The marine pollution is still a global lhot issues;artificial dumping activity is the main cause of this kind of problem,monitoring method was used to study the ocean dumping in order to grasp the marine environment change and pollution levels better.This paper emphatically summarized the research progress of traditional monitoring methods and remote sensing application technology in the study of marine dumping,and further analyse the limitations of traditional methods and polar orbiting satellites in the study of monitoring,and then discusses the progress and advantage of geostationary ocean color satellite in pracital application.Finally,combined with the actual application results,the paper pointed out the problems of geostationary satellite in the research process,and looked into the distance of future studies.
Heavy metal pollution in soil is one of the major environmental issues in the world.The issue that how to obtain the information of heavy metals content and distribution in soil rapidly and efficiently,which is the prerequisite for assessment and evaluation of pollution risk.Hyperspectral remote sensing has the advantage of offering non\|destructive and real\|time information while the study of retrieving soil heavy metals content by hyperspectral remote sensing is still under discussion.So this paper introduced and analyzed the achievements of inversion mechanism and modeling methods.The analysis shows that fully study of the adsorption properties of heavy metals in different soil types is the basis for deeply understanding inversion mechanism of heavy metals content in soils.This paper offered to strengthen the comparative analysis of different kinds of soil and different degree of pollution cases,develop the spectral signal processing methods,and introduce the algorithms for nonlinear models and setting up the model for chemical forms of heavy metals,which are the key research directions to estimate heavy mental content in soil by remote sensing.
Snow depth products derived from SSM/I passive microwave remote sensing data were evaluated against the ground\|based measurements across the Eurasian continent.The preliminary results show that the Chang algorithm underestimates and overestimates snow depth over the former USSR,China and Mongolia.The bias is 7.6,9.2 and 11.4 cm,respectively.In other words,the average error can be up to -24.3% for the former USSR,108.8% for China and 180.9% for Mongolia.Although the Che algorithm has relatively better results in China and Mongolia than the Chang algorithm,it still underestimates snow depth of the USSR by on average 21.3 cm or -68.6%,For the six snow types,the tundra snow and prairie snow,which always have less vegetation and relatively flat surface.Both the Chang and the Che algorithms have a trend to underestimate snow depth for thick snow cover and overestimate snow depth for thin snow.For the Che algorithm ,it underestimate snow depth for regions north of 40°N and overestimate the snow depth for regions south of 40°N.Similarly,the Che algorithm overestimates the snow depth for snow cover with <6.73 cm,in thickness and underestimate for snow cover with >6.73 cm in thickness.Accordingly,there is no single algorithm which can be used for global application in snow depth estimation,especially,in mountain and forested regions.
Based on the Advanced Integrated Emission Model (AIEM),this study established simulation database of multi\|angular bare soil emissivity at band\|C which contains a wide range of soil parameters,and uses the simulation data to analyze the relationship of the bare soil emissivity polarization differences between observation angles.Therefore,this paper used ω\|τ model to develop an inversion method to estimate vegetation optical depth,and using the measured values obtained by ground based microwave radiometer to invert winter wheat optical depth.The analysis result shows that the trend of inversion value of winter wheat optical depth is consistent well with the trend of measured values of LAI of winter wheat,which proves that the inversion method is feasible.
Multiple cropping is an important cultivated pattern of agricultural activities.Information related to cropping intensity of croplands is critical for food security,management of land resources,biogeochemical cycling and energy balance of agro\|ecosystems.Through analyzing the relationship between crop phonological metrics and time series vegetation index curve,this paper use the assembly of characteristic phases,which is composed by though and peak values of time series VI curve,to describe the crop growing processes,then extract the cropping intensity based on rank order morphological filter.We tested the proposed method in double\|cropping region of north china plain,with a window’s size of 13 as structuring element to filter the EVI time series data.The MODIS\|derived multiple cropping land information were validated with 30 meter resolution referenced multi\|temporal HJ data.The results indicated that the method can effectively extracted the cropping intensity with overall accuracy of 88.11% and Kappa coefficient of 0.765,which is favorable to extract the cropping intenstiy by setting a suitable structuring element window’s size in a wide range.
