Search

To evaluate the average dielectric property, the effective permittivity (EP) of multi-component media in low frequency region is the first issue. The most classical formula is the Maxwell–Garnett (...

Synthetic aperture radar (SAR) images are inevitably contaminated by speckle noise due to its coherent imaging mechanism. Speckle noise obscures the intrinsic radar cross section (RCS) information ...

All-weather, all-season microwave remote sensing is one of the most powerful technologies for monitoring natural disasters. Multichannel brightness temperature (Tb) measurements from satellite-borne passive microwave remote sensing has played an important role in retrieving quantitative physical information regarding global and regional weather and climate, atmospheric precipitation, land hydrology, and oceanic surface winds. However, in January 2008, during severe weather conditions with heavy snow and frost in the usually warm south of China, the operational algorithm for snow detection using multichannel Tb data failed to detect snow. In this article, based on the simulation of vector radiative transfer of a snowpack model of dense and sticky Mie ice particles, characteristic indexes of scattering and polarization differences, average indexes in the previous year under normal situation, and changes in antecedent indexes are newly defined and analysed. A new detection flowchart is designed to effectively detect the regional snow and frost disaster in 2008 in southern China.

As a first polar-orbit meteorological satellite, FY-3A FY – Feng Yun, which means wind and cloud, in China, the Microwave Temperature Sounder MWTS and Microwave Humidity Sounder MWHS are aboard the satellite for the purpose of retrieving atmospheric temperature and humidity profiles. A novel mixed MIX model, based on the Millimetre Propagation Model MPM and Rosenkranz ROS models, is developed for calculation of atmospheric absorption coefficients in multiple channels of the MWTS and MWHS using the China National Standard Atmosphere GB 1920-80 and US Standard Atmosphere of 1976. Using atmospheric radiative transfer with the in situ measurements of 836 radiosonde datasets including the altitude, pressure, atmospheric temperature and humidity from 16 sites in China and Japan, multiple channel brightness temperatures are simulated and compared with FY-3A MWTS/MWHS observations for data validation. Simulations using respective MPM, ROS and MIX models are also analysed. Furthermore, the statistical D-matrix is constructed from the MWTS/MWHS data and the aforementioned radiosonde measurements, and it is then applied to the retrieval of atmospheric temperature and humidity profiles over the north-western Pacific Ocean during 2008’s Phoenix typhoon.

The time domain finite element method (TDFEM) formulations have been presented using two kinds of temporal basis functions, B-spline and Lagrange interpolation polynomial, and the computational domain is truncated with second-order perfectly matched layer (PML). This is motivated by utilising good interpolation features of these temporal basis functions and better absorption performances of second-order PMLs. Numerical cases demonstrate that the proposed TDFEM schemes achieve more accurate results with about 5% to 7% increase of run time and 1.5% increase of memory usage compared with the TDFEM scheme with Sacks PML. However, the presented TDFEM schemes merely require 1/6th of the total run time for frequency domain FEM.

To recover original full phase values in inherently wrapped output, phase unwrapping has been one of the key techniques in digital image processing. However, to unwrap the phase of a large interferogram with dense residues, a conventional path-following algorithm may produce retrieval errors caused by placing long branch cuts and repetitive connection of residues. In this paper, two unwrapping algorithms, the method of multiple patches connecting opposite residues and the method of centre expansion searching opposite residues, are presented. Some examples of Environmental Satellite (ENVISAT) Advanced Synthetic Aperture Radar (ASAR) datasets demonstrate improvement compared to conventional methods in Interferometric Synthetic Aperture Radar (InSAR) data processing.

