （p = 0.091） reduced the GNDVI by 15.33% at the ASMSP site rather than the ASMWP and AMSP sites. Therefore， the effects of grazing on vegetation index， aboveground biomass and gross primary production varied with grazing seasons and grassland types.
Diseases and pests have become one of the biggest constraints on rice yield. Traditional plant protection technology mainly relies on the vision and experience of plant protection personnel， which is subjective， time-consuming and laborious， and difficult to meet the needs of large-scale real-time monitoring. The development of remote sensing technology provides a large-area， all-weather， multi-directional data acquisition method， which can provide crop and environmental information for the identification and classification of diseases and pests， it is an important means to monitor and forecast rice diseases and pests in a large area. On the basis of expounding the mechanism of remote sensing monitoring and prediction of rice diseases and pests， this paper summarizes the research progress of rice diseases and pests monitoring and prediction from many aspects， such as multi-scale remote sensing monitoring method， forecasting method， construction of rice disease and pest monitoring and prediction models， monitoring and forecasting system， etc. ， the existing problems and future development trends of rice disease and pest monitoring and prediction are prospected. With the development of information agriculture and the fusion of multi-source data， the accurate and intelligent remote sensing monitoring and forecasting of rice diseases and pests will become more and more mature.
Landslides are powerfully explosive and destructive， and are one of the natural disasters with high frequency in the world， causing serious damage to people's lives and properties. Accurate and rapid extraction of landslides and obtaining the distribution range of landslides after a disaster are extremely important for landslide disaster investigation and hazard assessment. The method of landslide monitoring based on high-resolution satellite remote sensing images is investigated. Firstly， the decoding characteristics of landslides on high-resolution satellite remote sensing images are introduced， while the research progress of landslide extraction methods and accuracy evaluation and analysis methods are discussed， and finally the advantages and shortcomings of current methods are summarized， as well as the development direction of future research. The results show that the deep learning method has greater potential， and the combination of deep learning and other automated interpretation methods should be strengthened in landslide monitoring in the future to solve the influence of sample size on the model results， realize the migrability of the model， and improve its automation.
Identifying an ecological network contributes to improving regional landscape connectivity and biodiversity， which is also an important means to maintain regional ecological security.Firstly， taking the Mekong River basin as the study area， and the ecological source was identified by the Morphological Spatial Pattern Analysis （MSPA） ， the importance of ecosystem services， and the connectivity.Then， the ecological risk assessment index of the Mekong River basin was constructed by using Spatial Principal Component Analysis （SPCA） to evaluate the ecological risk in the Mekong River basin. The ecological resistance surface was obtained by modifying the ecological resistance coefficient of land use/land cover with the generated ecological risk assessment index. Finally， the Minimum Cumulative Resistance （MCR） model and current theory were used to extract the key structures of ecological security network， such as ecological corridors， ecological nodes， and ecological obstacle points. The results showed that： （1） the overall ecological risk in the Mekong River basin was relatively low and the proportion of the area with the ecological risk grade of medium or below was 76.93%. （2） The difference of ecological resistance among different patches of the same land use/cover type could be fully excavated from the ecological resistance modified by the ecological risk assessment index， which was more consistent with the actual situation. （3） The ecological sources of the Mekong River basin were mainly forest land and water body， covering an area of 638 000 km2， which accounts for 10.92% of the total area. （4） A total of 74 ecological corridors， 86 ecological nodes and 37 ecological obstacle points were extracted from the study， providing a reference for ecological protection and restoration， biodiversity maintenance， and land use policy adjustment in the Mekong River basin.
In the field of earth observation， cloud detection is not only an important part in the quantitative application of remote sensing， but also a key step in the application of satellite meteorology. In recent years， remote sensing image cloud detection based on machine learning has gradually become a research hotspot in this field， and a series of research achievements have been obtained. Systematically describes the research progress of remote sensing image cloud detection based on machine learning at home and abroad in recent 10 years， dividing the algorithm models into traditional machine learning model and deep learning model. Moreover， the specific models of two categories are introduced in detail， and the advantages， disadvantages and applications of different models are compared and analyzed. This paper focuses on the Support Vector Machine （SVM）， random forest and other methods in traditional machine learning， and the neural network models in deep learning， including Convolutional Neural Network （CNN）， improved U-Net network and so on. On this basis， the existing problems in the research of remote sensing image cloud detection based on machine learning are analyzed， and the potential development direction in the future is discussed.