The data products of Microwave Scattermeter (QuikSCAT) and Microwave Radiometer (WindSAT) 10 m wind vectors are the largest coverage and longest time series for sea surface wind field at present.This paper compares the raw orbital data of QuikSCAT and WindSAT,with the buoy observation data in global ocean respectively.Comparisons showed that the average speed of QuikSCAT wind is slightly larger than buoy,and the average absolute bias are about 1 m/s;the average absolute bias of wind direction are less than 8°.The average speed of WindSAT wind is slightly larger than buoy,and the average absolute bias are not larger than 0.5 m/s,and the root mean square bias are about 1 m/s;the average absolute bias of wind direction are 10°or less.In global ocean,QuikSCAT and WindSAT wind vectors have high quality and can be relied.The transit time of QuikSCAT and WindSAT over the same point is synchronization.The wind vectors between QuikSCAT and WindSAT have good correlation and can be replaced each other.
February 11,2013,Landsat 8 was successfully launched in Vandenberg Air Force Base,California.The thermal infrared sensor can provide a new data source which used to retrieve land surface temperature,but there is no corresponding retrieval algorithm for it until now.According to USGS,given larger uncertainty in the Band 11 values,users should work with TIRS Band 10 as a single spectral band and should not attempt a split\|window correction by both TIRS Band 10 and 11.In this paper,we revised coefficients of the mona\|window algorithm to retrieve land surface temperature(LST) using Landsat 8 thermal infrared data (band 10),so that the new Landsat thermal infrared data could be effectively used.To validate the revised algorithm,the land surface temperature 20 ℃,30 ℃,40 ℃ and water vapor content 1,1.5,2,2.5 g·cm-2 were assumed,then the thermal radiance on the sensor were simulated respectively using MODTRAN atmospheric simulation program,and then retrieved the LST using the improved mona\|window algorithm.The result shows that a lower land surface temperature and lower water vapor content can keep a low error,and the average error is 0.74 ℃,which indicate that the revised algorithm able to provide an accurate LST retrieval from Landsat 8 thermal data.In the last,this paper retrieved the land surface temperature of Dianchi Lake Basin based on an image collected on April 20,2013,the result fit well with the actual conditions,and then analysed its distribute characteristic.
Terrestrial LiDAR systems have received lots of attention on three\|dimensional (3D) structure reconstruction for trees,especially on the branch skeleton generation.On the basis,a method is proposed to add leaves structure based on point density by dividing small cube in the canopy to reduce the influence of uneven distribution of point cloud,and combine gap fraction model to retrieval leaf area of a tree using terrestrial LiDAR data.It is successfully applied to reconstruct 3D trees using points simulated data by ray tracing algorithm as well as field measured points data.The relative error of leaf area between reconstructed and real structure is less than 0.9%.Meanwhile,the most relative error of directional gap fraction is also less than 4.0%.The experimental results prove that the method has gotten a satisfied consistency on visual sense and quantitative evaluation between the 3D structure reconstructed and real structure.In quantitative remote sensing,the method has lots of application values for simulation of canopy radiative transfer process as well as visualization.
Nitrogen is the largest amount of nutrition elements needed by crop.At the same time,it seriously affect growth and quality of crop.There are rapid,low consumption and non\|invasive advantages of Hyperspectral remote sensing on nitrogen content inversion.This paper proposed the method of using multi\|spectral index and SVM model,which could improve the nitrogen content inversion accuracy.Using Bacterial Speck of processing tomato actual spectral measurement and nitrogen content in the different growth stages and different severity level.Through the correlation analysis and the absolute coefficient R2 and F value of linear regression,we choose spectral index:PSSRb,ND705,GMI\|2,PTBSc.They are acted the input variable of SVM model,and invented the nitrogen content,and analyzed uncertainty.The result show:the correlation coefficient between real value and predict value is 0.711 and the value of MSE is 0.021;and the value of average relative error is 0.007.Compared with exponential model of the single spectral index:PSSRb and GMI\|2,the predict ability of SVM model is tetter.Then the correlation coefficient between real value and predict value is 0.720,which is maximum.So the multi\|spectral index and SVM model prediction has a better fitting effect than the single spectral index.
