The spatial representativeness error of in situ soil moisture (SM) is recognized as a major source of uncertainty when validating satellite SM products with a spatial resolution of tens of kilometers. Site underrepresentation is primarily caused by environmental heterogeneity, but their relationship remains poorly understood. Here, we assessed the spatial representativeness of in situ SM from 322 strictly screened stations worldwide relative to coarse-resolution (~0.25°) satellite footprint based on the extended triple collocation (ETC) method. We then evaluated the influence of the heterogeneity of four environmental factors (soil texture, land cover types, elevation, and vegetation coverage) on site representativeness. Moreover, we calculated SM variability within the satellite footprint based on 1-km SM data to explore its relationship with environmental heterogeneity. Results indicate that about 63% of the sites have relatively good spatial representativeness (ETC-derived correlation coefficient$\ge 0.7$). Soil texture and land cover exhibit greater heterogeneity across the mid and high latitudes of the Northern Hemisphere. The larger heterogeneity in elevation and vegetation coverage is primarily found in regions with significant ridges and dense vegetation, respectively. Land cover is the major factor influencing the spatial representativeness of SM sites, and the increase in the heterogeneity of land cover enhances SM variability, which negatively impacts site representativeness. The in situ SM can be more representative when the proportion of the land cover type where the site is located is higher or when there are fewer land cover types within the satellite footprint. Moreover, it is found that the newly proposed metric of the similar area ratio of sites, as a measure of land cover heterogeneity, can effectively reflect SM variability. This metric can also serve as a supplementary criterion for selecting representative sites, particularly in situations where sites are sparse and the ETC method is inapplicable. These findings provide useful references for robust evaluation of satellite SM products based on in situ measurements (e.g., in situ SM upscaling and SM site deployment).

This study examines the microwave emission properties of snow-covered sea ice, modeled as a layer with rough top and bottom boundaries. The emission model is based on the first-order solution to the radiative transfer equation (RTE). This equation describes the brightness temperatures that propagate both upward and downward in the layer, and it is solved numerically using the eigenvalue method. Additionally, the model takes into account contributions from irregular boundaries as well as surface and volume scattering interactions. The study looks at the variability of brightness temperature and the growth of different ice types. It uses modulation theory to consider the roughness of the boundaries, showing how this affects the emission from various types of snow-covered sea ice. We used polarization ratio-gradient ratio (PR-GR) space to analyze ice-grown transitions quantitatively. The thresholds of PR were found to be sensitive to changes in ice concentration. It was observed that the threshold of PR did not change with the roughness of the top boundary. We also found that the roughness of the ice surface has a more significant impact on emission than that of the snow-covered ice. It also observed that roughness effects are more potent in H-polarization than in V-polarization for each of the seven ice types. The roughness effects are weaker at large look angles, indicating stronger volume scattering. The study concludes that the presence of roughness in the ice layer’s boundaries leads to noticeable variations in brightness temperature.

The high-resolution synthetic aperture radar (SAR) images reveal more details about the complex-structured targets. In this study, we simulate the SAR echo and images taking the Great Belt Bridge over the sea surface as an example. By decomposing the bounces, we show that the scattering feature patterns come from the different reflections or scatterings. The target models include three types: two suspension cables of the same height and suspension cables with curvatures. A detailed deck model was also constructed to compare the environment of bridge construction with and without a roadway deck. The results demonstrate the line characteristics of different slant ranges in the simulated SAR images. We can appropriately interpret the scattering features from such complex targets by comparing the results with the actual SAR images.

In this study, a model-based neural network approach is proposed to retrieve ancillary parameters related to oil pollutants at sea. The proposed methodology consists of two pillars. First, an electromagnetic scattering model is used to generate radar backscatter for slick-free and slick-covered sea surface at variance of incidence angle, faction of water into the oil and oil thickness. Then, these radar backscatter values are combined to generated a metric adopted fro the retrieval process, namely the damping ratio. Second, an artificial neural network is first trained on the simulated damping ratio DR and then applied to actual synthetic aperture radar imagery to retrieve oil thickness and seawater volume fraction. Results, obtained processing synthetic aperture radar scenes collected during the Deep Water Horizon oil accident by the L-band uninhabited aerial vehicle synthetic aperture radar (National Aeronautics and Space Administration - Jet Propulsion Laboratory), show the soundness of the proposed methodology.

The uncertainty inherent in validating satellite-derived soil moisture (SM) products is significantly attributed to the spatial underrepresentation of in situ SM measurements. The main reason for this phenomenon is the varying environmental conditions (called as environmental heterogeneity) within the satellite footprint. To better understand this issue, we assessed the spatial representativeness of in situ SM from 383 strictly screened stations worldwide relative to the coarse-resolution (~0.25°) satellite footprint and analyzed the effects of four environmental factors (i.e., soil texture, land cover, elevation, and vegetation coverage) using the extended triple collocation (ETC) technique. Results show about 63% of the sites have satisfactory levels of spatial representativeness (ETC derived correlation coefficient ⩾0.7). Land cover is the foremost factor affecting the spatial representativeness of SM sites. The in situ SM can better represent the true variability of SM when the proportion of land cover types where the site is located is higher or there are fewer land cover types within the satellite pixels.

The spatial resolution of existing satellite soil moisture products is very coarse (~25 km), and thus there is usually significant spatial heterogeneity in the land surface covered by satellite footprints. However, the effects of spatial heterogeneity on satellite soil moisture products are largely under-studied previously. The study firstly evaluated seven satellite soil moisture products comprising SMAP (DCA, IB, and MTDCA), SMOS-IC, AMSR2 (LPRM and JAXA) and FY-3C at a global scale using the extended triple collocation (ETC) method. Then, the skills of these products under a wide range of surface heterogeneity including heterogeneity in vegetation coverage, terrain, land cover, and soil texture were ascertained. The results indicate: (1) SMAP-IB and SMAP DCA products generally outperform others, followed by SMOS-IC and SMAP MTDCA which also exhibit satisfactory performance; (2) heterogeneity in vegetation coverage, terrain, and land cover generally decreases the R value of satellite soil moisture products, while heterogeneity in soil texture has an insignificant effect on product skills; (3) L-band products demonstrate greater stability compared to C/X-band datasets across various surface heterogeneity.

This paper examines the influence of Earth’s ellipsoidal shape on the system performance of the Moon-based SAR (MBSAR). Specifically, the corresponding impacts on the beam-pointing accuracy and imaging performance are considered. Following the observation geometric model, we formulate the vectors of light-of-sight (LOS) and target of interest (TOI) under the Ellipsoidal Earth model. By considering Earth observation at different epochs, we analyze the positioning errors of the TOI and imaging deterioration in the MBSAR. The results show that the impact of the Earth ellipsoid, depending on the look angle and orbital elements of MBSAR, varies spatiotemporally. In addition, the synthetic aperture time can contribute when considering the influence of Earth ellipsoid on the imaging performance in the MBSAR.

Satellite soil moisture products have great potential for many applications, such as drought monitoring and landslide warning. However, these applications often require accurate and continuous soil moisture records, and the missing values in satellite soil moisture datasets often hamper the usefulness of these products for such applications. This study firstly proposed and compared three approaches, i.e., linear regression, linear rescaling, and random forest to fill the missing values in the SMAP soil moisture products in both temporal and spatial dimensions, based on the seamless ERA5 data from 2016 to 2019. Then, a total of twelve auxiliary data were incorporated into the training datasets of random forest to improve the accuracy of gap-filled SMAP data. Finally, the gap-filled SMAP data were compared with the original SMAP data and validated by in situ measurements from 1071 sites worldwide. The results indicate: 1) when using only the ERA5 datasets, the random forest performs better than linear regression and linear rescaling methods in the training phase, but its skill degrades noticeably in the validation phase; 2) by adding the auxiliary data, the performance of random forest improves significantly in the validation phase; 3) the gap-filled SMAP data maintain or even exceed the accuracy of the original SMAP soil moisture, demonstrating the effectiveness of the proposed gap-filling method.

This article employs the modulated correlation function (MCF) model to describe the ocean surface in radar backscattering. Surface correlation functions derived from wind wave spectra: Apel, Elfouhaily, Kudryavtsev, and Hwang models are examined. The spectral properties of the above spectra are also analyzed, including the height, slope, and saturation spectra. The results suggest that the MCF model is applicable in depicting spatial correlation of ocean surface. At the same time, the spectrum of MCF may not be proper in describing the energy cascade of ocean waves. The comparisons of backscatter are made among the MCF model and wind wave spectra, with regard to dependences of radar frequency at L-, C-, X-, and Ku-bands, wind vector, and incidence angle. The results indicate that the MCF model yields the overall minimum errors in radar backscatter against geophysical model functions (GMFs). We validate the model with the airborne and spaceborne radar measurements and show that the MCF model gives good agreement at the C-band but only moderate at the Ku-band. Meanwhile, the backscatters calculated based on different wind wave spectra can significantly deviate from each other, related to the magnitudes of the short-wave spectrum. Besides, the breaking wave effect on radar backscatter is also discussed.

