Satellite visible and infrared observations have their own advantages in precipitation remote sensing,it`s proved that cloud optical thickness and effective radius,retrieved from visible and infrared observations,and have obvious effects in discriminating precipitation.Based on two existing precipitating clouds identification schemes,infrared bright temperature is added and the abilities of discriminating precipitation events from infrared bright temperature,optical thickness and effective radius are explored in this study.Comparisons to PR standard rain products from 2006 to 2007 boreal summers are conducted to verify the IPCTC results,in which three dimidiate-forecast factors are utilized and three existing precipitating-cloud identification schemes are also evaluated.The study shows that,by adding infrared bright temperature information,the capability of the new scheme in identifying precipitation has significantly improved.After adding infrared bright temperature information,weakly rainy areas can be identified more effectively,and both the missing and false alarm areas are spatially very smaller and narrower than the existing methods,which gives better performance both in land and oceanic areas.Especially,in oceanic areas,about 79% of rainy pixels observed by PR are correctly determined,and the probability of false detection remains just 8%,and the CSI,ETS and HSS factor values increase significantly.The IPCTC,which is proposed on these bases,is available for common multi-channel visible and infrared sensors,and can be applied in real-time satellite precipitation monitoring.
Land Surface Temperature(LST) is required for a variety of climatic,hydrologic,ecological and biogeochemical studies.With the launch of more and more high temporal resolution of geostationary metrological satellites,it is urgent to estimate LST data from these satellites.In order to meet this demand,this study aims to estimate LST based on diurnal evolution cycle of LST and split-window method.First,according to the split-window algorithm and diurnal temperature cycle model,the method of inversing LST is induced;then,the new method was validated by a diurnal brightness temperatures datasets corresponding to FY-2D,which was simulated by MODTRAN model with several input parameters under clear sky,including near surface air temperature,atmospheric water,surface temperature and emissivity,and viewing angles;Last,LST in XinJiang region are estimated using the proposed method with FY-2D data collected on september,30,2010,and are compared with the products from MODIS.The results show that RMSE of LST using proposed method reach 1.5 K based on simulated data;the trends of LST from FY-2D and MODIS are similar and the correlation coefficients reach 0.5 with the RMSE of 4.4 K.However the method is tested only in west-north region that is drier and the change of diurnal atmospheric water is less.Furthermore,the method is only fitted with the less cloudy or no cloudy condition.
For the applied survey and map with airborne InSAR,to solve the problem of height’s difference in the overlap area of multiple InSAR strips,unite calibration method using connect points in overlap area of multiple InSAR strips is proposed.The proposed method can reduce the needed GCPs’ number,solve the problem of height’s difference in multiple InSAR strips,and broke the bottleneck of using InSAR technique to survey and map in large area.The method of using GCPs to estimate the linearity trend of Doppler centroid frequency along azimuth direction in long strip InSAR data is also proposed.Some airborne InSAR data derived by Institute of Electronics,Chinese Academy of Sciences,were used to do calibration experiments with the proposed processor.Their results demonstrated that is efficient.
Heterogeneities of snow properties include the variation of snow depth and the snow cover fraction in the horizontal direction and the variation of grain size and density within a snowpack profile in the vertical direction.These heterogeneities produce great uncertainties to the snow depth or Snow Water Equivalent (SWE) retrieval from passive microwave data,as well as the validation of the results.In this study,we illustrate and discuss these problems based on the field observations of snow microwave radiation characteristics,the comparison of different snow remote sensing products and the modeling of snowpack radiation transfer characteristics,which will provide reference for the algorithm development of microwave remote sensing of snow and result analysis and validation.