Cloud detection is the basis for related applications using satellite remote sensing images. Aiming at the problem that the cloud detection process is easily disturbed by the complex surface environment， a cloud detection model based on the extreme gradient boosting（XGBoost） is proposed. The method uses Top-Of-Atmosphere （TOA） reflectance， brightness temperature and spectral indices to form a feature space； Then， Bayesian optimization was used to adjust the hyperparameters of XGBoost model. To test the performance of XGBoost in cloud detection， Landsat 8 remote sensing images of different cloud scenes were selected as test data， and the cloud detection results of XGBoost， random forest and decision tree were compared. The results showed that the cloud identification performance of the XGBoost cloud detection model proposed in this paper was better than that of random forest and decision tree， which showed the potential of XGBoost in cloud detection；and the F1 score and Kappa coefficient of XGBoost can reach more than 73% and 71% respectively. The achieved accurate cloud detection and can provide certain support for subsequent researches of cloud detection.
To explore the capabilities of a set of machine learning methods for vegetation classification in typical humid mountainous areas of south China， four types of the machine learning models， including Decision Tree （DT）， Random Forest （RF）， Support Vector Machine （SVM） and Adaptive Boosting （AdaBoost）， were used to build the vegetation classification methods based on the UAV remote sensing images， field observation data， and digital elevation models. A suit of performance matrics such as classification accuracy， kappa coefficient， mean square error， user accuracy， and producer precision were selected to quantify the performance of the four machine learning methods. The results show that the AdaBoost model has the highest classification accuracy for identifying the forest vegetation types indicating that the AdaBoost model has an obvious advantage for distinguishing the forest vegetation types. Regarding the classification of non-forest types， the performances of the four methods differ relatively large with the highest accuracy in the RF model. In general， the four models can achieve good classification results in the typical humid mountainous areas of south China. The AdaBoost model has the highest classification accuracy and Kappa coefficient， while the SVM model has the relatively lowest accuracy. Auxiliary feature information such as topographic factors and texture features provide important information for improving the classification accuracy.
Using the new generation of Geostationary Meteorological Satellite Himawari-8 data， a new adaptive threshold decision tree low-temperature fire spot detection algorithm is proposed. The algorithm is based on the localization recognition results of clear sky pixels and background pixels， using data at channel 2.3 μm and channel 0.86 μm. Taking Shanxi Province as the research area， the identification results were verified using the data on April 24，2020 and February 20，2021. The results of the new algorithm show： （1） the new algorithm can identify small fire spots earlier （sampling point 40 minutes in advance）； （2） the new algorithm has better accuracy on identifying the fire spot with small range and low temperature on grassland and cultivated land； （3） the new algorithm solves the problem of false alarms and misses of fire spot， providing a new idea for identifying fire spot as soon as possible and monitoring fire disaster effectively.
Synthetic Aperture Radar （SAR） radiometric calibration is the basis of SAR quantitative applications， and the calculation of the integral response energy of point targets is an important part of SAR radiometric calibration. In order to optimize the radiometric calibration method of high-resolution SAR system， according to the energy calculation model of integral method and the introduction of image context information， this paper proposes a calculation method of point target response energy of high-resolution SAR image based on the sliding window， which is verified by the airborne SAR data obtained by Xinzhou 60 remote sensing aircraft. The results show that the method can reduce the interference of background clutter， speckle noise and ground point target pointing error in SAR images on point target position selection， improve point target position selection accuracy， and the integral response energy obtained based on images is closer to the theoretical value of point target. Therefore， the method can enable the high-resolution SAR system to achieve higher radiometric calibration calculation accuracy.
Buildings in remote sensing images are an important source of urban big data collection and analysis. Developing large-scale and high-precision models for extracting buildings from remote sensing images is significance for building spatio-temporal big data platform of smart city. It also helps to the promote of urban intelligent computing. Current building extraction models usually use large-scale convolutional neural networks or multiple networks in series， supplemented by other boundary refinement algorithms to improve the accuracy of extracting buildings. However， the large-scale and complex network models consume high computing resources and require more training time or computing power， which is not conducive to large-scale and rapidly train network and do prediction. Also， the large-scale and complex models limit the deployment and application on portable and other terminal devices. Therefore， based on the idea of a cancellation balance method， this paper proposes a lightweight full convolutional neural network model and feature fusion scheme for large-scale and rapid building extraction from remote sensing images. The model parameters are reduced by about 40% compared with before lightweight. GPU memory usage is reduced by 33.61%， and the average training time and prediction time are reduced by 32.40% and 26.31%， respectively. The MIoU accuracy of the model after feature fusion is about 74.14% in the public data set， which meets the expectation of lightweight for building extraction models under the premise of ensuring high-precision accuracy.