In this study, the potential of electromagnetic scattering models to retrieve quantitative parameters of sea oil spills is investigated using an artificial intelligence (AI)-based approach. The backscattering coefficient of a slick-covered sea surface is predicted using the advanced integral equation model augmented with the model of local balance (MLB), an effective dielectric constant model, and a composite medium model to include the effect of an oil slick. Damping ratios (DRs), predicted for different oil parameters (namely, the oil thickness and seawater volume fraction), are used to train and test a four-layer neural network. Once successfully tested, the neural network is applied to an uninhabited aerial vehicle synthetic aperture radar (UAVSAR) image collected during the DeepWater Horizon (DWH) oil spill accident to retrieve the oil slick thickness and volume fraction of seawater in the oil layer. The inversion results show that the thicker (i.e., 2–4 mm) emulsions are located in the south and west of the slick and they are surrounded by thinner (i.e., < 1 mm) oil films. In addition, the seawater volume fraction in the oil slick is found to be about 20%–30%. Results are contrasted with optical data and previous studies of the same accidental oil spill, showing qualitatively good agreement.

Surface parameters (roughness and soil permittivity) are crucial for characterizing backscattering from bare soil. However, the estimation of roughness parameters (root-mean-square (rms) height and correlation length) depends on the sample surface size. The conversion between dielectric constant and soil moisture is disturbed by the dielectric model. These estimation biases significantly compromise the reliability of backscattering coefficients derived from analytic modeling, numerical simulations, and experimental measurements. In this study, we illustrate the statistical relationship between sample surface size and estimation bias of surface roughness at varied accuracy levels. To quantify the estimation bias of surface roughness with sample surface size and the estimation bias of soil permittivity, we analyze the propagation from the estimation bias of surface parameters to the backscattering coefficient error by the advanced integral equation model (AIEM) model. By comparing it with measurement data, we quantitatively confirm the impact of roughness parameter estimation bias. Ultimately, quantifying the backscattering coefficient error as a function of sample surface size and incident angle allows for selecting the optimal sample surface sizes suitable for L-band synthetic aperture radar (SAR) simulation and soil moisture retrieval, along with their applicability over various incident angles. This study suggests sample surface sizes for estimating roughness parameters and backscattering coefficients at various levels of accuracy.

The comprehensive and robust assessment of diverse global-scale satellite-based soil moisture products from various satellite data sources (e.g., different frequencies and incidence angles) and retrieval algorithms is essential for the refinements as well as applications of these products. To date, soil moisture retrieval algorithms and products are rapidly evolving and their updated iterations are ongoing. In support of the validation activities of recently developed/reprocessed satellite soil moisture products, the study firstly assessed eight commonly-employed satellite soil moisture datasets comprising SMAP (DCA, IB, and MTDCA), SMOS-IC, AMSR2 (LPRM and JAXA), FY-3C, and ESA CCI on a global scale using three different strategies, i.e., ERA5 reanalysis soil moisture dataset with similar spatial resolution to satellite products,in situmeasurements from densely-instrumented networks worldwide with mitigated spatial mismatch between ground site and satellite pixel, and the Extended Triple Collocation (ETC) method that can obtain error indicators relative to ground truth. The skills of these products under a broad range of vegetation density, land cover and climate types, and surface heterogeneity (heterogeneity in terrain, land cover, soil texture, and vegetation coverage) were also examined. The results indicate: (1) different soil moisture products show overall consistency in skill ranking under three different evaluation strategies, except for SMAP DCA, SMAP-IB, and SMAP MTDCA in terms ofRvalue; (2) ESA CCI, SMAP-IB, SMAP DCA products generally perform better than the others under three strategies, and SMOS-IC and SMAP MTDCA also show satisfactory performance concerning ubRMSD andRvalues; (3) vegetation density exerts visible influences on satellite soil moisture datasets. Specifically, the C/X-band (AMSR2 and FY-3C) and L-band (SMAP and SMOS) products display the optimal skills under sparse and moderate vegetation coverage respectively, and the impacts of vegetation density on C/X-band products are evidently stronger than those on L-band datasets. The errors of satellite soil moisture data also increase as the increase of heterogeneity in terrain, land cover, and vegetation coverage, while the effect of heterogeneity in soil texture on the skill of satellite soil moisture products is insignificant; (4) the skills of L-band products are more stable than those of C/X-band datasets under different ground conditions.

We analyzed the spatiotemporal variations of the imaging swath and region using the lunar-based synthetic aperture radar (LBSAR) for Earth observation. The ground and slant swaths following the Earth-observing geometry and synthetic aperture radar (SAR) configurations were formulated. The bounds of the imaging swath were highlighted for appropriately configuring the LBSAR. The results suggested that the imaging swath is determined by the coupling of orbit elements, near-look angle, antenna beamwidth, and to some extent by the Earth’s ellipsoidal shape and LBSARs site location on the Moon’s surface. The side-looking direction, despite its modest impact on the imaging swath, plays a dominant role in ascertaining LBSARs imaging region. Furthermore, spatiotemporal variations of ground swath in various epochs were analyzed to emphasize the challenge and importance of characterizing Earth observation capability in the LBSAR.

We analyze the bistatic scattering of a gamma-distributed rough surface to explore the scattering properties in the context of polar angular patterns and specular angular beamwidth as a dependence on the degree of non-Gaussianity. We focus on the incoherent scattering because of its importance in remote sensing of rough surfaces, e.g., soil and sea ice. We use a 3 dB angular width, the width of the range of scattering angle, as a measure of angular broadening and dropping-off of the scattered power. Some insights into the physical significance are perceived. For the co-polarized scattering in the incident plane, as the skewness and kurtosis of Gamma height probability density (HPD) surface decreases, the scattering becomes narrower and concentrates more at the specular direction and has a faster decay than in the Gaussian surface at large angles. As the shape parameters increase, the incoherent scattering concentrates more at 0° of scattering angle for the cross-polarized scattering in the cross-plane. The narrow 3 dB angular width is located at the larger shape parameters (or low skewness and kurtosis) with a small to moderate root mean square (rms) height. The overall 3 dB angular width of the exponential power spectral density (PSD) surface is smaller than the Gaussian PSD surface. This study reveals that angular broadening is narrower, and the dropping-off of the scattered power is more significant for Gamma height distributed and exponential-correlated surfaces than the Gaussian surface.

A comprehensive exploration of the transformer and dual-receiver system-based direction-of-arrival (DOA) estimation is presented in the context of sea surface scattering, particularly under varying sea conditions. A bidirectional encoder representation from transformer (BERT) with a physics-based loss function is utilized to process two individual channel radars. The datasets are the radar scattering coefficients of sea surface simulated at C-band for copolarizations and cross-polarizations. Through detailed analysis of simulated datasets and root mean square error (RMSE) evaluations, the model’s performance is investigated across different observation modes, namely, the co-polar (CP), co-azimuth (CA), full-bistatic (FB), and Beaufort wind scale from 3 to 5. Our study demonstrates that the bidirectional encoder representation from transformer model, employing a physics-based loss function, outperforms the baseline long short-term memory (LSTM) model, especially under high noise levels and with larger datasets. Significant correlations between wind conditions and DOA accuracy are observed, highlighting the bidirectional encoder representation from transformer model’s adaptability to dynamic environmental factors, particularly under increased wind scales. The choice of observation mode, with CP and FB consistently outperforming CA, proves pivotal. Precise simulation of speckle variations and optimized observation mode selection are identified as crucial avenues for enhancing the model’s practical utility.

It is known that co-polarized backscattering reaches saturation and damping quickly at high wind speeds. Also, the cross-polarized scattering has higher sensitivity to high wind speed. Motivated by these inherent backscattering characteristics, we explore the linearly and circularly polarized bistatic scattering behaviors at high winds on the whole scattering plane. Numerical results show that, for HH polarization, the deep scattering valley appears at the cross-plane. In contrast, VV polarization is not so at cross direction but tends to be so in the forward region. The scattering valley locates in the incident plane for HV and VH polarizations; that is, the depolarization effect is relatively weak in the incident plane. The difference between HV and VH polarizations is complementary to backscattering in the context of winds retrieval. The circular polarization can better retrieve high winds, e.g., 40m/s.

