The polarized Bidirectional Reflectance Distribution Function (pBRDF) for the rough ocean surface was deduced and was coupled with the Successive Order Scattering Vector Radiative Transfer model(SOSVRT).The SOSVRT-pBRDF model has been tested against with other existing models and demonstrated its accuracy.Some study results were computed by the coupled model are given.It is shown that,the pBRDF model could compute the polarized reflectance characteristics of the rough ocean surface well;the degree of polarization on the ocean surface is high in specular reflection direction and affects the polarization intensity at the top of the atmosphere significantly.By adding the polarization boundary condtion,the SOSVRT-pBRDF model may compute the radiation spectrum in the ocean\|atmosphere system more accurately,it is of very important for the remote sensing in the earth-atmosphere system.
The missed and false alarmed precipitation is the main drawback of high-resolution satellite precipitation estimation.In this study,the missed and false alarmed of CMORPH(Climate Prediction Center morphing) with spatial resolution of 0.1°×0.1°and temporal resolution of an hour were calculated by comparing with ground observations.Then the false alarmed ratio was calculated and compared with precipitation ratio and precipitation areas.The results show that:① There is a huge number of false alarmed data in CMORPH which is much more than the missed data,and leads to the precipitation areas of CMORPH enlarged compared with ground observations;②The relationship between the false alarmed area and the total precipitation area is very strong in June,July and August,and the correlation coefficients are 0.9133、0.9474 and 0.9482 respectively.So the area of false alarmed can be estimated by the total precipitation area of CMORPH;③ When the rainfall is below 5mm,the false alarming ratio is related to the rainfall value.As the rainfall value increasing,the false alarming ratio tends to decline;④ The spatial distribution of the false alarming ratio present obvious regional characteristic.The false alarming ratio in southeast,northeast is the lowest,while in the west of china is the highest.All of the conclusion above can be used as the scientific basis in building correction model of CMORPH.
The daily evapotranspiration(ET) of Central Asia and Xinjiang has been estimated from the national geostationary meteorological satellite data and ground observation for 30 years.The Surface Energy Balance System(SEBS) model can be applied in the ET estimation.Results show that the ET during the 2005 growing season in Xinjiang and Central Asia were 2 168.68×108 and 9 741.03×108 m3 respectively.The total ET for May\|September of 1980,1990 and 2005 were 8 960.64×108,9 134.37×108 and 9 085×108 m3 respectively.Through analyzing the precipitation of study area and water balance in Xinjiang,model simulation of ET is acceptable.The actual ET describes the water variation quantitatively with the participation of evaporation process and reveals the variation rules of ET process in arid areas.
Grasslands not only support animal production,but also play an important role as an ecological safety barrier and they maintain herders’ livelihood as well as inherited culture.However,a series of ecological problems are becoming more and more serious because of increasing grassland degradation.Monitoring the ecological status of grassland in real-time and accurately is important.The spectral characteristics of the common grasses in central and eastern part of the Three River Source were analyzed by measuring spectral data of various plants.The spectral reflectance curve of grassland vegetation was processed by methods of first\|order differential,continuum removing and normalized differential ratio to extract the spectral characteristics of typical grassland vegetation.Dominant species of Kobresia tibetica and Kobresia parva were identified with high precision through spectral analysis,which provides the strong basis of applying hyperspectral remote sensing in grassland monitoring.
For a flat ocean surface,the emissivity can be accurately calculated from the Fresnel formula for a given seawater complex dielectric constant and a local incident angle.Consequently,the microwave complex dielectric constant of sea water is an important input parameter for sea surface emissivity.As far,many classic seawater complex dielectric constant models have been released,which are widely applied in spaceborne microwave ocean remote sensing field.Firstly,some classic models and their main characteristics and differences were reviewed.Secondly,using the satellite data and collocated atmospheric and oceanic products,the simulation accuracy of the five classic models was analyzed by comparing calculation and measurement of WindSat brightness temperatures.The results show that the microwave seawater complex dielectric constant model in Fast Emissivity Model-5 has relatively higher calculation accuracy and more faster running speed.