Ocean is one of the significant parts of the earth,which provides abundant resources for human beings.However,pollutions take place at different levels every year,in which oil pollution plays an important role.Oil pollution is mainly caused by oil leak from streamers and well platform,as well as by explosion of submarine pipe lines.Each accident may cost a direct economic losses ranging from millions to billions.Thus it is meaningful to develop oil spill monitoring.This paper uses a SAR data of Envisat monitoring marine oil spill,introduces and analyzes general steps and implementation of SAR images oil spill detection,chooses mean,variance,equivalent looks and speckle reduction capability as filtering effect evaluation indicators.Through these indicators,this study evaluates seven filters as follows,Lee,Enhanced Lee,Frost,Enhanced Frost,Gamma,Local sigma and Bit error,to compare,analyze and then draw a conclusion that enhanced Lee filter is the most appropriate for the images in the research.Then does target detection to images with single threshold segmentation,maximum entropy segmentation and unsupervised classification,thus divides image area to foreground region and background region roughly,at the same time,classifies by texture features of images.In the processes of texture features selection,this paper manually selects some oil spill areas and non-oil spill areas as regions of interest,and statistics of general texture features including mean,variance,homogeneity,contrast,dissimilarity,second moment and correlation of texture features of SAR images in regions of interest.The result indicates that mean,homogeneity,dissimilarity and second moment have the maximum differences between oil film and non-oil film,and these four texture feature parameters are selected and combined with results of targets detection and original images classification by BP neural network.The SAR data of Cosmo-Skymed with one meter resolution are used as basic data to verify the accuracy.Verification shows that the highest accuracy combined with maximum entropy is 86%.The second accuracy combined with single threshold is 81%~84%,the Isodata accuracy combined with unsupervised classification is 75%,the accuracy just combined with original images is 70%,the lowest accuracy without any targets detection is 57%.So targets detection plays a significant role in marine oil spill.This paper finally defines maximum entropy method as the best marine oil spill detecting method,and puts forward the direction of marine oil spill detection with SAR images in the end.
The development of theory and technology about Polarimetric SAR Interferometry,provided a broad space in research and application,just one of those applications is extract DEM in vegetation coverage area.In this paper,making use of Singular Value Decomposition,Numerical Radius and Phase Diversity ways to extract DEM with processing of filtering,unwrapped phase,base-line estimation and so on.At the end,compared optimization’ results with the single polarimetric interferometry such as HH,HV and VV. It has been observed that optimization ways can reduce the interferometric noise and the phase unwrapping residuals,and improve the precision of the DEM.Meanwhile,Numerical Radius way has a better result than Singular Value Decomposition and Phase Diversity ways;Interferogram with integrated Contoured Median and Goldstein two-step filter method can further improve and optimize the quality of the interferogram,reduce residual almost,improve the precision of the DEM,but if two-step filter window setting big will affect the area of non-water-covers while can improve the water-covers area.
The similarity of spectral feature in visible/near-infrared band between cloud and snow has been an important influence which degrades the recognition accuracy of cloud and snow,especially the panchromatic images.In this paper,a novel and feasible method was presented to automatically identify cloud and snow from panchromatic images.The method made full use of two different analytical techniques:the spectrum threshold segmentation and the texture analysis.These two approaches discriminated the image from two different aspects.At first,the cloud or snow was distinguished from the background utilizing the difference of spectral feature,so the proportion of cloud or snow in the image was got.And then the samples’ fractal dimension which could reflect the texture feature of cloud and snow from an image were calculated to get the distribution of the fractal dimension values.At last,by comparing the proportion with the distribution,the automatic identification of cloud and snow was realized.The experimental results of the actual panchromatic images by Beijing-1 indicate the feasibility and accuracy of the method.The method could be also applied for other high-resolution panchromatic images because of the universality of the texture feature.
For high-resolution remote images with tremendous feature,a new method of image registration is proposed by combining contourlet transformation with SIFT.Firstly,the discrete contourlet transformation decomposes the original image into a low-frequency sub-band and several high-frequency sub-band.For high-frequency sub-bands,image edge feature points are extracted by an appropriate threshold.For the low-frequency sub-band,image SIFT feature are extracted by SIFT.Then,coarse control point pairs are searched out with normalized corss-correlation matching and SIFT matching before accurate control point pairs are hunted by RANSAC.The experimental results show that,as for high-resolution remote sensing image registration,this method can extract more acculately control point pairs,get higher computing efficiency and matching ratio than these methods based on NSCT or SIFT.
The H/a-Wishart classifier is a kind of common and reliable classification methods for the polarimetric SAR(PolSAR) image,of which the classification precision still needs to be improved.A new classification algorithm for the PolSAR image is introduced in this paper,which is based on the genetic algorithm.This method decomposes polarimetric SAR data to extract features,then initials classification by the H/α plane and finally does reclassification by iterating the genetic algorithm.The disadvantages of genetic algorithm are‘premature’ and slow convergence.This paper improves the mutation operator of genetic algorithm by combining the H/α plane so that the algorithm can make use of the polarimetric scattering mechanism to decrease the searching field and improve the efficiency.Adopting the data of the NASA-JPL Laboratory and high resolution PolSAR data of The 38th Research Institute of China Electronic Technology Group to research,the results show that the refined genetic algorithm can improve the accuracy of classification and get better results than both the H/a-Wishart classifier and standard genetic algorithm.