The Geladandong region, a source of the Tuotuo River, the primary source of the Yangtze River, is an essential ecological protection region in China. The study of glacier classification accuracy in Geladandong is still limited. Based on the Landsat remote sensing images, the second glacier inventory, a digital elevation model, and meteorological data, we used the random forest method to accurately classify glacial and non-glacial regions. We compared the changes in annual and interannual glaciers between 1999 and 2020. Besides, we quantitatively assessed the influencing factors of glacier change. The results can be summarized as follows. During the 21 years from 1999 to 2020, the annual and interannual area of glaciers in Geladandong has been shrinking, mainly caused by insufficient precipitation to compensate for the melting of the glaciers due to rising temperatures.

Soil moisture (SM) is influenced by changes in meteorological elements and vegetation, as well as by environmental heterogeneity. Due to the prevalence of extreme events in the past two decades, this complexity is growing in the 21st century. In this study, the global spatiotemporal trend of SM and its possible influencing factors were investigated by using the satellite-based ESA CCI SM from 2000-2021. The results reveal global SM generally declines at a rate of -1×10-4m3m-3yr-1, dominated by a drying trend in the southern hemisphere. From a global perspective, the driving force of precipitation and vegetation on SM fluctuation is stronger than that of temperature. Different environmental variables (e.g., land cover, soil texture, and terrain) have different regulatory effects on SM changes. In contrast to other types, there are general wetting trends of SM in croplands, savannas, forests, loam soils, and high elevations (> 1000 m). The response of SM to temperature and vegetation is greatly limited under barren and forests, respectively. The impact of temperature on SM is enhanced with increasing clay content or decreasing sand content. The high positive correlation between precipitation and SM is rarely influenced by environmental factors compared to temperature and vegetation.

Estimation of surface soil temperature (ST) using passive microwave remote sensing is still a great challenge, especially under diverse land surface conditions. The microwave radiation imager (MWRI) on board Fengyun (FY)-3D satellite is the latest Chinese FY-3 series multiband microwave sensor in orbit, providing a valuable opportunity for estimating surface ST on a global scale. In this study, the FY-3D brightness temperature from 10.7 to 36.5 GHz and the topsoil temperature simulations from ERA5-Land and GEOS-FP from 2019 to 2020 were used to build the Multi-Band Temperature Retrieval Model (MBTRM). The universal global optimization algorithm was employed to establish the MBTRM on a per-pixel basis globally. Ground measurements from a total of 1221 sites worldwide were adopted to fully examine the performance of the proposed MBTRM. Moreover, four widely used passive microwave-based ST models/products were compared. The results reveal the proposed MBTRM performs the best with an averaged root mean square difference (RMSD) of 3.03 K and 3.79 K at descending and ascending overpass respectively and a correlation coefficient larger than 0.9. The advantage of the newly developed MBTRM is that it is user friendly (with explicit formula) and can be applied globally with satisfactory accuracy, and provides a good supplement to optical satellite based temperature products.

This paper examines the radar penetration into a rough soil surface with a vertical moisture profile. Numerical analysis shows that the penetration depth occurs at a larger dynamic range of incident angle in V polarization but in H polarization. The maximum polarization difference occurs at low frequencies (e.g., L band) and at a large incident angle (e.g.,70°~85°), which are rarely used due to low backscattering returns. Of the two roughness parameters, the RMS height significantly influences on the penetration depth more than the correlation length. The dependence of penetration depth on the wave polarization moderates when the surface becomes rougher. Results also suggest that the penetration depth is sensitive to the inhomogeneity of moisture profiles due to the temporal evaporation process, indicating that the penetration depth is difficult to quantitate and an equivalent model to estimate it may be inappropriate, or at least it is difficult to establish.

In this paper, we proposed a modulated correlation function to characterize the multiscale property of the sea surface and adopted it in the AIEM (Advanced Integral Equation Model) to calculate the radar backscattering. Comparisons of NRBCS (Normalized Radar Backscattering Cross Section) with GMFs (Geophysical Model Functions) predictions and radar measurements are conducted in various wind conditions. Good consistency and accuracy confirm the proposed model’s accuracy and applicability in predicting radar backscattering. In addition, the relations between two modulation parameters and wind vectors are analyzed at C-band. The effective correlation lengths determined from the modulated correlation function show strong wind dependence, besides the incident angle and frequency in the context of radar backscattering.

This paper studies the orbital perturbation effects on the azimuthal resolution in lunar-based synthetic aperture radar (LBSAR). We derive explicit expressions for the Doppler frequency modulation rate (DFMR) and beam-crossing velocity using the antenna beam pointing and orbit models. Following that, the azimuthal resolution is expressed in line with orbital elements and SAR configurations. The results show that the long-term orbital variations caused by accumulated perturbation effects significantly affect the azimuthal resolution, which, in effect, produces aperiodic variations in the azimuthal resolution. Such a phenomenon is most distinguished for a large LBSAR look angle, leading to a fluctuation of over 30% or even larger in the azimuthal resolution across different cycles. Additionally, the errors give rise by short-term orbital perturbations could impact azimuthal resolution to a lesser extent, with corresponding fluctuations consistently below 3%. The findings reveal that it is imperative to consider the irregular variability of azimuthal resolution due to orbital perturbations in the LBSAR.

Lunar Terahertz Surveyor for Kilometer-scale Mapping (TSUKIMI), a collaborative mission for lunar exploration, is introduced. We numerically explore terahertz scattering and emission from the lunar surface as an initial step forward. Regarding the characterization of the lunar surface, the power-law spectrum is adopted to model the roughness. While for the dielectric constant of the lunar surface, tentatively simplifying into a homogeneous condition. The scattering and emission models are briefly described. To investigate the terahertz scattering effect from the lunar surface, we conducted the preliminary experiments with dielectric constant and surface roughness using THz- TDS and T-Ray 4000 Picometric, respectively. The measurement results are subsequently presented.

The Fengyun (FY)-3D satellite launched in November 2017, provides the latest multi-frequency brightness temperature (TB) of Chinese FY-3 series satellites. In this study, the FY-3D TB at five frequencies from 10.7 to 89 GHz during 2019 to 2020 were intercalibrated against AMSR2 TB using two intercalibration methods, i.e., the commonly used global linear regression method and the global per-pixel linear regression method proposed in the study. The effects of different environmental variables (e.g., land cover and its heterogeneity, climate types, water body fraction, terrain and vegetation coverage) on the calibration accuracy were investigated. The results reveal the global per-pixel linear regression method performs much better than global linear regression method with an averaged root mean square difference (RMSD) of 2.93 K at ascending overpass and 2.34 K at descending overpass. The water body fraction has the greatest influence on calibration accuracy, followed by terrain, land cover heterogeneity, and vegetation coverage. The RMSD is relatively lower in savannas and barren lands as well as in arid, cold and tropical climatic zones than that in other land cover and climate types.

This study examines the speckle properties of SAR (synthetic aperture radar) images of non-Gaussian rough surfaces. Three height probability density functions were considered: Gaussian, exponential, and Weibull forms, while two forms of the power spectral density were considered: self-affine fractal and exponential. Computer realizations of the rough surface ensemble samples were done, followed by a physics-based simulation of SAR complex images of the rough surface. Speckle distributions were analyzed against the fully-developed speckle model. Illustrative examples with various surface roughness scales in the context of SAR imaging are given. When the surface HPD is non-Gaussian, image speckle is no longer following the Rayleigh distribution, regardless of the PSD. The speckle of exponential HPD surface profoundly away from Ravleigh model, especially under self-affine fractal PSD.

This paper analyze polarimetric signatures SAR image of an electrically large cargo ship sitting over a sea surface. For the systematic and illustrative purpose, we applied a fully coherent SAR image simulation in streamlining the data flow from backscattered field to complex single look images recently developed to generate high-resolution full polarization SAR images. Three scenes of the ship heading parallel, orthogonal, and 45° to the azimuth direction were simulated. An attempt is made to exploit the polarimetric analysis via the Y4R decompositions. Results reveal that the polarimetric features of sea surface and its interactions with the cargo ship at different headings are quite distinct. For headings parallel to the SAR azimuth direction, richer and diverse polarimetric signatures are presented, but they lose or diminish when the heading is 45° to the azimuth, where surface scattering becomes predominant. A large look angle may attribute this change.