To provide a reference for parameter inversion and the model optimization,the Extended Fourier Amplitude Sensitivity Test (EFAST)was used to global sensitivity analysis of the PROSAIL model parameters.This paper quantified the sensitivity,and screened out the parameter which had the greatest impact to the model results.The results show that:in the red band,the parameters Cab,Ns,Hspot and LAI,are more than 0.1 of total sensitivity index.The total sensitivity index of Cab is high to 0.489,which is the primary impact parameter to the PROSAIL model simulation results.In the near\|infrared band,the parameters LAI,Cm,ALA,Ns and Hspot are more than 0.1 of total sensitivity index.The sensitivity index of LAI is high to 0.512,which is the greatest impact parameter to the PROSAIL model simulation results.
The remote sensing inversion models for the aboveground biomass of the Honghe Wetland were established by TM data and the field investigation of aboveground biomass around the Honghe Wetland. The spatial distribution of aboveground biomass of the Honghe Wetland and the relationship between the biomass and the DEM were analyzed.The reasons for within different biomasses,the correlation between biomass and elevation differences were analyzed too.The result showed:The multiple regressions had a higher fitting precision than other regressions,and R2 was 0.813;The total biomass of Honghe Wetland was 2.4856×108g and the average was 934.7105 g/m2 in 2007;The biomass mainly concentrated between 600 and 1 200 g/m2.The composite analysis of the biomass and DEM showed that the biomass was between 0 and 600 g/m2 which had a good correlation,the correlation coefficient was 0.79839.The biomass was between 600 and 1 200 g/m2 and above 1 200 g/m2 which showed a poor correlation.
Tomographic technology has unique advantage in forest vertical structure information extraction,and it has been applied for forest vertical structure parameters extraction,hybrid environment separation and so on.Polarization Coherence Tomography(PCT)is a tomography technique relying on a priori information using Fourier-Legendre series to reconstruct target vertical structure based on interferometric coherence.Single or dual\|baseline PCT technique still can be used for forest vertical structure profile extraction.Some scholars questioned this method as it has too little baselines to get vertical structure information.The objective of our research is to analyze the principle,approach and present problems of forest vertical structure profile extraction using single baseline PCT technology.Specific method is as follows:first,vertical distribution of mean backscattering power in different forest scenarios are collected using forest simulation data;second,forest vertical structure profile(vertical distribution of relative reflectivity)is produced by PCT;third,similarities and differences between vertical distribution of mean backscattering power and forest vertical structure profile are analyzed;fourth,physical significance of forest vertical structure profile will be concluded.The results show that:①Forest vertical structure profile reconstructed by PCT approximately express the vertical distribution outline of mean backscattering power.②Forest vertical structure profile and mean backscattering power vertical distribution of different polarization are different.③The vertical distribution of mean backscattering power contains more details.
With the capabilities to observe all weather,all time and penetrate the crops,Synthetic Aperture Radar (SAR) has become a major tool to retrieve parameters of crop.This paper develops a multi-scattering model of maize based on Matrix Doubling algorithm,validates the model by polarimetric radar of RadarSat\|2 data.The physical parameters dataset of the ground and maize are collected as input parameters to simulate the characteristics of maize in different phenophase by this model.Different structure characteristics of maize result in different backscattering characteristics.At low incidence angle,C-band frequency,a strong relationship between soil moisture and the depolarization ratio in dB ( VV-HV) is presented in the seeding stage of maize.At high incidence angle,C-band frequency,the sensitivity between leaf water content and HH-HV is also observed in the ripening stage of maize.