Propagation and localisation of cylindrical waves in a two-dimensional (2D) isotropic and homogeneous random medium is studied using the stochastic functional approach. By expanding the random permittivity fluctuation in the form of a Wiener integral equation, and representing the wave fields by a linear combination of outgoing and incoming waves, the scalar Helmholtz equation is solved in the cylindrical coordinates system. An analytical expression of the cylindrical wave is derived and demonstrates the localisation phenomenon, as well as the wavenumber fluctuation in the random medium. Comparing with the waves in non-random medium, the wave transfer equation between plane wave and cylindrical wave in random medium shows an additional exponential factor to indicate the modulation effect owing to the medium randomness in both the amplitude and phase. Numerical simulations are presented to illustrate the functional dependence of the localisation phenomena.

The Dempster-Shafer (D-S) algorithm is improved to fuse different features of multi-temporal Synthetic Aperture Radar (SAR) images to detect changes in urban areas. Firstly, D-S theory is developed by not only considering the certainty of the sources, but also considering the average support of the sources to different subsets in the frame of discernment, in the process of evidence combination. In this way, it can assign conflict information in deferent sources reasonably and present a more reliable combination result. Secondly, amplitude ratio and distance between probability density distribution functions in Pearson graph of different temporal SAR images are extracted to present the change features in different scales. Lastly, the improved D-S algorithm is applied to fuse the two different features to detect change information of SAR images. An example of the Shanghai Lujiazui area using European Remote Sensing 2 (ERS-2) SAR images demonstrates the accuracy of the improved fusion algorithm.

To implement target reconstruction from multi-aspect SAR data, a simple method of first calibrating multi-aspect data is presented. It requires that at least one aspect is calibrated beforehand, and other aspects are then calibrated with respect to this calibrated aspect. A natural object, such as flat bare ground, is usually chosen as a reference target, and is expected to preserve identical scattering for all aspects. Thereafter, the channel imbalance factors are estimated from the distribution of the phase difference and amplitude ratio of co-polarized, hh and vv, echoes of the reference target, and are then used to compensate the whole SAR images. This approach was applied to the calibration of four-aspect airborne Pi-SAR (Polarimetric and Interferometric Synthetic Aperture Radar) data for target reconstruction. Based on the same principle, another potential application is calibration of descending (ascending) data using the calibrated ascending (descending) data.

Phase unwrapping is a key step in extracting digital elevation models (DEMs) from interferometric synthetic aperture radar (InSAR) data. A new two-dimensional (2-D) phase unwrapping algorithm based on ant colony optimization (ACO) is proposed, which is used to configure the shortest path linking all residues in an interferogram. Using an optimization strategy to establish the branch cuts, the unwrapping error can be significantly reduced. Simulated and real InSAR image datasets are studied to evaluate the performance of the algorithm. The results of the simulated datasets show that the errors of the algorithm are lower than some conventional methods, and the results from a real InSAR image dataset demonstrate that the ACO approach has no isolated regions in comparison with some conventional approaches. It indicates that our ACO algorithm is an optional compromise strategy between preferable phase unwrapping precision and time-consuming computation.

Based on the Gaussian scale-space, a Gaussian comparison function is presented for extracting the linearly road features in aerial remote sensing image. Combining the geometric and radiometric features, the curvilinear structures of the roads are extracted based on locally oriented energy in continuous scale-space. Curvilinear features of roads are verified, grouped and extracted by using both topologic and geometric methods. This algorithm is applicable to extracting the road features in different scale such as rural roads or urban highways, and significantly reduces the computation complexity of line tracing. Some discussions on the zero drift of the Gaussian comparison function are also presented.

A hybrid algorithm of the back propagation artificial neural network and genetic algorithm (BP-ANN/GA) was developed to optimize initial weights and to enable fast convergence of the BP-ANN. This algorithm was applied to the classification of urban terrain surfaces with fused data of Landsat ETM+ and ERS-2 SAR. An example of the Shanghai Pudong area using both Landsat ETM+ and ERS-2 SAR image data demonstrates well the classification when applying our fusion algorithm.