The Lunar-Based SAR (LBSAR) has plenty of advantages, for example, large-scale mapping, high temporal resolution, and long-period operation, for earth’s observation. However, because of the extremely high orbit, there is a round-trip propagation time delay on the order of several seconds for the LBSAR signal. Consequently, the conventional “stop-and-go” approximation used for the synthetic aperture radar (SAR) signal modeling is no longer applicable in the LBSAR. Generally, such an approximation raises two groups of effects. One is the effect of the antenna displacement during the pulse duration, which gives rise to the center frequency shift and the frequency modulation (FM) rate variation in the chirp signal. The other one is correlated with the range history and manifests as the Doppler error. These two effects might, respectively, give rise to image distortions in LBSAR imaging. This letter quantitatively analyzes the effects of the “stop-and-go” approximation on the LBSAR imaging performance. Theoretical analysis shows that the impact of the antenna displacement during pulse duration contributes little to LBSAR imaging, and thus it can be reasonably ignored in signal processing. In contrast, the azimuth imaging is sensitive to the Doppler error owing to the “stop-and-go” approximation. Under this effect, there is a severe position deviation in the azimuth direction, although the focusing quality is almost uninfluenced. To compensate for the Doppler error adequately, an effective method is proposed to retrieve the range history of the LBSAR with “nonstop-and-go” configuration. Finally, point target responses are simulated for verifying the theoretical analysis and proposed method.

This article presents the computation of the backscattered field in simulation of synthetic aperture radar (SAR) raw data. We develop an improved Kirchhoff approximation to estimate the surface fields induced by the incident waves. The improved Kirchhoff approximation is a second-order iterative solution of the integral equations governing the targets’ electric and magnetic current densities. In computing the second-order re-radiated fields, we applied the shoot-bouncing-ray (SBR) technique to enhance propagation path tracking efficiency. The geometrical theory of diffraction (GTD) was employed to account for the diffraction fields from the edges and wedges, which constitute the total scattered fields collected by SAR in synthetic aperture. As the SAR moves along the azimuth direction, the backscattered signal computation is repeated as the antenna beam crosses the targets. This procedure demands heavy computational resources but requires no priori assumptions of radar cross-sections (RCSs) nor speckle statistics. The raw data is generated as an output signal of the SAR system response into which the backscattered signal is input. We chose three types of the target to demonstrate our approach to confirm the effectiveness. The first set of targets are three dihedral reflectors – two of them were rotated to constitute three unique scattering matrices. The results show that the polarimetric information, both relative amplitudes and phases, are well–preserved. The second target is a rough surface with exponential power spectral density and Gaussian height probability density. We examined image speckle statistics of multi-looking image. The third target type is an electrically large cargo ship (container) sitting over a sea surface. Polarimetric analysis via the Pauli and Y4R decompositions reveals that the polarimetric features are well preserved. Simulation results demonstrate that the present approach is fully coherent in streamlining the data flow from backscattered field to complex single look images within the SAR imaging scene.

In this study, the bistatic radar scattering coefficients related to an oil-covered sea surface are predicted by modeling both the oil damping effect on surface roughness–through the advanced integral equation method–and the oil modification on the dielectric properties of the scattering surface. The bistatic scattering is analyzed in the whole upper scattering space under different radar frequencies, incidence angles, wind speeds, and oil thicknesses. Numerical predictions show that the scattering energy of an oil-covered sea surface is generally higher in the forward scattering zone than that in the backward one. In addition, the oil damping effect is the main mechanism ruling the scattering behavior in the backward region. The information related to the bistatic scattering geometry is also explored to retrieve oil thickness, representing one of the key parameters for radar-based marine oil pollution observation. A new index is proposed to quantify the sensitivity of bistatic scattering coefficients to oil thickness in different cases: single-polarization features, dual co-polarization features, namely, the polarization ratio (PR) and the normalized polarization difference index (NPDI), and dual-angular scattering features. Numerical results show that the bistatic scattering coefficients result in an enhanced sensitivity to oil thickness with respect to the monostatic case. The single HH-polarized scattering coefficients show better oil thickness sensitivity in the backward region, while the VV-polarized ones are more sensitive to oil thickness in the forward region. The combination of dual-polarized scattering coefficients significantly improves the oil thickness sensitivity compared to single-polarization radar observations, especially in the forward region. PR outperforms NPDI, even though the latter can suppress the effect of wind speed. The combination of dual-angular observations can significantly increase its sensitivity of oil thickness in the backward region but at the expense of reduced sensitivity in the forward region.

Oils floating on the sea surface can be observed as “dark” patches on radar images since the backscattered signals from the contaminated area are reduced in two dominant ways. First, oil slicks could damp short gravity and capillary waves on the sea surface responsible for backscattering energy. Second, the oil-covered sea surface permittivity decreases significantly if the oil film is sufficiently thick or mixed with seawater, i.e., oil emulsion. In this article, the geometry of the oil-covered sea surface is accounted for by the damping of sea waves, which is described by the model of local balance (MLB) combined with the sea wave spectrum. The radar backscattering is predicted by the advanced integral equation method (AIEM) model. The reflection coefficients are calculated based on a layered-medium model to analyze the impact of oil thickness and emulsions on the radar scattering. Numerical simulations demonstrate that: 1) the sensitivity to oil thickness and water content of the oil spills increases when the radar frequency increases; 2) the backscattering signals exhibit a nonlinear behavior with respect to oil thickness; and 3) high wind speed can generally narrow the difference between the radar backscattering from the clean and oil-covered sea surface, while the incidence angle has little effect. Numerical simulations are then compared with the multifrequency synthetic aperture radar observations acquired during the Gulf of Mexico Deepwater Horizon (DWH) oil spill accident and the 2011 Norwegian Clean Seas Association for Operating Companies (NOFO) oil-on-water exercise. Comparison results show that it is possible to estimate the oil thickness at reasonably good accuracy.

As the second generation of Chinese polar-orbiting meteorological satellite missions, the Fengyun (FY)-3D satellite provides the latest multi-frequency brightness temperature (TB) of FY-3 series satellites. The microwave radiation imager (MWRI) boarded on FY-3D has similar sensor configuration as Advanced Microwave Scanning Radiometer 2 (AMSR2), and thus the intercalibration of these two sensors can make their TB data more consistent and continuous to facilitate their joint applications. In this study, the FY-3D H-pol and V-pol TB at five frequencies from 10.7 to 89 GHz during 2019 to 2020 were calibrated against AMSR2 TB over land. Two categories of intercalibration methods were compared, including global intercalibration method, i.e., global linear regression, and per-pixel-based intercalibration methods, i.e., per-pixel linear regression joint global linear regression, per-pixel linear regression joint inverse distance interpolation, per-pixel linear regression joint nearest neighbor interpolation, and global per-pixel linear regression. Furthermore, the effects of diverse environmental variables (i.e., land cover and its heterogeneity, climate types, water body fraction, terrain and its complexity, soil texture, and vegetation coverage) on FY-3D calibration accuracy were fully investigated. The results indicate that all five approaches can reduce the bias between FY-3D and AMSR2 TB, and the root-mean-square difference (RMSD) also reduces accordingly. Among them, the global per-pixel linear regression method performs the best with the lowest averaged RMSD of 2.93 K (at ascending overpass) and 2.34 K (at descending overpass), followed by the per-pixel linear regression joint inverse distance interpolation. The global linear regression method performs the worst with the largest RMSD of 4.69 and 3.82 K at ascending and descending overpass, respectively. The RMSD is relatively larger in temperate and polar climate zones, as well as in grasslands and croplands than in other climate and land cover types. The calibration errors generally decrease as the altitude increases, while they increase with the increase in land cover heterogeneity. The water body fraction exerts the greatest impact on the calibration accuracy, and the RMSD reaches 3 K when the water body fraction is greater than 15%. Soil texture, terrain complexity, and vegetation coverage generally have little influence on the calibration accuracy. These findings can provide a good reference for the intercalibration of satellites with similar configuration to generate long-term climate data records.

The signal propagation of the lunar-based synthetic aperture radar (LBSAR) is affected by perturbations of the lunar orbit, wherein the apsidal precession that exerts a significant impact on the LBSAR imaging performance of the LBSAR deserves special care. Accordingly, the orbital determination used to maintain well-focused quality and high geometric fidelity in the existing SAR system becomes critical for the LBSAR. In this article, through establishing criteria for the orbital determination of LBSAR based on its imaging performance under the influence of apsidal precession, we investigate the accuracy requirements for the LBSAR orbital determination in terms of the position and velocity determinations. Analysis results show that the required accuracy for the LBSAR position determination depends on the geometric fidelity in the range direction, while the accuracy requirement for the velocity determination is dominated by the azimuth positioning accuracy. The focusing quality is not a primary issue for the LBSAR orbital determination. In addition, the far look angle of LBSAR accounts for the highest accuracy requirement in the position and velocity determinations; thus, it can be treated as the optimum look angle for the LBSAR orbital determination. It is also found that both velocity and position determinations are challenging in the${z}$-direction for the LBSAR.