The signal\|to\|noise ratio (SNR) is a very important parameter for assessing the quality of remote sensing images.The accurate measurement of the SNR parameter from the corresponding remote sensing image through suitable methods can contribute to the prediction and improvement of the information extraction performance of the remote sensing image for data providers and data users.The existing methods of measuring SNRs are mostly aimed at the median/low spatial resolution images,and are unsuitable for the high spatial resolution images.Considering the limitations of existing methods of measuring SNRs,a method was applied according to the characteristics of high spatial resolution remote sensing images.Based on the adaptive DN subinterval divisions,the noise values of all the DN subintervals were estimated by the mean values of the minimum local standard deviations of the certain percentages,and then the SNRs of corresponding DN subintervals were estimated further.The method overcomes the difficulty of choosing homogeneous areas caused by more detailed spatial information,edge information and texture information.The results show that the improved method has good applicability for estimating the SNRs of high spatial resolution remote sensing images.
How to improve the accuracy and speed of image segmentation algorithm that has a significant effect upon the interpretation of remote sensing images.In this paper,a region\|growing\|based method is proposed for remote sensing image segmentation.This method is composed of two procedures which are local the best merge and global best merge.The first step focuses on performing high\|speed image segmentation.In implementation,a threshold is introduced into the first step to reduce erroneous region merges.The second step aims at increasing segmentation accuracy.In order to raise the running speed,an advanced data structure and red\|black tree are utilized to implement the second step.Finally,simulated remote sensing image and Orbview\|3 high resolution images are used to carry out segmentation experiment.A supervised image segmentation accuracy evaluation method is utilized to quantitively validate the performance of our algorithm.The experiment result indicates that the proposed method can achieve satisfactory segmentation in terms of segmentation accuracy and speed.
,planning and protecting of coastal zones.Nowadays,it becomes one effective measure to investigate inshore aquaculture with remote sensing image processing.In this study,index of spectral feature,addresses rapid detection of offshore aquaculture activities,were constructed by analyzing the spectral feature of the fish raft and the kelp/porphyra zone in area of Sanduao Sea around Ningde City in Fujian Province.Then an algorithm of threshold detection based on the texture of the index was proposed to identify the fish raft and the kelp/ porphyra area.The model was selected to remove sea\|background,and the Otsu was used to discriminate fish raft and kelp/ porphyra area.The experimental results showed the proposed method can monitor the different aquaculture area in a fast and accurate way,and ensure the accuracy to reach higher than 90%.
High\|resolution reconstruction techniques are investigated for HaiYang\|2 Scatterometer (HY2\|SCAT) backscatter measurements for a variety of application purposes including land and polar ice detection and the studies of cryophere and biosphere.Three resolution enhancement algorithms,including Additive Algebraic Reconstruction Technique (AART),Multiplicative Algebraic Reconstruction Technique (MART) and Scatterometer Image Reconstruction (SIR),are compared by simulation approach.The results that SIR has the best performance in noise suppression and resolution enhancement of the backscatter measurements than AART and MART algorithms.This study is verified using HY2\|SCAT Level\|1B (L1B) backscatter measurements.The SymbolsA@0 measurements over some sites with particular backscattering characteristics(e.g.islands) are reconstructed.
The construction of DEM in mountain plantation landscape,is the basic content of describing landscape topography,plays an important role in extracting information of plantation landscape such as size,structure and living stock volume,the study has important significance.In this paper,orthophotos and DEM were extracted based on images captured by UAV platform by means of splicing stereo pairs,and were compared with GPS measured data,ASTER GDEM data and SRTM data.The result shows that,the DEM extracted based on UAV images has the closest difference with GPS measured result (RMSE=8.96),which has higher precision than ASTER GDEM (RMSE=13.68) and SRTM (RMSE=11.81) data;The closest result(RMES=1.813) of max tree heights and max elevation differences in each sample plot,shows that UAV DEM has an ability to reflect more hierarchical information between canopy and ground,which shows a greater potential in the construction of DEM in mountain plantation landscape.