The Chinese Moderate Resolution Imaging Spectroradiometer (CMODIS) was loaded on the China's SZ‐3 spacecraft. Using an empirical line method, the CMODIS radiance is converted to the water‐leaving reflectance, and is applied to inversion of the suspended sediment concentrations in the Yangtze River estuary. The concentrations ranging between 0 mg/L and 1000 mg/L are well validated by the field measurement data. This study demonstrates an example for the feasibility of the CMODIS data for concentration retrieval of the suspended sediment.

The polarization index (PI) of special sensor microwave/imager (SSM/I) radiance is derived from a radiative transfer equation and applied to identifying spatial and temporal variations of terrain surface moisture. The index PI, its deviation from the multiyear monthly average , ΔnPI, and a scattering index SI are utilized to monitor flooding of the Huaihe River, China, in July 2003. An anomaly index K, based on statistics of the indices PI and SI, is defined to quantitatively evaluate the flooding situation over vast areas.

To automatically detect and analyze the surface change in the urban area from multi-temporal SAR images, an algorithm of two-threshold expectation maximum (EM) and Markov random field (MRF) is developed. Difference of the SAR images demonstrates variation of backscattering caused by the surface change all over the image pixels. Two thresholds are obtained by the EM iterative process and categorized to three classes: enhanced scattering, reduced scattering and unchanged regimes. Initializing from the EM result, the iterated conditional modes (ICM) algorithm of the MRF is then used to analyze the detection of contexture change in the urban area. As an example, two images of the ERS-2 SAR in 1996 and 2002 over the Shanghai City are studied.

A massive sandstorm enveloped most of northern China during Spring 2002. Monitoring the evolution of sandstorm and desertification has become one of the most serious problems for China's environment. Since 1989, one of the most advanced and operational passive microwave sensors is the Defense Meteorological Satellite Program (DMSP) Special Sensor Microwave Imager (SSM/I) operated at seven channels (19, 37, 85 GHz with vertical and horizontal polarization and 22 GHz with vertical polarization only). In this paper, the sandstorm and desertification indexes, SDI and DI, are derived from the radiative transfer equation, and are employed with multi-channel measurements of the DMSP SSM/I for monitoring the sandstorm and desertification in northern China. Some SSM/I data in 1997 and 2001 are employed. The algorithm of the Getis statistics is developed to categorize the spatial correlation and its evolution during these days. It is demonstrated that the SSM/I indexes, SDI and DI, and its Getis statistics are well...

Observation of polarimetric backscattering from vegetation canopy takes into account multiple scattering from random nonspherical scatterers. Based on the difference of the first- and high-order Mueller matrices solution of the vector radiative transfer (VRT) equation, an iterative method for parameter retrievals of fractional volume per unit area of random scatterers and underlying surface moisture is developed. By using both the numerical simulation and airborne synthetic aperture radar (AirSAR) measured data of the polarimetric Mueller matrix, inversions of the fractional volume per unit area of spatially inhomogeneous canopy and underlying surface moisture are discussed.

To solve 3D-VRT equation for the model of spatially inhomogeneous scatter media, the finite enclosure of the scatter media is geometrically divided, in both the vertical z and horizontal (x, y) directions, to form very thin multi-boxes. The zero-th order emission, first-order Mueller matrix of each thin box and an iterative approach of high-order radiative transfer are applied to deriving high-orderscattering and emission of whole inhomogeneous scatter media. Numerical results of polarized brightness temperature at microwave frequency from inhomogeneous scatter model such as vegetation canopy are calculated and discussed.

A massive sandstorm enveloped most part of northern China during the spring season 2002. Monitoring the evolution of sandstorm and desertification has become one of the most serious problems for China's environment. Since 1989, one of the most advanced and operational passive microwave sensors is the DMSP SSM/I (special sensor microwave) imager operated at seven channels (19, 37, 85 GHz with the vertical and horizontal polarization and 22 GHz with vertical polarization only). In this paper, the sandstorm and desertification indexes. SDI and DI, are derived from the radiative transfer equation and are employed with multi-channel measurements of the DMSP SSM/I for monitoring the sandstorm and desertification in northern China. Some SSM/I data in 1997 and 2001 are employed. The algorithm of Getis statistics is developed to categorize the spatial correlation and its evolution these days. It is demonstrated that the SSM/Iindexes SDI and DI, and its Getis statistics are well applicable for monitoring t...