The possibility of using HJ-1A/B CCD data for regional burned area mapping with a two-phase algorithm was investigated.Four kinds of vegetation indices were tested for the seperability between burned area and other land types.The result showed that the Burned Area Index (BAI) and Global Environment Monitoring Index (GEMI),based on the reflectance of RED and NIR (Near Infrared) bands of HJ-1A/B CCD,presentel good discrimination capability.Rigid multi-thresholds based on BAI and GEMI were used to discriminate the more severe burned area to minimize the commission error in the first phase.And the “seed” pixels extracted from the first phase were used to generate in the second phase.Loose thresholds were applied as the generation criterion in the second phase to obtain the burned region boundary,and this phase will minimize the omission error.The accuracy assessment showed the burned area commission error was 5.5%,and the omission error was 12.7%.The results underline the feasibility of the two-phase algorithm for regional burned area mapping by HJ-1A/B CCD data.
According to the Shadow characteristic in high resolution remote sensing image,a new Shadow Index (Shadow Index,SI) on Shadow detection is proposed based on principal component transformation and HIS model.Polidorio algorithm,NDUI and SI were used in 2 test areas,and the results showed that the SI can effectively distinguish shadow region with water and partial blue objects.The histogram of three algorithm shows that SI histogram shaped two wave peaks and a wave trough,which is helpful for the histogram threshold value method that divided the area into the shadow area and none shadow area.The experimental results indicate that the new shadow index is simple and effective.
In remote sensing image process,geometric correction is an important component that can reduce or eliminate differences between remote sensing image and real geographic shape.But the effect of conventional technique for UAV image geometric correction is not satisfactory,it improves the locally-weighted-fitting technique by using control circle to regulate its local corrected region and adding auxiliary control point to the cover area.By contrast geometric correction analysis experiment,it shows that the locally-weighted-fitting technique improved may be more adequate and reasonable use of local information,has stronger adaptability for the distribution of control points,and can effectively improve the accuracy of the image geometric correction.
Satellite-derived NDVI time series are often contaminated by negative atmospheric conditions and sunsensor-surface viewing geometries.The reconstruction of high quality NDVI time-series is crucial to the detection of long-term vegetation cover changes and the remote sensing of vegetation phenology.In this paper,MOD13Q1 time-series data covered in Yunnan province were employed to address the performance effectiveness of time-series data reconstruction methods (e.g.linear interpolation,Savitzky-Golay filtering,asymmetric Gaussian and double logistic function-fitting) through integrating with different quality setting (e.g.UI5,UI5-CSS,UI3,UI3-CSS).The results show that seasonal and regional variations in the number and the maximum gap length of invalid pixels of time-series data were mainly controlled by local climate.A comparison of four selected methods revealed that the superiority of the robustness and fitting capability of asymmetric Gaussian and double logistic function-fitting methods over the other fitting techniques.The maximum gap length of invalid pixels in time-series data is an important data quality indicator reflecting the feasibility for meaningful reconstruction.Concentrated clouds and precipitation in the rainy season is a crucial factor of influencing the fitting accuracy of the reconstructed time-series data in some parts of the study area.The reconstructed NDVI time-series data show that the NDVI values are higher in the rainy season than those in the dry season,higher in the western than those in the eastern,higher in the southern than those in the northern, and higher in the river valley than those in the uplands in the study area.
k-NN classifier has been widely used in remote sensing image classification due to its simple concept and easy implementation.However,it may produce a large amount of “salt and pepper” noise and wrong classification results in high resolution remote sensing classification because of its rich texture information.According to the first law of geography,this paper attempts to present a new geographically weighted k-NN classification method (GWk-NN) to solve these problems by incorporating geographical statistical models into the traditional k-NN classifier.First of all,the spatial distribution characteristics of the training samples of each land cover class have been calculated through conditional probability function;Secondly,two kinds of geographical statistical models (exponential and Gaussian model) are fitted and the suitable weighting model for each land cover is selected by the least residual error.Finally,a subregion of Nanjing city (SPOT5,2.5 m spatial resolution) is taken as an example to illustrate the validation of GWk-NN method.By comparing the classification results of GW k-NN and k-NN,the study finds that the new GWk-NN classifier can reduce or even eliminate the “salt and pepper noise” and eventually improve the classification accuracy by making use of spatial and spectral signature of the remote sensing images.