P.euphratica and T.ramosissimaare are indicator species of ecological environment in arid desert area,the extraction of their tree\|crown is the basis of quantitative monitoring of desert habitat by means of remote sensing.In this paper,Taking P.euphratica and T.ramosissimain in the lower reaches of Tarim River as the study object,the method of single data source SVM (Support Vector Machine)based on spectrum characteristics,SVM method based spectrum and texture characteristics,object\|oriented classification method and maximum likelihood classification method was used to extract tree\|crown from the QuickBird image.Single data source SVM method and maximum likelihood classification method is applied to classify the image which contains only spectrum characteristics.Other methods were carried out as follows:Firstly,calculating the textural measures by grey level co\|occurrence matrix and determining the optimal parameters for texture information by principal component analysis.Secondly,the optimal texture bands and the spectrum bands were combined into a new image.Finally,the support vector machine method and object\|oriented classification method was applied to classify the new image.The results show that:(1)The classification accuracy of SVM method based spectrum and texture characteristics is 9.65%,wich higher than that of single data source SVM method,the estimated accuracy of average crown diameter of the former is 7.18%,which higher than the later.The result indicates that texture is an important factor to improve the classification accuracy in high resolution images;(2)The tree\|crown extraction accuracy of object\|oriented classification is the highest.Its classification overall accuracy is 86.47%.The accuracy is 15.67% higher than single data source SVM method,6.02% higher than SVM method based on spectrum and texture characteristics,and 22.58% higher than maximum likelihood classification.Its estimated accuracy of average crown diameter is 87.45%,which suggests that object\|oriented classification method can effectively extract tree\|crown information in high\|resolution image and is better than the other classification methods.
Spaceborne microwave scatterometer is the only sensor for directly measuring sea surface wind speed and wind direction at the same time,which plays an important role in marine scientific research.Objects’ backscatter coefficient can become valid physical quantity including position information by geolocation,therefore,geolocation is an important part of scatterometer data preprocessing,and is closely related to the quality of remote sensing data.Rotating range-gated fanbeam Scatterometer is combined with advantages of fixed fanbeam scatterometer and pencil beam rotating scanning scatterometer,which hasn’t been in orbit.Based on the special working type,its scanning type was analyzed in detail,an appropriate geolocation method was proposed,geolocation results were calculated,afterwards error sources were analyzed,the satellite attitude influences on geolocation results were discussed,finally the relative accuracy of geolocation was evaluated using HY-2 satellite data products.
Medical image segmentation,which is difficult to solve problem in the image segmentation,has important application in the medical field.For the features of medical image and the differences of the image segmentation algorithm,this paper prensents an improved medical image segmenration algorithm using the level set.Firstly this paper proposes the velocity function and the core of level set algorithm by the curve evolution emulation.Secondly,we use rough segmentation to image by choosing velocity function,changing the region of strong gray value into that of small gray value,and finding out the object region in image according to emulation evolution exactly.Finally,we fulfill exact segmentation in medical image through initial algorithm with the help of the chosen velocity function based on times of iterative operation.Experimental result shows that the proposed algorithm is able to identify the region of tumors exactly,as well as it is robust to image segmentation efficiently.
With the development of LiDAR technology,the size of acquired data,known as point cloud has increased rapidly.However,the visualization of massive point cloud proves difficult due to the limited computer memory.To effectively visualize the massive point cloud,numerous data structures have been proposed.Among them,pyramid index structure is an effective method of the fast display for massive point cloud.In order to overcome the display delay caused by limited computer memory,this paper introduces pyramid index structure and proposes an innovative index structure that called superimposition pyramid index structure to improve the performance of the fast display for massive point cloud.First,point cloud data is divided into several cells.Within each cell,the point cloud data is then rearranged according to the structure of superimposition pyramid index proposed in this paper.Finally,each cell is organized to form an index file.The above structure has been implemented and the experiment on point cloud data from airborne LiDAR proves that the superimposition pyramid index structure effectively reduces the consumption of memory cost and accomplishes the fast display for massive point cloud.〖JP〗
In order to improve image quality of band 5 and band 27,stripe noises were acquired by Moderate Resolution Imaging Spectroradiometer (MODIS) level 1B.Based on MODIS scanning characteristics a method of use the max mean of each swath was proposed to judge the stripe noises.As the time of destriping noises,according to the thought of single line stripe interpolation at band 5,an interpolate method was proposed to replace the adjacent multi\|line strpe noises at band 27.Finally,comparison diagram,mean diagram and numeric analysis between original data and processed data were compared to validate the effect of destriping noises.The results show that the method can judge all the stripe noises exactly at both bands,and can remove the strpe noises well.The process of destriping is easily and suitable for the complex remote sensing scenes.