A simple linear mixing model of heterogeneous soil\|vegetation system was applied to bi\|angular observations made by AATSR to retrieve soil and foliage component temperatures over heterogeneous land surface during the period of growing season records of component temperatures,from May 30th to August 24th in 2008 in Zhangye Oasis.Comparison the retrieved component temperatures from the AATSR data with and without proper geo-registration between nadir and forward view images indicates that the registration between the nadir view and the forward view is an important process for obtaining reliable retrieval of soil and vegetation component temperatures.The retrieved component temperatures by AATSR were compared with the results retrieved from airborne measurements made by WIDAS and ground measurements of long-wave radiation for validation,and the validation results indicate that both vegetation and soil temperatures retrieved by AATSR can reflect the spatial and temporal trend of component temperatures correctly.
In order to repair the Digital elevation model(DEM) derived from airborne Interferometric Synthetic Aperture Radar(InSAR),many studies have to identify the error patches which are caused by water or shadow.A new method is proposed to deal with this problem automatically.Firstly,gross error points are detected in DEM.Then let the points grow by using regional growth algorithm in SAR image.After that we can extract the water body and shadow areas.Secondly,we identify the water and shadow areas by constructing constraint conditions based on the height difference along range direction and radar depression angle.Airborne high\|resolution InSAR data are used to validate our method.The result shows the identify ratio of water and shadow areas can be more than 92%.Water and shadow of landform can reach satisfying effect,but small shadow result from trees is easily misidentified due to the inherent noise in InSAR DEM and other factors.
This paper aims to study the relationship between urban expansion and thermal environment changes in Hangzhou,China.The study is of great importance in the field of city planning which helps the cities sustainable development.By analyzing multitemporal Landsat satellite images of the city,the information of urban expansion,Land Surface Temperature (LST),built\|up land,and vegetation in 1989,2000 and 2010 were collected for quantitative analysis.The study results show that during the past 21 years,Hangzhous urban heat island area expanded greatly in accordance with the citys built\|up area expansion.Nevertheless,statistics shows that the percentage of the citys high temperature area has decreased gradually,the Urban\|Heat\|Island Ratio Index (URI) declined from 0.78 to 0.71.This suggests that the urban heat island effect has mitigated due to the reduction of built\|up land density.Quantitative analysis proves that the contribution of built\|up land to LST rise is greater than that of vegetation to LST decreasing.On the whole,despite some minor mitigation,Hangzhous urban heat island phenomenon was still in a strong status during the 21\|year study period.
As the geological weak zones,faults are always eroded as linear geomorphologies which can be distinguished as lineaments by the color variation in remote sensing images.Lineaments help to understand the spatial distribution of faults,and then estimate the stability of rocks and predicate the underground mineral resources distributions.For any pixel in the remote sensing image,STA (Segment Tracing Algorithm) first determines a searching window centered at it,then finds the most continuous direction from the several directions divided in the searching window.Next STA identifies the pixel whether is a line using a threshold value which can be obtained using geostatistics.Then STA distinguishes whether the line pixel is a valley or a ridge by analyzing the variation along the direction.Those valley line pixels in a certain distance and angle tolerance are connected as a lineament.While those reduplicate line pixels should be deleted.From the application of the STA in the eastern district of Gejiu tin mine,the extracting lineaments coincide well to the measured faults,and fit the trend of measured faults well.But some parameters of STA need to be given by operators,the algorithm still should be improved.
Meng Jihua,Du Xin,Zhang Miao,You Xingzhi,Wu Bingfang
(Institute of Remote Sensing and Digital Earth,Chinese Academy of Sciences,Beijing 100094,China)
Large-area crop condition information is not only helpful in producing proper agricultural policy and making sound decision in grain trade,but also a premise for crop production prediction.Crop bio-physical parameters(biomass)derived from remote sensing are the major indicators for crop group condition evaluation.When these parameters are used in large-area crop condition evaluation,the parameter difference due to phenophase difference and that difference due to crop condition are mixed,which is hard to distinguish.This will induce much uncertainty in the large\|area crop condition monitoring with these bio\|physical parameters.Taking Shandong and Henan provinces as study area,this study estimated winter wheat biomass and phenophase with 250 m MODIS data.Then the relationship between phenophase and biomass at certain phonological stage for winter wheat was studied.This relationship was used to normalize the winter wheat biomass to same phonological stage.The elimination of uncertainty due to phenophase difference in crop condition monitoring was explored.