The circulation and flow of water and heat resources directly affect the coupling relationship of regional ecosystems. Qilian Mountain National Park is one of the first national park system pilot areas in China， and also an important ecological barrier in northwest China. Using meteorological drought monitoring index （SPEI） and ecological drought monitoring index （TVDI）， the spatial and temporal distribution of heat resources were analyzed in Qilian Mountain National Park， during the period of 1989~2019. The Theil-Sen Median trend analysis and Mann-Kendall trend test were utilized to detect the spatiotemporal characteristics and trend on both seasonal and annual scales. Comparing the suitability of two drought index and further research on the periodicity was by Morlet wavelet analysis. The results show that SPEI is more effective in assessing hydro-thermal feature in the area studied； There was a slow change from drought to wetting as a whole， occurred around 2014；Severe drought events mostly occurred in summer； Taking Sunan Station-Yeniugou Station as the boundary of dry-wet trend， significant humidifying trends exist in most parts of the western region， while a zonal humidifying region runs through the western boundary； Under different time scales， SPEI has two types of periodic changes of 5~9 years and 15~20 years that run through the entire study period； The area is currently in a wet period， and there will be a continued wet trend in the future.
FengYun-4A is the first of the second generation of Chinese geostationary meteorological satellite series （FY-4 series）. The Advanced Geosynchronous Radiation Imager （AGRI） onboard the FY-4A is the new generation of geostationary imager with 14 spectral bands for providing measurements with high spatial and temporal resolutions. However， due to lack of green band， the application RGB composite image is limited， and the reconstruction of green band from other spectral bands is needed for enhancing the application. A spectral conversion method is developed based on Deep Learning （DL） technique. DL model is used to establish the relationship among spectral bands for simulating the FY-4A green band. Considering the channel characteristics of FY-4A/AGRI， AQUA/MODIS data are selected as the benchmark for spectral conversion. First， the relationship between FY-4A/AGRI and MODIS visible band is established， and the FY-4A/AGRI visible band is spectrally adjusted to MODIS benchmark. Second， using MODIS data with adjusted AGRI visible band for training， a DL model called Deep-Layer Perceptron （DLP） is developed for spectral conversion between green band and other visible bands. Finally， it is applied to FY-4A/AGRI for green band reconstruction. Verification results indicate that， the mean bias between MODIS green band reconstructed by the model and the observed value is within 0.01 in reflectance， and the mean bias of AGRI green band reconstructed by the model and MODIS observed value is within 0.02 in reflectance. This method can be applied to reconstruct the FY-4A/AGRI green band for various applications.
The cross-verification of HY-2B sea surface wind field and significant wave height products in the Arctic region is an essential prerequisite for the promotion and applying related products in high latitudes. It’s an important link in testing the quality and application scope of China's autonomous marine dynamic series satellite products. Based on the wind field and significant wave height data obtained by HY-2B satellite launched in 2018， cross-validation analysis between the autonomous marine dynamic satellite and similar international satellite products in the Arctic region was carried out. The observation data of the past two years show that：（1） the wind speed and direction of HY-2B satellite and MetOp-A satellite have high consistency and accuracy. The correlation coefficients of wind speed and direction were 0.97 and 0.99， and the root mean square error was lower than 1.03 m/s and 12.28°.（2） The significant wave height data of HY-2B and Jason-3 in the Arctic region have fewer matching pairs within the given space-time range， but the consistency and accuracy of the data are both high. The correlation between the two satellites is 0.99 and the root mean square error is 0.28m.（3） The autonomous HY-2B satellite wind field and significant wave height products are highly consistent with the international similar satellite products in the Arctic region， meeting the accuracy requirements of operational operation. Related products can meet the promotion and application of sea surface wind field and significant wave height in high latitude regions such as the Arctic.
The inversion of nutrient element content in plants through plant reflectance spectrum information can quickly monitor the states of wetland plants. Taking Suaeda salsa in the Liaohe Estuary Wetland as the research object， three machine models including Random Forest （RF）， Support Vector Machine Regress（SVR） and Backpropagation Neural Network （BPNN） are selected based on canopy hyperspectral data to model inversing carbon， nitrogen， phosphorus element content and ecological stoichiometry， and by the First Derivative （FD）， correlation analysis sensitive bands extraction method， principal component analysis to improve the accuracy of the inversion model. The results showed that the first-order differential processing can significantly improve the sensitivity of hyperspectral information to element and ecological stoichiometric characteristics， and also improve the modeling accuracy； comparing the cross-validation result of the test data with the inversion result of the train data show that over-fitting were had occurred during BPNN modeling， the SVR model has the worst accuracy in the two inversions， and the RF model has the most stable inversion effect and the highest accuracy. The research results can provide a basis for the inversion of wetland plant elements and ecological stoichiometric characteristics.