Online hydrological or land surface models can provide functions for a wide range of users with data analysis,forecasting,warning and decision supporting.With the increasing development of science data sharing,how to utilize those abundant online data in a more convenient way becomes a new challenge for online model.Current online model services have deficiencies in standardizing the interfaces for passing data and parameters which are heavily varied with model.In addition,they are lack of close collaboration with scientific data center generally.This paper proposes a novel framework for online model services for hydrological and land surface models in which model server,data server and client application closely collaborate.In accordance with a general work flow of using an online model service,we established standardized interfaces through online model services and data services are offered.This paper elaborates the architecture,major interfaces,and key techniques used to solve main challenges during implementation,which include the means of communicating data and parameters with model server and formatting them into the formats desired,retrieval of data from the data server,and the retrieval of simulation results upon request.The proposed framework takes advantage of strong computing capability of a server and the rich resources of a data center,and it is able to be replicated and extended to other similar applications.
Leaf area index is a crucial parameter of vegetation structure which provides key information for earth surface process simulation and climate change research on the global and regional scales.Focusing on the meadow steppe in Hulunber,Inner Mongolia,the accuracy of MODIS/LAI product in this region was analyzed.Firstly,the in situ LAI data were acquired through six field experiment in 2013,and the regression models between in situ LAI and vegetation indices were established;then,six high resolution LAI maps from corresponding HJ\|1A/B CCD were obtained;finally,comparisons between MODIS/LAI products and high resolution LAI images were conducted to reflect the accuracy of MODIS/LAI products.Results showed that MODIS/LAI product can correctly indicate the grassland growth and phenology.However,because of the uncertainty of the input data,uncertainty still exists between MODIS/LAI product and ground situation with a mean absolute deviation of 0.59 m2/m2.Results also showed that MODIS overestimates in situ LAI throughout the grass growing season with an average relative error of 40%.At the begin and end of the growing season,a bigger overestimation occurred of MODIS/LAI products because of the surface heterogeneity caused by the soil background,and at the middle of growing season,overestimation decreased with a relative error of 30%.The results are important to understand regional produce accuracy and further improve model predictions of Hulunber meadow steppe.
Chang algorithm and improved Chang algorithm are the simple empirical algorithms of snow depth inversion algorithms of passive microwave remote sensing.In order to evaluate the applicability of the improved Chang algorithm in Northeast China,this paper analyzed and validated improved Chang algorithm.In spatial analysis,this study selected 84 field sampling points and 48 meteorological stations to analyze and validate the improved Chang algorithm.The results showed that when the underlying surface is forest improved Chang algorithm underestimated the snow depth of 3.6 cm,however when the underlying surface is farmland improved Chang algorithm overestimated the snow depth of 1.5cm.In the time series analysis,this study selected snow depth data of four meteorological stations from 15 November 2012 to 28 February 2013 to analyze and validate the improved Chang algorithm,and four meteorological stations are Wuying,Huzhong,Qingan and Bayan respectively.The results showed that when the underlying surface was forest improved Chang algorithm underestimated the snow depth.It underestimated the snow depth of 13.7 cm for Wuying and 8.3 cm for Huzhong.However when the underlying surface was farmland improved Chang algorithm overestimated the snow depth.It overestimated the snow depth of 3.4 cm for Qingan and 0.8 cm for Bayan.The results also showed that when the underlying surface is farmland the accuracy of the improved Chang algorithm is better than that when the underlying surface is forest in spatial analysis and in the time series analysis.Moreover,the snow depth of improved Chang algorithm inversion was increasing and the depth of meteorological stations was constant.The possible cause was that snow grain size was increasing.