The processing of urbanization makes land use and vegetation coverage change significantly.The impact of urbanization on vegetation phenology quantitatively causes more and more attention.Based on NDVI time series images from 2001 to 2006 of Beijing-Tianjin-Tangshan region,this paper got the spatial distribution pattern of the vegetation phenology and calculated the metric of vegetation phenology in urban areas and buffer zones of Beijing,Tianjin and Tangshan.Then this study analyzed the trend of urbanization impacts on vegetation phenology metrics.The results showed that from 2001 to 2006,the urbanization of Beijing caused the vegetation in the urban area where had an earlier start of growing season(SOS),later end of growing season(EOS),longer Growing Season Length(GSL) and smaller NDVI amplitude.But the urbanization of Tianjin and Tangshan made the vegetation in the urban area later SOS,earlier EOS,shorter GSL and smaller NDVI amplitude.The result also revealed there is a correlation relationship between the impact of urbanization on vegetation phenology and the type of city expansion.
With the deepening of the research and application of remote sensing,the quality of remote sensing data as well as the uncertainty problem has caused wide attention,especially the combination between remote sensing information and GIS makes the research content and system of the uncertainty of remote sensing,the source of data uncertainty,processing method etc,which has become a hotspot in the field of remote sensing.Based on the spatial resolution of 30 m of TM images and DEM data on September 10,2006,this study automatic extract the scope of the snow by the method of NDSI,then compare with the result of visual interpretation,taking roughness as measurement,analyse land cover and terrain factor which affect its uncertainty.The results show that:The bare land’s spectrum curve and the snow spectrum curve are in the shadow area are very similar,so it’s hard to distinguish bare land from snow in this area,in addition,the vegetation in shadow area are misclassification as snow because of their low reflectivity;Extract the scope of the snow by the method of NDSI in half shady slope,its uncertainty is the minimum,but in sunny slope and half sunny slope the uncertainty is the maximum.The uncertainty of extract the scope of the snow by the method of NDSI is smaller when the slope is bigger.
Understanding of the radiant effect of water content on the satellite microwave remote sensing is very important to the applications such as cloud parameter retrieve and the use of satellite data affected by cloud and precipitation in numerical forecast and so on.By using the Community Radiative Transfer Model(CRTM)with the output from numerical forecast model WRF,the effect and sensitivity of water content,including cloud water,rain water,ice,snow and graupel,on the satellite microwave remote sensing are studied.The characteristics of the simulation bias of satellite microwave remote sensing are analyzed.Also,the investigation of the sensitivity degree of the satellite microwave observation to the magnitude,particle radius and the vertical profile of water content is presented.The result shows that the satellite microwave remote sensing is greatly affected by the water content in several channels.The cloud and rain water mainly increase the brightness temperature.The effect of rain water is larger than that of cloud water.The largest temperature increment in AMSUA channel 2 brought by cloud and rain water is 8 K and 17 K,respectively.The ice,snow and graupel decrease the brightness temperature.The effect of graupel is the largest.The largest decrement in AMSUB channel 2 is -18 K.Ice has limited contribution to the temperature descreasing.The sensitivity of satellite microwave remote sensing to the magnitude of water content corresponds well with the radiant effect of water content.The satellite microwave observation is not sensitive to the effective radius of cloud and ice particle and the sensitivity of satellite microwave remote sensing is large to the effective radius of rain,snow and graupel particles.The sensitivity degree change has become complex with the frequency.In addition,the sensitivity of satellite microwave remote sensing to the vertical profile of water content is presented by the change of channel affected.AMSUB channels 3~5 are the most sensitive to the vertical profile of water content.