The imaging performance of the lunar-based SAR (LBSAR) is susceptible to the orbital perturbation effects. In particular, the perturbations of orbital inclination and right ascension of ascending node (RAAN) could give rise to the temporally varying orbit drift of LBSAR and further lead to Doppler errors in the radio signal. As a result, the LBSAR image performance might be influenced by such effects. This study comprehensively probes into the phase error induced by perturbations of orbital inclination and RAAN, and its effects on the LBSAR imaging performance are further explored. It is found the LBSAR imaging performance in terms of focusing quality and geometric fidelity are affected by the perturbations of orbital inclination and RAAN, wherein the deterioration of focusing quality is closely associated with the synthetic aperture time. In this regard, the azimuth resolution on a decameter level is optimum for Earth observation of LBSAR with satisfactory image quality. Regarding the geometric fidelity, the accuracy requirement for the orbit determination of the LBSAR under perturbations of the orbital inclination and RAAN is proposed. The analysis results show that the LBSAR orbit determination in terms of the position and velocity determinations are most strenuous in the z-direction. Finally, point target responses are simulated to illustrate the preceding analysis.

We present a generalized power-law roughness spectrum to account for the spatial anisotropy effects on the radar scattering of a rough surface, where both the correlation anisotropy and the scaling anisotropy are accounted for. The spatial anisotropy is essential to correctly interpret the radar scattering from an agriculture field where both plow and sow are practiced. We investigate the dependence of the backscattering coefficient on the correlation anisotropy and the scaling anisotropy through a model simulation. A drastic change in backscattering strength is observed due to the anisotropy. The correlation anisotropy and the scaling anisotropy generate similar backscattering angular behavior, implying that in the context of spatial anisotropy, merely using correlation length in scattering modeling is insufficient. Equivalently, the correlation length retrieved from the backscattering coefficients perhaps is not unique. Fair use of the generalized anisotropic power-law roughness spectrum in conjunction with the scattering model is illustrated by comparing the backscattering coefficients with experimental measurements. However, the anisotropy complicates the roughness description in terms of surface parameters retrieval because we can generate similar backscattering angular patterns by combining different correlation anisotropy and scaling anisotropy. When the soil moisture is of primary interest, a more suitable radar observation geometry to minimize the spatial anisotropy influence is desirable.

The synergy of active and passive microwave measurements has attracted considerable attention in recent years since they offer complementary information on the characteristics of the observed target (e.g., soil moisture), which motivates the launch of NASA’s Soil Moisture Active Passive (SMAP) mission. An assumption of a near-linear relationship between active and passive measurements has been made in the SMAP active–passive baseline algorithm, which is essential to downscale coarse-resolution radiometer brightness temperature (TB) using high-resolution radar backscatter ($\sigma ^{0}$) but has not yet been fully tested under a wide range of ground conditions. Motivated by this, we first examined the validity of the linear assumption by using concurrent and coincident SMAP active and passive observations under diverse environmental factors (e.g., land cover, climate types, terrain and its complexity, soil texture, vegetation coverage, soil moisture, and its dynamics). We also adopted SMAP enhanced TB to evaluate the performance of the disaggregated TB at the same grid resolution of 9 km. The results reveal there is a generally good linear relationship between$\sigma ^{0}$(no matter in dB or in linear unit) and TB at a global scale. There is no significant difference in the correlation among the four polarization combinations ($\sigma ^{0}_{\text {hh}}$versus TBh,$\sigma ^{{0}}_{\text {hh}}$versus TBv,$\sigma ^{{0}}_{\text {vv}}$versus TBh, and$\sigma ^{{0}}_{\text {vv}}$versus TBv) with the$\sigma ^{{0}}_{\text {vv}}$and TBhcombination displaying an overall slightly higher correlation. The linear relationship between$\sigma ^{{0}}$and TB is significantly affected by environmental factors. Particularly in bare soils and densely vegetated areas (e.g., large forest fraction and vegetation coverage), and arid and polar climate zones, the linear correlation between active and passive measurements worsens, whereas it is favorable in moderate vegetation and soil moisture as well as large soil moisture dynamic conditions. Interestingly, the linear correlation generally decreases as sand content increases while increases with the increase of clay content. The absolute linear correlation coefficient is higher with larger soil moisture dynamics. When compared to SMAP enhanced TB, it shows the linear assumption may have more influence on the correlation (i.e., temporal evolution) of downscaled TB than its absolute accuracy. These findings can enhance the understanding of the geophysical relationship between radar and radiometer signatures, and thus benefit active–passive joint algorithms for future satellite missions.

The Tibetan Plateau, known as the “Third Pole” of the world, is extremely sensitive to global climate change. Reliable soil moisture information is essential for understanding the impact of the Tibetan Plateau on the Asian monsoon. The study assessed a total of four satellite soil moisture products, namely the SMAP-SCA (V7), ESA CCI (V0.52), AMSR2 LPRM (V001), and FY-3C (V1) over the Tibetan Plateau using three densely-instrumented soil moisture networks (i.e., Maqu, Naqu, and Pali). Moreover, the possible error sources of these products were thoroughly investigated over the Tibetan Plateau, which had not yet been fully explored in previous studies. The results reveal the ESA CCI (V0.52) and SMAP-SCA (V7) generally perform better in the three networks with higher correlation coefficient ($R$) and lower standard deviation of the difference (STDD) compared to other products. The bias of ESA CCI (V0.52) is demonstrated to be dependent on the GLDAS Noah soil moisture, but it correlates much better with soil moisture measurements and also displays lower STDD value than the GLDAS Noah. SMAP brightness temperatures demonstrate much higher sensitivity to soil moisture than the AMSR$2~C$-band and FY-3C$X$-band measurements regardless of vegetation, surface roughness, and land cover heterogeneity. The auxiliary surface temperature used in SMAP-SCA (V7) also performs better than that used in AMSR2 LPRM (V001) and FY-3C (V1) though it is slightly colder than the ground measurements. These factors contribute to the better performance of SMAP-SCA (V7) soil moisture product compared to other satellite datasets. The AMSR2 LPRM (V001) product produces evidently larger absolute values than ground observations, but it can track the temporal trend of soil moisture in sparsely vegetated areas (Naqu and Pali). The FY-3C (V1) product exhibits some abnormal saturation values in Maqu with the highest vegetation biomass, while it performs better in Naqu and Pali with relatively lower vegetation coverage. The surface temperature derived from AMSR2 LPRM and FY-3C shows large uncertainties over the Tibetan Plateau which may be caused by the limited data used to calibrate the empirical surface temperature models.

This paper presents a physics-based machine learning (ML) approach to estimate the direction-of-arrival (DOA) over sea surfaces. We designed a recurrent neural network (RNN) to accept two receiving channel radar data to ensure angular coherence between the receiving signals from different directions, given the scattered signal by rough surfaces having angular memory. Specifically, the radar scattering coefficients of sea surfaces were simulated at C-band and for both co- and cross polarizations; investigations show that the sea speckles significantly and negatively impact the dual receiver’s DOA estimation performance; to mitigate the speckle interferences, we further designed an optimal configuration of dual-channel observation for DOA estimations over the sea surface, i.e., the co-polar (CP) observation mode performed best compared to that of the co-azimuth (CA) and the full-bistatic (FB) observation mode, and at moderate sea speckle impacts, the root-mean-square error of CP observation mode was about 1° for incident angle and 5° for incident azimuth angle estimations, results demonstrate the superior performances of the proposed physics-based neural DOA estimator.

We illustrate the wave scattering from a multiscale rough surface modeled by a modulated correlation function. The modulation ratio, defined as the ratio of baseband correlation length and the modulated length, determines the degree of multiscale roughness. The dependence of bistatic scattering on the modulation ratio are investigated. Numerical results show that without considering the multiscale roughness, the scattering coefficients are overestimated at a small incident angle region but underestimated at a large scattering region. Radar wave scattering from multiscale rough surface is highly frequency selective. As an application example, we compare the model predictions with two independent sets of measurement data. The results demonstrate that the model predictions with modulation effects are in good agreement with the measurement data.

Emulsified oil slicks can not only damp short gravity and capillary waves on the sea surface, but also reduce the permittivity of the contaminated area. This paper simulates the backscattering coefficients of emulsion oil covered sea surfaces based on AIEM, with the damping model described by model of local balance (MLB). The sea surface covered by emulsion oil with finite thickness is modeled as a layered-medium to calculate the composite reflection coefficients. The simulation results of oil-covered sea surface are compared to those of clean sea surfaces and discussed in terms of incidence angles and frequencies of EM waves, oil thickness and wind speeds.