To measure completely polarimetric scattering of a layer of randomly distributed, chirally active small particles, the scattering amplitude function of weakly chiral, electrically small spheroid is derived for the first time. It is used to construct the Mueller matrix solution for polarimetric scattering of four Stokes parameters from a layer of random chiral-spheroids. Copolarized and cross-polarized backscattering σ c , σ x upon the incidence polarization, that is,\qy2 the elliptic and orientation angles (χ, ψ), can be numerically obtained. It has been known that the maximum of copolarized backscattering σ c is largely located at (χ = 0°, ψ = 0°, 180°) or (χ = 0°, ψ = 90°) depending on spatial orientation and particle shape (e.g., needles or disks). However, our results show that chirality can asymmetrically shift this location. It is due to the chiral parameter, which appeared in the depolarized scattering amplitude functions. This shift can be evaluated by the measurement of four scattered Stokes para...

In order to numerically simulate bistatic scattering from a fractal rough surface at low grazing angle (LGA) incidence, a hybrid approach of the forward-backward method (FBM) with spectral accelerate algorithm (SAA) to solve the magnetic field integral equation is developed. Numerical bistatic scatterings from one-dimensional perfectly conducting fractal rough surface with and without a regular object presence are accomplished. Accuracy and efficiency of FBM/SAA are discussed. Numerical simulations demonstrate functional dependence of bistatic scattering upon fractal geometry, especially at LGA. The linear relationship between the envelope slope of bistatic scattering pattern and fractal dimension is discussed.

To numerically simulate bistatic scattering from fractal rough dielectric surface at low grazing angle (LGA), a hybrid approach of the forward-backward method (FBM) with spectral accelerate algorithm (SAA) is developed. The impedance matrix of the electric or magnetic field integral equations is decomposed into four submatrices for fast FBM algorithm. Bistatic scattering from and transmitting through one-dimensional (1-D) fractal rough surface with any dielectric constant are calculated. Angular reflectivity and transmission demonstrate the energy conservation, and its functional dependence upon the surface parameters are discussed. A linear relationship between the envelope slope of the bistatic scattering pattern and the fractal dimension is also discussed.

In this paper, by using the joint probability density functions (PDFs) of the scattering amplitude functions, the PDFs of four Stokes parameters are derived for multi-look polarimetric synthetic aperture radar (SAR) data. The PDFs are formulated analytically with dependence on the look number and mean values of the scattered Stokes parameters. The statistics histograms of four Stokes parameters are validated by the AirSAR imagery at multi-P, L, and C bands observed over some typical sites such as water, sparse trees, and forest. An effective look number might be specially assigned to provide an excellent match with the AirSAR data. Potential application of the statistics of four Stokes parameters to classification of the terrain types is studied.

Two key parameters to affect microwave backscattering from the land surface are the surface roughness and soil wetness. A novel genetic algorithm is developed for multi-parameter retrieval of land surface roughness and soil wetness from angular backscattering observations. Parameters of wetness and roughness are encoded into genes. Genes are constituents of chromosomes, which undergo optimal selection based on a natural evolutionary process in the genetic algorithm. The theoretical model of a two-scale rough surface is employed for computation of the cost function. Results retrieved using this genetic algorithm are compared well with ground data measurements. This study presents an example of the genetic algorithm for application of multi-parameter retrieval in remote sensing.

It is known that the angular correlation function of scattered fields from a randomly rough surface has a conspicuous enhancement on the memory line, and becomes much smaller away from the line. The finite-element method is employed to solve the problem of scattering from a scatter target situated over a randomly rough surface. Numerical results show that the width of the angular correlation function becomes broadened owing to the presence of the target. It can be used to detect the presence of a target placed over the rough surface.