Ocean surface mineral oil and biogenic oil are both presented as dark features in synthetic aperture radar (SAR) image,so it is difficult to discriminate between mineral oil and biogenic oil by the single polarimetric SAR image.This paper analyzes the polarization ratio of ocean surface backscattering coefficient of the mineral oil films and biogenic oil films,and then proposes a method based on cross-polarization ratio of multi-polarization SAR images to distinguish mineral oil with biogenic oil.At last,SIR-C multi-polarimetric SAR data are used to prove the effectiveness of the proposed algorithm.
Clouds have significant impacts on the use of the facilities for optical,infrared and microwave observations and communications in the fields of meteorology,topography,and land resource survey.In order to analyze the cloud occurrence frequency along the line of sight over China and its neighborhood,the calculation method based on active measurements is developed.Then,the CloudSat/CALIPSO observations from June 2006 to March 2009 are used and analyzed.The results show that the cloud occurrence frequency is strongly dependent on the view angle,altitude and season.Furthermore,it reveals that the cloud occurrence frequency along the line-of-sight is independent of the azimuth.In the scope of the downward observations,there is a minimum cloud occurrence Frequency value at a particular angle.Additionally,the cloud occurrence Frequency distributions at different heights are also analyzed.This paper shows that the cloud occurrence significantly differs over different areas.Especially,it strongly depends on the latitude in the upper troposphere,but there are some maximums over tropical regions.
There are close relationships between locust occurrence and their habitat.Therefore,the key of efficiently control of locust plague carrys out locust habitat monitoring.In this study,Bohai development zone in Hebei province was selected as the study area and multi-temporal data from the HJ-1 satellites were chosen to monitor the locust habitat.The locust habitat were monitored by the use of classification based on four different schemes of images band combination and two kinds of classifiers,including the maximum likelihood classifier and RuleGen Decision Tree classifier.The results show that there is no obvious difference between the two kinds of classifiers in terms of the overall accuracy of locust habitat classification.Precisely,the execution efficiency of RuleGen decision tree becomes low when the different category data exist.On the other hand,the overall accuracy of locust habitat classification with the single temporal combination schemes dated on May 20 is only 76.43%,Kappa coefficient of 0.7396,but after adding NDVI time-series information,the overall accuracy can reach 93.93%,Kappa coefficient of 0.9323.Therefore,taking advantage of multi-temporal data,which have similar spectral characteristics easy to classify,the impact of mixed pixel can be reduced largely and the classification accuracy of locust habitat can further be improved.
The study area is located at the north of Gaize county,Ali region,belongs to Ban Nu metallogenic belt.Natural and geographical environment in this area is very poor,and exists low level of geological research,slow progress in geological survey and mineral resource survey.Seen from the mine have been found,the main minerals of study area is placer gold,rock gold and copper and gold mine,but also distributs salt lakes and coal mines,this area boasts favorable mineralization background and great ore mineral potential.Therefore,prospecting in this area via remote sensing is an economical,convenient and effective means.This paper is supported by the ENVI software,according to the spectral characteristics of Aster multi spectral image data and alteration mineral assemblages,and make use of the methods of ratio and principal component analysis to draw alteration information of metal mineralization related to minerals and mineral assemblages from Aster remote sensing data.Geological data are integrated with the results of remote sensing image processing to analyze mining factors to divide prospective districts and delineate prospecting targets.
Ximen Island Marine Special Protected Area is the first national marine special protected area in Zhejiang province.Remote sensing,geographic Information System technologies,and landscape theory are used to analyze landscape dynamics.In this study,SPOT-5 remote sensing image in 2006 and WorldView-1 remote sensing image in 2010 were acquired for the wetland area.The landscapes in the wetland area are classified as seaweed,breeding,mangrove forest and bare mud based on interpretation signs.Areas and landscape indices are calculated .Then changes of every landscape type are analyzed and predicted by Markov model.The results show that Shannon diversity index and evenness index in the wetland of the protected area increase,which means that the landscape types in wetland are becoming diversity.In addition,the predictions show that areas of breeding and bare mud will decrease,and the areas of seaweed and mangrove forest will slowly increase,which reveals that instead and human have gradually recognized the importance of wetland resource.