There are few studies on predicting fully bistatic scattering from the rough surface of Mars, though some bistatic radar observations have been made, such as in the MARS EXPRESS mission. To better understand the interaction of radar signals with a planetary surface in bistatic radar observations, the topographic-scale roughness of Mars, characterized by a two-dimensional power spectrum density (2D-PSD), is examined in view of its global roughness variations and scale dependence on geological units. The analysis shows that the Martian 2D-PSD is strongly dependent on the geological units and that it lies between Gaussian and exponential functions, with a power index equal to 1.9. The bistatic scattering coefficients are calculated by an advanced integral equation model (AIEM) with the 2D-PSD as the input. It shows that the specific surface roughness spectrum and the dielectric inhomogeneity should be taken into account in interpreting the bistatic radar scattering response.

There have been considerable interests in the lunar-based synthetic aperture radar (LBSAR) for monitoring large-scale geoscience phenomena. However, the signal distortions given rise by lunar orbital perturbations, especially the apsidal precession, are particularly severe in the LBSAR. The apsidal precession effects can induce a coordinate drift of the LBSAR, which can further lead to the variation in the range history of the LBSAR. As a result, LBSAR’s image performance might be affected. In this letter, we thoroughly investigate whether the apsidal precession effects cause the phase decorrelation in the signal of the LBSAR, and how such effects impact the LBSAR imaging. The theoretical result shows that the impact of the lunar apsidal precession mainly results in the first-order and second-order Doppler errors, which further influence the geometric location and focusing quality along the azimuth direction. Numerical simulations using the point target response show good consistency with the theoretical analysis. To this end, the lunar apsidal precession effects deserve special care in the LBSAR for high imaging quality.

Electromagnetic scattering and emission from 1-D rough surfaces with multiple elevations are studied using full-wave simulations. Both the root-mean-square (rms) heights and the surface length are large compared to the wavelength. A novel multilevel steepest decent-sparse matrix canonical grid (MLSD-SMCG) method is proposed to address limitations in the original SMCG. The uniform Nystrom method and neighborhood impedance boundary condition (NIBC) are also incorporated in solving the dual surface integral equations (SIEs) of the method of moments (MoM). Simulation results are illustrated at L-band for soil and ocean surfaces. The surface rms heights and lengths are up to 1.43 and 243.8 m corresponding to 6 and 1024 wavelengths at 1.26 GHz, respectively. For ocean surfaces, the wind speeds up to 20 m/s are considered, and the entire spectrum is included to capture all relevant surface length scales. Numerical results indicate the proposed approach is computationally efficient and accurate. Energy conservation checks in simulations are at $10^{-4}$ for ocean scattering and emission. Also, the effects of wind-driven roughness on ocean emissivity are further investigated using the proposed approach in terms of wind speed and observation angle for both polarizations.

Numerical simulation of random rough surface finds wide applications in scientific disciplines, e.g., radar remote sensing of terrain and sea. In scattering simulation of rough surface, not only energy conservation must be ensured, but also, perhaps equally important, the surface inherent properties must be preserved. However, the proper choice of surface and grid sizes that are statistically representative poses a problematic issue. This study applied the entropy measure to determine such parameter settings by examining the relative error of sample entropy associated with roughness parameters and by noticing the fact that a rough surface with certain roughness parameters, including power spectrum density function, must have unique sample entropy. It is found that if the two criteria are met, proper choice of surface length and grid size is attainable to warrant minimum uncertainties of rough surfaces and maximum information content for different roughness spectra density functions under different correlation lengths. The feasibility and superiority of the proposed entropy-based method are validated in terms of minimum error of roughness parameters and also the energy conservation in bistatic scattering coefficients of rough surfaces generated using obtained simulation parameters.

This article presents a numerical study of the wave scattering from a modulated rough surface characterized by a modulated correlation function that shows zero crossings along the lag distance. The zero crossings imply the presence of multiscale roughness. In such a way, we treat the rough surface to include a baseband correlation length and a modulation length. The modulation ratio, the ratio of baseband correlation length, and the modulated length determine the degree of multiscale roughness. We then examine the dependence of bistatic scattering, both in level and angular trends, on the modulation length. The results indicate that as the modulation ratio increases, the scattering coefficients reduce in the incident plane. Comparing the model predictions with two independent sets of measurement data demonstrates a good agreement of the scattering coefficients between the experimental data and the model predictions with modulation effects. The proposed modulated rough surface concept offers higher flexibility to model the radar scattering of multiscale rough surface.

This article presents a new index, polarization-conversion ratio (PCR) to characterize depolarized bistatic scattering from rough surfaces with dielectric inhomogeneity and spatial anisotropy. We then investigate the dependence of PCR on both surface and radar parameters. Numerical results show that the distribution of PCR on the scattering plane varies with the polarization state of the incident wave and incident angle. The PCR clusters more in the cross-plane for horizontally polarized incidence. However, for vertically polarized incidence, the PCR disperses as “triangular shape” on the whole scattering plane with a sharp valley occurring in the incident plane. The following points can be drawn: 1) the inhomogeneity effectively enhances the PCR in the cross-plane; 2) the effect of anisotropy on the PCR is relatively weak, because the scattering is less affected by correlation length; 3) the impacts of surface rms height on the PCR are negative on the whole scattering plane; and 4) as the background permittivity increases, at the horizontally polarized incidence, the PCR is enhanced in the backward and forward regions, while at vertically polarized incidence, it is enhanced in the incident plane and the forward region. As is demonstrated, the PCR is an effective measure of the sensitivity of depolarization, making it potentially useful as a new reliable index for surface parameter inversion.

This paper assessed the accuracy of four passive microwave (PM) sea ice concentration (SIC) products in polar regions by using twelve scenes MODIS images under clear-sky conditions. The SIC products include the DMSP SSMIS with Arctic Radiation and Turbulence Interaction Study Sea Ice (ASI) algorithm (SSMIS/ASI), the GCOM-W AMSR2 with NASA Bootstrap (BT) algorithm (AMSR2/BT), the Chinese Feng Yun-3B with enhanced NASA Team (NT2) algorithm (FY3B/NT2), and the Chinese Feng Yun-3C with NT2 (FY3C/NT2). An adapted optimal threshold method (i.e., the Otsu algorithm) was adopted to automatically classify the MODIS images into sea ice and water which were then aggregated to compare with the PM SIC. The results show that the averaged bias of PM SIC (PM SIC minus MODIS) is less than 6% in Arctic and ranges from -8% to 1 % in Antarctic, and the averaged root mean square error (RMSE) is less than 16% in the whole polar regions. Overall, the SSMIS/ASI product has better performance in Arctic, while the FY3/NT2 has lower bias and RMSE in Antarctic. Meanwhile, it is observed that the error metrics of PM SIC vary noticeably when compared to different MODIS images which may be caused by the diverse surface characteristics of different sea ice types.

The effectiveness of refined interferometric synthetic aperture radar (InSAR) filter (refined filter) was validated using both simulated and real interferometric data. However, the threshold as a key parameter in the refined filter is determined by repeated attempts, and the deterministic method has some disadvantages, such as strong subjectivity, long time consuming, and bad adaptation. This letter makes full use of the robustness of inverse distance weighting (IDW) to simplify the refined filter. In the proposed filter, using an 11 × 11 window and an improved IDW, the computational burden of I(ejψz)I for 16 windows per pixel is avoided, and the need for setting the threshold is eliminated. From the beginning, the initial window angle is determined using four preprocessing windows. The filtering direction of the center pixel is affected by the initial window angle of each pixel, the degree of homogeneity of each pixel, and the distance of the pixel from the center pixel in the 11 × 11 window. The improved weighting method is then used to further refine the window angle of the center pixel. Numerical experiments by simulated and real InSAR data were used to validate the proposed approach. By comparison, the proposed filter greatly improves the efficiency of the refined filter and exhibits the superiority in filtering performance compared to commonly used filters, particularly for regions of low coherence, high-coherence-gradient, and high-phase gradient.

Due to the earth and moon's revolutions, the Doppler centroid of the moon-based synthetic aperture radar (moon-based SAR) varies dramatically along the orbit of the moon when the SAR system is looking broadside (nonsquinted). However, the Doppler centroid should be kept as small as possible to avoid range ambiguity and to alleviate difficulty with focus. This letter presents the Doppler properties along the orbit of the moon in accordance with the antenna coordinate system. Based on the Doppler analysis and the phase scan, we propose a method for Doppler centroid steering to minimize the Doppler centroid frequency without rotating the platform. The new method can accurately compensate the Doppler centroid to zero, because it considers the effects of the lunar orbit and the relative motion between earth's target and moon-based SAR. To validate the proposed method, we also derived the lower and upper bounds of the look angle. Subsequently, we performed simulations in accordance with Jet Propulsion Laboratory Development Ephemeris 430 (JPL DE430). Finally, the performance requirement of the phase scan is analyzed so as to validate the proposed method.