Flooding is always closely related to the regional geographical conditions. The globally optimized algorithms of data validation for the DMSP SSM/I, such as a rule TB22v-TB19v > 4K, are not applicable to monitoring flood without specific knowledge of regional geography. This Letter investigates the SSM/I data during the rain and flood time of June-July 1996 in southern China, and proposes a flooding index FI = TB37h - TB85h and its regional threshold value F0 for monitoring flooding. Flooding is considered to occur when FI < F0. In two examples selected from the watery Wuhan and mountainous Wuyuan regions, temporal and spatial variation of the flooding index FI and another two indices, the scattering index SI and surface polarization index PI, are discussed. Numerical simulation of radiative transfer modelling for these indices is also presented.

Observations by active ERS scatterometer and passive SSM/I radiometers on the same area and for the same time period are correlated using the data measured by ERS-1 and multichannel SSM/I over snowpack located in Greenland, Siberia, the Alps, etc. It is demonstrated that multichannel analysis by both active and passive sensors is very helpful for monitoring temporal variation of complex snow covers. Back-scattering from multilayer, strongly fluctuating random media is calculated for ERS data simulation. The radiative transfer of thermal emission from multilayer media of dense scatterers is solved for SSM/I data simulation. Simulations of both back-scattering and emission show good correlation and functional dependence on highly-variable snowcovers. Theoretical results also compare well with observation data.

Numerical simulation of coupled vector radiative transfer equations for a multi-layer model of dense scatterers is developed. The effect from ice layer formed during melting is studied to take account of anomalous spectrum of brightness temperature at the SSM/I channels. This approach is applied to analysis of SSM/I data observed over the southern and central Greenland snow, which can show distinctly different spectral characteristics, T B19 > T B85 or T B19 < T B85. Ice layers produced in melting and ice grain sizes can significantly affect functions of thermal emission at SSM/I channels.

The accuracy of snow depth estimation is affected significantly by the regional surface type. We have developed a theoretical model of vector radiative transfer (VRT) for snowpack/vegetation canopies at SSM/I channels. The vegetation canopy is modelled by a layer of nonspherical particles, and the snowpack is modelled as a layer of dense spherical particles. By numerically solving two coupled VRT equations for multi-layer models of different surface types such as tree/snow, grass/snow and snowpack only, two scattering indices SI1 = T B19v - T B37v and SI2 = T B22v - T B85v are obtained for a variety of snow depths (SD) and ice-grain sizes. These results are combined as a mesh graph in the figure of SI1 versus SI2 . When the SSM/I TB data is observed, its location in the mesh graph can indicate the estimation of SD. Our results compare well with the SSM/I data of the U.S.A. east coast January blizzard, 1996.

Retrieval of the biomass parameters from active/passive microwave remote sensing data is performed based on an iterative inversion of the artificial neural network (ANN). The ANN is trained by a se...

Multi-frequency microwave radiometers (3-09, 5-4, 9-6, 36GHz) designed by the Changchun Institute of Geography have been used to measure dual-polarized brightness temperature of different crop canopies in the People's Republic of China. A numerical modelling of radiative transfer of multi-layer random media is developed. All parameters calculated by the empirical formulations arc chosen in a consistent way. Numerical results arc compared with the experimental measurements. The functional dependence of brightness tempera-lure on frequency, polarization, observation angle and crop parameters such as moisture content, crop height, etc. arc numerically discussed.

By using the fluctuation-dissipation theorem, the strong fluctuation theory is applied to the calculation of brightness temperatures from a bounded layer of random discrete scatterers with the zeroth- and first-order approximations. Various functional dependences on wavelength, polarization, observation angle, medium depth, scatterer constituents, and other physical parameters are discussed. Theoretical results are favorably matched with experimental data for passive remote sensing of snowpacks.