Land cover production is a hot research topic in remote sensing.Because of the huge spatial scales,global land cover products have longer production cycles,which lead to a large time scale gap in two adjacent products.Correspondingly,rapid generation and update technology is required to improve the time resolution of the global land cover products.Taking the method of mean significant statistical test to realize the rapid renewal of land cover type product using data of historical MODIS Land Cover Type products and Surface Reflectance products.The study data of Ningxia autonomous region in one MODIS tile whose tile identifier is h26v05 has passed visual inspection and accuracy assessment;the overall accuracy is 90.03%.By analyzing and demonstrating trends and reasons for the changes of water bodies,mixed forests,grasslands,urban areas and barren or sparsely vegetated.The result shows that taking the method of mean significant statistical test to realize the automatic rapid renewal of land cover type product is feasible.
Support vector machine (SVM) shows great performance in many classification algorithms,with the merits of high precision,generalization ability and high\|dimensional data processing ability.Therefore,It has been widely used in remote sensing classifications.SVM classification,combining with texture features,has been the research focus of remote sensing applications.Since texture features can overcome the phenomena of “the same thing with different spectrum and different things with the same spectrum” in remote sensing images.Multi\|scale texture features were used to distinguish features in different scales space,which were difficult to distinguish in single scale.The study was mainly focus on texture features selection and classification with stratified samples.Firstly,using ALOS pan and multispectral remote sensing images,8 kinds of texture features in different scales and directions were extracted,based on the Gray Level Co\|occurrence Matrix;Secondly,with the help of the characteristic curve of land types,texture features of mean,homogeneity and dissimilarity in multi\|scale were selected,based on the spectral classification results.Finally,the sample stratification method was used in the SVM classification of land cover,which combined spectral with these three kinds of multi\|scale texture features in different directions.The sample stratification was implemented as follows:firstly,select the initial training samples to classify the land types;secondly,select new samples from the misclassified plots of the initial classification results,then put these new samples and the initial training samples together constituted the second-level training samples.If the second\|level training samples met the needs of classification,they were the final training samples.If not,using the same method to select higher-level training samples.The experimental results show that the overall accuracy and Kappa coefficient of the SVM classification,which combined spectral with multi-scale texture features,using the third\|level training samples,which are improved by 8.11% and 0.11 respectively,compared with that based only on spectral features,using the initial level training samples.Texture features made overall classification accuracy improved 4.13%,especially for those land cover types whose texture features is strong;while stratified sample contributed to 3.98%,and the accuracy of single classification has different degrees of improvement.Results illustrates the method used in this paper is effective.
The image analysis software,eCognition has been used more than ten years.This software adopts an object\|oriented method to improve classification accuracy of high-resolution data.One of the key technologies is the Fractal Net Evolution Approach (FNEA) to be used for multi\|scale segmentation.In this paper,we analyzed some shortages of the FNEA.For example,the FNEA only adopts the spectral and shape features.This paper introduced the multi\|scale edge feature into the multi-scale segmentation process to improve the segmentation results,which have smooth edges and are consistent with the targets as soon as possible in the earth s surface.In theory,this improvement can reduce over-segmentation and under-segmentation and improve results of the FNEA.We conduct experiments to segment high-resolution imagery by fusing the edge and spectral features.The result proves that our method can advance the segmentation results and reduce the over\|segmentation and unde-segmentation.