The spatiotemporal coverage of a Moon-based synthetic aperture radar (SAR) is analyzed based on the imaging geometry, upon which the spatial coverage and image formulation rely. The distance from the Earth to the Moon-based SAR and bounds of the grazing and azimuthal angles jointly determine the coverage area on the Earth's surface. Meanwhile, the ground coverage of the Moon-based SAR is determined by the bounds of the grazing and azimuthal angles and geographic coordinates of the nadir point at a specified time. Moreover, the temporal variation in the spatial coverage is pertinent to the temporally varying nadir point of the Moon-based SAR on the Earth's surface. Furthermore, numerical simulations using the lunar ephemeris data are carried out to complement the analysis and to illustrate the spatiotemporal coverage. Finally, a guideline for the optimal site selection of a Moon-based SAR is proposed. In conclusion, a Moon-based SAR has the potential to perform long-term, continuous Earth observations on a global scale to enhance our capability to understand the planet.

This article attempts to examine the Brewster effect of incoherent scattering from a lossy inhomogeneous rough surface with a vertical dielectric profile. Five typical dielectric profiles are selected for the purpose of illustration. In calculating the reflection coefficients, a transition model is used to convert the angle of incidence to local incidence, which accounts for the surface roughness. Numerical results show that the Brewster angle gradually moves to the large incident angle with an increase in the background dielectric constant and in the surface root-mean-squared (rms) height. The angular dependence of reflection coefficients, both the level and the trend, is slightly affected by the correlation length. The scattering strength is much more sensitive to the rms height than to the correlation length. The results could be useful in the retrieval of vertical soil moisture content when proper radar observation is available.

Soil moisture is a pivotal hydrological variable that links the terrestrial water, energy, and carbon cycles. In this article, a new soil moisture (SM) index (SMI), which aims to capture the temporal variability of SM, irrespective of cloud cover and solar illumination, was developed by using the L-band SM active passive (SMAP) radiometer observations. The SMI was proposed on the basis of two key foundations: 1) vegetation and roughness have similar effects on “depolarization” of microwave emission, while SM enhances polarization differences and 2) vegetation and roughness generally impose positive effects on surface emissivity, while SM and emissivity are negatively correlated. Based on the two physical principles, it is possible to decouple the effects of SM and those of vegetation and surface roughness in a 2-D space independent of vegetation type and roughness condition. The proposed SMI was then validated byin situmeasurements from five dense SM networks covering different vegetation and climatic conditions and also compared with SMAP passive and European space agency climate change initiative (ESA CCI) SM products at a coarse resolution of 36 km, and SMAP-enhanced passive and Japan Aerospace Exploration Agency (JAXA) advanced microwave scanning radiometer (AMSR2) SM products at a medium resolution of 9 km. The results show that the new SMI is able to well reproduce the temporal dynamic of SM with a favorable averaged correlation coefficient value of 0.87 and 0.84 at 36 and 9 km, respectively, higher than that of SMAP passive (0.80), SMAP-enhanced passive (0.77), ESA CCI (0.69), and JAXA AMSR2 (0.53). After removing the systematic differences between satellite and site-specific SM data by using the cumulative distribution function (CDF) matching technique, the SMI can achieve an average root mean squared error (RMSE) of 0.031 and 0.036 m3m−3at 36 and 9 km during the validation period, respectively, lower than that of the satellite SM products. In addition to surface temperature, the SMI does not need any further information from other sensors [e.g., the optical normalized difference vegetation index (NDVI) or leaf area index (LAI) data] to guarantee an all-weather monitoring. Therefore, it has great potential to estimate SM variability on a global scale.

Land surface temperature (LST) is a key parameter for many fields of study. Currently, LST retrieved from satellite thermal infrared (TIR) measurements is attainable with an accuracy of about 1 K for most natural flat surfaces. However, over urban areas, TIR measurements are influenced by 3-D structures and their radiation that could degrade the performance of existing LST retrieval algorithms. Therefore, quantitative models are needed to investigate such impact. Current 3-D radiative transfer models are generally based on time-consuming numerical integrations whose solutions are not analytical, and are therefore difficult to exploit in the methods of physical retrieval of LST in urban areas. This article proposes an analytical TIR radiative transfer model over urban (ATIMOU) areas that considers the impact of 3-D structures and their radiation. The magnitude of this impact on TIR measurements is investigated in detail, using ATIMOU, under various conditions. Simulations show that failure to acknowledge this impact can potentially introduce a 1.87-K bias to the ground brightness temperature for street canyon whose ratio “wall height/road width” is 2, wall and road temperature is 300 K, wall emissivity is 0.906, and road emissivity is 0.950. This bias reaches 4.60 K if road emissivity decreases to 0.921, and road temperature decreases to 260 K. ATIMOU is also compared to the discrete anisotropic radiative transfer (DART) model. Small mean absolute error of 0.10 K was found between the models regarding the simulated ground brightness temperatures, indicating that ATIMOU is in good agreement with DART.

This paper presents an enhancement of the refined InSAR filter (refined filter). In the proposed filter, using an 11x11 window and an improved inverse distance weighting (IDW), the computational burden of |〈ejΨz〉|for 16 windows per pixel is avoided, and the need of setting the threshold is eliminated. From the beginning, the initial window angle is determined using four pre-processing windows. The filtering direction of the center pixel is affected by the initial window angle of each pixel, the degree of homogeneity of each pixel, and the distance of the pixel from the center pixel in the 11x11 window. Then an improved weighting method is applied to refine the window angle of the center pixel. Numerical experiments using both simulated and real InSAR data are carried out to validate the proposed approach. It is demonstrated that the proposed filter improves the efficiency of the refined filter and yet shows excellent filtering performance compared to commonly used filters, particularly for those regions of low coherence, high-coherence-gradient, and high phase gradient.

This paper evaluated the consistency of four microwave remotely sensed sea ice concentration (SIC) products with respect to a reanalysis SIC dataset in polar regions during the period of 2015-2017. The remotely sensed SIC products include the Chinese Feng Yun-3B with enhanced NASA Team (NT2) sea ice algorithm (FY3B/NT2), the Chinese Feng Yun-3C with NT2 (FY3C/NT2), the DMSP SSMIS with Arctic Radiation and Turbulence Interaction Study Sea Ice (ASI) algorithm (SSMIS/ASI), and the GCOM-W AMSR2 with NASA Bootstrap (BT) algorithm (AMSR2/BT). The OISSTV2 (NOAA Optimum Interpolation 1/4 Degree Daily Temperature Analysis Version 2) dataset was adopted as a reference to compare and evaluate the performance of four satellite-based SIC products. The results show that remotely sensed SIC values are generally in good consistency with OISSTV2. Meanwhile, it is observed that different products have different bias in polar regions. Overall, the SSMIS/ASI product has better performance during the whole period, demonstrating that ASI algorithm may be more potential in SIC estimation. Our results also illustrate the spatial and temporal distribution characteristic of discrepancy between microwave remotely sensed SIC products and reanalysis dataset for the whole Arctic and Antarctic regions. The large difference for all the four SIC products mostly occurs in summer and marginal ice zone, indicating a great deal of uncertainty of satellite SIC products in this period and areas. The results will be useful to find possible errors in the satellite SIC products for further algorithm improvement.

Radar wavelength and observing geometry together determine the radar probing spectral width within which the Bragg resonance occurs. In this paper, we present numerical study of radar bistatic scattering from sea surface, and analyze the performance of bistatic radar on observing ocean surface. Bistatic radar in sensing the wave number range, wind speed, and wind direction are comprehensively discussed for general bistatic geometry. Moreover, the frequency dependence is also demonstrated. The results show that, compared to the mono-static radar, bistatic radar exhibits wider covering of the spectrum width, and can better catch the multi-scale marine dynamic process. In addition, bistatic radar, with receiver a slight shift off the specular angle, can capture the scattering of wind-generated high-frequency components of wave spectrum. Therefore, it is expected that bistatic observation provides greater sensing capability of wind field.

This paper presents full polarization wave scattering from a spatially anisotropic rough surface with inhomogeneous dielectric profile modeled by a transitional layer as a function of depth. The spatial anisotropy is described by directional correlation function. In this study, both linearly and circularly polarized scattering are investigated in light of azimuthal dependence, degree of anisotropy, and inhomogeneity. Numerical results show that the backscattering at small scattering angle exhibits strong dependence on degree of anisotropy, but turns to a strong azimuthal dependence at larger scattering angle. The bistatic scattering in the incidence plane reveals strong dependence on degree of anisotropy near the specular direction. In virtue of dielectric inhomogeneity, the scattering pattern reveals several distinct features including: HH polarization increases significantly in the backward region but decreases slightly in the forward region; the forward and backward scattering of VV polarization are enhanced; HV polarization can be greater than VH polarization; both LR and RR polarizations on the whole scattering plane are enhanced. Furthermore, backward scattering is stronger at 0°or 180° of azimuthal angle, but at 90° of azimuthal angle the forward scattering is stronger.