The observation experiment is the base of validating the remote sensing products,and the observation value is necessary to ascertain the accuracy of remote sensing products in the pixel scale and enhance the application value.In the paper,based on the theory and method of remote sensing product validation,the framework of observation experiments was proposed in a way of exploring the conception of the two target ranges validating the remote sensing products with the different resolutions.The two target ranges have the characteristics of uniform and non\|uniform obtained through statistical analysis respectively.And as an example of Huailai Remote Sensing Test Station,Chinese Academy of Sciences,the feasibility building two target ranges was analyzed,the current progress of constructing the two target ranges was summarized,and the improvement direction was suggested.And so,the paper could provide a train of thought to validate remote sensing products.
It is well\|known that ocean covers 70% of the globe which play a fundamental role in climate,meteorology,environment,economy.Therefore,there is a need to monitor,understand and predict all the processes that occur in the ocean and its surface.In particular,wind and surface waves are key parameters which affect the marine meteorology,ocean dynamics,marine resources,pollution,economy,coastal environment.To investigate the wind and surface wave of ocean,a wave spectrometer (SWIM:Surface Waves Investigation and Monitoring) and a scatterometer (SCAT) were proposed to measure ocean winds and waves.This paper presents the introduction of these two payloads,SWIM and SCAT,to built a small satellite for the above mentioned purposes.Based on the mission analysis for the design of this kind of active microwave satellite,it demonstrats the key aspects for the system design,engineering feasibility and validation methods which is dedicated for this small satellite with this combination application of SWIM and SCAT instrument to reach the mission requirements of this project.
This paper presents a model of the alteration mineral mapping base on multi\|level separate background.The understructure of this model is the physical mechanism of spectral features and remote sensing technology.The procedure of this model is “analysis of spectral characteristics of mineralization and alteration rock\|\|calculation of apparent reflectance image\|\|‘background’ multi\|level separation\|\|alteration information enhancement\|\|alteration mineral mapping”.This model can quickly,accurately extract alteration information.This paper chooses Mount Hadu Hula in Yanqi County,Xinjiang,as study area and the new Landsat8\|OLI multi spectral data is as the data source.Then the model is used to extract jarosite altered rocks information in this study area and it has extracted more than 80 jarosite altered rocks dew points.In these points,there are two concentrated region of jarosite altered rocks,which provides important clues of prospecting work for the study area.
This study analyzed the spatial distribution of vegetation NPP in Qilian Mountains from 2000 to 2010 based on MODIS NPP.The study showed that the vegetation NPP had a decreasing trend from east to west in Qilian Mountains due to effects from precipitation.Average vegetation NPP values of Qilian Mountains were only 121.95 g C/(m2·a).NPP values of different vegetation types could be sorted in a descending trend:evergreen broadleaved forests,plain grasslands,evergreen coniferous forests,meadows,farmlands,alpine and sub\|alpine grasslands,desert grasslands,deciduous coniferous forests.Overall,the average NPP of Qilian Mountains had been increasing during the past 11 years.The NPP of most of the area (about 47.30%) in Qilian Mountains showed an increasing trend,while less than 2% area showed a decreasing trend.Precipitation is a major factor of vegetation NPP change in Qilian Mountains.The effects of temperature on the NPP is not obvious,and irrational human activities may be the important reasons of Vegetation NPP reduction in part of Qilian Mountains.
Remote sensing is an important method to monitor sea fog.In order to get the sea fog products with the time resolution higher than those from polar orbit satellites and the space coverage larger than those from ground based measurements,the data from Japanese MTSAT geostationary satellite were used to retrieve the sea fog in Zhejiang province and its adjacent sea area.Normalized Differential Fog Index(NDFI) and classified of the solar elevation angle were applied to extract the sea fog from cloud and other feature types since 2008,and the sea fog results of remote sensing were compared to the manual observation about nearly five years and visibility automatic observation hourly in the spring of 2013.The results showed that the overall accuracy of remote sensing retrieved were 68.4% and 69.6%,and the spatial distribution was similar as the observation.It is feasible to monitor the sea fog using the geostationary meteorological satellite at real time,and has a good application prospect.Hit percent of the different time indicated that,the sea fog method by remote sensing was more suitable for winter and spring as well as daytime.