Vegetation is an important part and plays a very important role in the watershed ecosystem.The sustainable development of regional resources could be assessed by vegetation cover change very well.In order to study the processes of vegetation cover change of Loess Plateau in recent 33 years,this study took Malian river watershed of Loess Plateau as an example.The Plateau would be divided into three functional zones. such as:Mr Tung Chi-plateau agricultural zone,Meridional ridge forest zone and northern Farming and pastoral zone and used Landsat images of four periods in 1977 (MSS),1987 (TM),2000 (ETM+) and 2010 (TM).The methods of supervised classification and unsupervised classification were used in this paper.The transfer matrix was used to analyse the process of vegetation cover change,and calculated the annual rate of change for vegetation cover.The results show that the restoration of Mr Tung Chi-plateau agricultural zone was bigger than that of Meridional ridge forest zone from 1977 to 1987,however,all the vegetation in northern Farming and pastoral zone was of degradation except the restoration of the grassland in large area.During the period 1987~2000,there was a turning piont for the change of vegetation.All the functional zones showed the degradation of vegetation,and with the most typical in the Mr Tung Chi-plateau agricultural zone.From 2000 to 2010,every functional zone was the main vegetation recovery,the restoration of vegetation in Mr Tung Chi-plateau and northern agricultural and pastoral zone is very fast.The Malian river basin shows a trends of degradation in the first period,the further degradation in the second period and the restoration in the last period.
Spatial scale as the basis to the geoscience study,it’s important scientific and practical significance for the coastline where is the interaction of sea-land-air-phase.Based on multi-source remote sensing images such as QuickBird,SPOT5 and Landsat5,Shijiutuo island coastline was extracted,the accuracy evaluation model was constructed by the number of pixels.In the light of fractal theory,scale effect of fractal dimension for coastline remote sensing information was studied by scales translation model.In the context of spatial correction,coastline extraction accuracy error caused by scale effect was analysed for the dynamic characteristics analysis of Shijiutuo island coastline evolution.The results are shown as follows:①the overall fractal dimension of the Shijiutuo island coastline is 1.0111.②Scale effect of coastline varies with the different types of coastline,the fractal dimension of artificial and silt coastline of Shijiutuo island is 1.0123 and 1.0089 respectively,the former is slightly greater than the latter.③After scale translation,the length of coastline increases with the scale,which is called as “marginal effect”.
The spectral backscattering coefficient is one of the important inherent optical properties,and also is an important input parameter of water color quasi-analytical inversion model.It is highly related to the concentration of water constituents in the water.This study reviews the progress in the water backscattering that includes the backscattering properties of water,various phytoplankton and other suspended inorganic particles,at the same time,this paper also discusses the backscattering spectral varition of the different particles.Otherwise,this study describes the theoretical basis and the way how to achieve the backscattering coefficient,systematicaly analyzes the variation characteristics of particulate backscattering coefficient and backscattering ratio,and finds that particle-size distribution and average particle refractive index are the main influencing factors.In order to know if the chlorophyll concentration influence the water backscattering properties,the relationship between them is discussed in the paper.Finally,the future development perspective and research fields are proposed.
Cyanobacteria bloom in relatively smaller and medium sized inland lakes can be detected by HJ-1 satellite more efficiently.However,various types of results will be retrieved from the same remote image due to different atmospheric correction algorithms.Three typical atmospheric correction algorithms are experimented in this study to implement the radiometric correction of multi-temporal remote images,and then acquired global and local features in several regions of interest as well as statistics,in order to compare their effects in the dynamic monitoring of cyanobacteria bloom.Because biomass can be calculated by vegetation indices obtained through multi-spectral arithmetic,the study utilized that method and analyzed both the source of diversities of NDVIs and the impacts on the estimate of cyanobacteria biomass.The results demonstrate quantitative descriptions of cyanobacteria bloom are incongruous because of different correction algorithms.Then the study specifies and characterizes the uncertainties generated by the experimented correction algorithms,and presents recommendations on how to utilize these algorithms more effectively.