A supervised method combining Middle Level Feature (MLF) with Support Vector Machine (SVM) was proposed for land\|use classification of full polarization Synthetic Aperture Radar (SAR) image.Supervised method is chosen to directly distinguish the actual land-use categories.The MLF that is used for striding over the semantic gap between the low-level polarization scattering characteristics and the high\|level semantics of land-use categories is got from the result of classic unsupervised classification methods for full polarization SAR image.The MLF of a pixel is calculated by counting the frequency of the middle-components in a feature supporting region centered on the pixel.Here the middle-components refer to the unsupervised clustering categories obtained from the low\|level polarization characteristics.The proposed method is tested on a Radarsat-2 full polarization data covering WUHAN area and good classification performance and potential of further improvement are shown.The comparison between the supervised classification method combining SVM and the classic polarization characteristics is given.The proposed method and different methods for getting the middle-components and feature supporting windows with different size are studied on their impact on the final classification performance.
The decision tree algorithm is a non\|parametric,nonlinear supervised classification method.This study takes Landsat TM images of August 1,2010 as the basic remote sensing information source,choosing the intersection of Bairin Right County,Linxi County,Hexigten County and Wengniute County,the center region of Chifeng City,Inner Mongolia,China as the study area and by repeatedly modified to improve the training data set,then select twenty\|one characteristic variables combinations whicht consist of five different characteristic variables combinations,and they include six original bands and the NDVI based on them,principal components (PC1,PC2 and PC3),eight texture features (Mean,Variance,Homogeneity,Contrast,Dissimilarity,Entropy and the Second Moment and the Correlation) and three topographical features (DEM,Slope and Aspect),then this study utilizes the typical decision tree algorithm C5.0 to classify them,and the final results were compared with that of maximum likelihood classification results.The result shows that the decision tree classification result is better than maximum likelihood classification result ,especially,when the combinations of characteristic variables are appropriate,which is able to effectively use the related auxiliary information,then their final classification results are more satisfy the users demand.
High-resolution remote sensing images and GIS ancillary datasets such as parcels are combined to perform land use/cover change mapping in the urban\|built areas.NAIP datasets,a novel high\|resolution aerial remote sensing images in The National Agriculture Imagery Program,are used in these works.After trial and error image segmentation pursuing for good processing results,an objcet\|oriented image classification framework based on decision tree rules,combined with the cadastral datasets as a secondary data,was built to improve high\|resolution remote sensing image classification on the high\|density urban areas.The classification accuracy of object\|oriented remote sensing image classification combined with geographic auxiliary data are better than only using the remote sensing images.Experiments studies showed that roads,building and others are excellently extracted.Comparing with the conventional object\|oriented classification,the overall classification accuracy of this novel methodology increased from 84.08% to 89.79%.Such a result reveals that auxiliary data can effectively improve the accuracy of high\|resolution remote sensing image classification.
Remote sensing technology has been applied in petroleum prospecting for many years.With the development of remote sensing technology,hyperspectral remote sensing has been introduced to petroleum prospecting and a lot of progress have been made in recent years.The principle and workflow of petroleum prospecting by using hyperspectral remote sensing are briefly described.Subsequently,the microseepage of hydrocarbon and its spectral expression on the surface,i.e.the reflectance spectral feature corresponding to the seeped hydrocarbon,the induced rock and mineral alteration and geobotanical change,and the current problems as well are discussed.The current hyperspectral data used for petroleum prospecting and their feature are summarized.The further research should be focus on some key issues,such as the spectral feature extraction,data processing and analysis,and the possible resolution is proposed.
Long time series of remote sensing data products have been widely applied in various fields,such as monitoring environment system on global or regional scales,analyzing vegetation and land cover dynamics,and extracting information of vegetation phenology.However,due to the effects of cloudy,aerosols,solar angle and bidirectional reflectance etc.,serious residual noise exists in the long time series data sets which hinder their future use.Therefore,reconstruction of long time series data sets is the most important work before application.In this paper,common used data reconstructing methods were reviewed firstly.Then the comparison works of the time\|series reconstructing methods were summarized in detail.The parameter definition was also discussed for common used reconstructing methods.Assessment methods for reconstructing results with their problems were analyzed.A detailed introduction for the softwares used for time series and their development were made.Finally,the research focus and developing direction were proposed for the future work based on a comprehensive analysis of the available literature.