In this paper, we compare the imaging performance for the point target obscured by random media of three commonly used techniques, namely: synthetic aperture radar (SAR), time reversal (TR), and time reversal-multiple signal classification (TR-MUSIC). For quantitative evaluation, both 3-dB beam width and geometric accuracy of a point target response are used. The results of three methods are all degraded by the presence of random media, and become poorer with denser media in which the effects of the scattering thickness is the main factor causing imaging degradation. Among the three techniques, TR-MUSIC offers the best imaging performance and suppress the grating lobes under a sparse array. Taking advantage of this, in TR-MUSIC we use a special array to enhance the imaging performance and to reduce the complexity of the radar system.

In this paper, a new soil moisture index (SMI), which aims to capture the temporal variability of soil moisture, was developed by using the L-band SMAP radiometer observations. This index is proposed on the basis of two key foundations: 1) vegetation and roughness have similar effects on "depolarization" of microwave emission, while soil moisture enhances polarization differences; 2) vegetation and roughness generally impose positive effects on surface emissivity, while soil moisture and emissivity are negatively correlated. Based on the two physical principles, it is possible to decouple the effects of soil moisture and those of vegetation and surface roughness in a two-dimensional space independent of vegetation type and roughness condition. The proposed SMI was then validated by in-situ measurements from five dense soil moisture networks covering different vegetation and climatic conditions, and also compared with SMAP and ESA CCI official soil moisture products. The results show that the new SMI can well reproduce the temporal dynamic of soil moisture with a very favorable averaged correlation coefficient value of 0.87, higher than that of SMAP (0.80) and ESA CCI (0.69). The unique advantage of the proposed SMI is that it does not need field observations of soil moisture, roughness, or canopy biophysical properties for calibration purposes, and does not involve any empirical coefficients. Therefore, it has great potential to estimate soil moisture variability on a global scale.

Active and passive microwave signatures respond differently to the land surface and provide complementary information on the characteristics of the observed scenes. The objective of this paper is to explore the synergy of active radar and passive radiometer observations at the same spatial scale to constrain a discrete radiative transfer model, the Tor Vergata (TVG) model, to gain insights into the microwave scattering and emission mechanisms over grasslands. The TVG model can simultaneously simulate the backscattering coefficient and emissivity with a set of input parameters. To calibrate this model, in situ soil moisture and temperature data collected from the Maqu area in the northeastern region of the Tibetan Plateau, interpolated leaf area index (LA!) data from the Moderate Resolution Imaging Spectroradiometer LAI eight-day products, and concurrent and coincident Soil Moisture Active Passive (SMAP) radar and radiometer observations are used. Because this model needs numerous input parameters to be driven, the extended Fourier amplitude sensitivity test is first applied to conduct global sensitivity analysis (GSA) to select the sensitive and insensitive parameters. Only the most sensitive parameters are defined as free variables, to separately calibrate the activeonly model (TVG-A), the passive-only model (TVG-P), and the active and passive combined model (TVG-AP). The accuracy of the calibrated models is evaluated by comparing the SMAP observations and the model simulations. The results show that TVG-AP can well reproduce the backscattering coefficient and brightness temperature, with correlation coefficients of 0.87, 0.89, 0.78, and 0.43 and root-mean-square errors of 0.49 dB, 0.52 dB, 7.20 K, and 10.47 K for σHHo, σVVo, TBH, and TBV, respectively. In contrast, TVG-A and TVG-P can only accurately model the backscattering coefficient and brightness temperature, respectively. Without any modifications of the calibrated parameters, the error metrics computed from the validation data are slightly worse than those of the calibration data. These results demonstrate the feasibility of the synergistic use of SMAP active radar and passive radiometer observations under the unified framework of a physical model. In addition, the results demonstrate the necessity and effectiveness of applying GSA in model optimization. It is expected that these findings can contribute to the development of model-based soil moisture retrieval methods using active and passive microwave remote sensing data.

Microwave backscattering models play key roles in surface scattering modeling and soil moisture inversion in active microwave remote sensing. However, numerous evaluations indicate that significant discrepancies between the model simulations and radar observations remain, and these discrepancies are regarded to be attributed to inaccuracies in the models. What do such discrepancies originate from is unclear and has not been comprehensively analyzed. To this end, this paper presents an uncertainty analysis to explore the intrinsic reason for the discrepancies between the backscattering model simulations and radar observations. The probability distribution function and the corresponding statistical characteristics are introduced to describe the uncertainty in the model outputs. We find that the scale dependence of the key model inputs leads to significant uncertainties in the model inputs, and the uncertainties are transferred into the model outputs. Thus, the discrepancies between the model simulations and radar observations are intrinsically caused by the spatial scaling and related uncertainties of key model inputs. In short, the scale mismatch between the model inputs and remote sensing pixels is an intrinsic factor that causes the discrepancies between the model simulations and radar observations. This finding suggests that the scaling effect of model inputs should be carefully considered when using the backscattering models at the pixel scale, and equivalent inputs matched at the corresponding scales should be developed for remote sensing applications. Thus, this analysis insights into the scale dependence of inputs for backscattering models and suggests to provide scale-matched inputs where the models are applied at different scales.

Wind wave spectrum describes the quasi-periodic nature of the ocean surface oscillations and plays an indispensable role in the study of microwave electromagnetic scattering from sea surface. A reliable spectrum model suitable for radar cross section (RCS) predictions at different radar frequencies is desired. This paper evaluated the performances of five common spectrum models (i.e., Fung spectrum, Durden-Vesecky spectrum, Apel spectrum, Elfouhaily spectrum, and the newest version of Hwang spectrum, H18) on the normalized radar backscattering cross section (NRBCS) simulations based on advanced integral equation model (AIEM) at L-, C-, X-, and Ku-bands versus incidence angle, wind direction, and wind speed by comparing with the model and measured data for validation. These results indicate no single wave spectrum of them is satisfying for all the four radar frequencies, e.g., Apel and H18 spectra are better for L- and C-bands, Apel spectrum for X-band, and Elfouhaily and H18 spectra for Ku-band. Given this, three average composite spectrum models are constructed using different spectral models (i.e., all five spectra, Apel + Elfouhaily + H18, and Apel + H18) to simulate NRBCSs, similar to that of the individual spectrum model. It is concluded that the combination of Apel and H18 spectra overall performs best among the individual one and other composited spectra in like-polarized NRBCSs versus incidence angles, wind directions, and wind speeds, for wind speed greater than 30 m/s where the combination of the five spectra work well at Ku-band.

In this paper, effects of the earth's curvature and lunar revolution on the performance of the moon-based synthetic aperture radar (SAR) are examined by a comprehensive analysis of the motion-induced Doppler frequency and Doppler rate on which azimuthal imaging relies. The motion effects include the earth's self-rotation related to the earth's curvature and lunar revolution around the earth. An extended hyperbolic range equation (EHRE) is proposed in line of the equivalent velocity and equivalent squint angle, and then the signal model based on the EHRE is established to simulate the moon-based SAR image from which the imaging performance is analyzed. Theoretical analyses show that the earth's curvature is a dominant factor in determining the moon-based SAR's Doppler and azimuthal resolution. Furthermore, the earth's curvature distorts the SAR image by way of rotating the azimuth imaging from the cross-range direction within a certain skewed angle. The overall effects of lunar revolution generate a velocity correction factor and a deviate squint angle, which subsequently deteriorate the azimuthal resolution and image focusing. Results also show that effects of the earth's curvature and lunar revolution are in connection with relative positions of the ground target and moon-based SAR. To this end, numerical simulations using point target response is carried out to accentuate the necessary for taking account of the Doppler error induced by the lunar revolution.

This paper examines the properties of bistatic scattering from an inhomogeneous rough surface, which, in this paper, is modeled by the transitional layer as a function of depth. The lower medium of the rough surface is horizontally uniform but vertically inhomogeneous. Both linear and circular polarizations are investigated in light of the dependences of transition rate, background dielectric constant, and surface roughness. The presence of dielectric inhomogeneity generally leads to several features that do not appear in the homogeneous surface, such as the scattering coefficient on the whole scattering plane is enhanced; the dynamic range of HH and VV over the azimuth plane is reduced; HV can be greater than VH; and the difference of LR and RR is decreased. With the increasing transition rate, the scattering coefficients for both the linear and circular polarizations are enhanced. As the background dielectric constant increases, the scattering responses of the linear and circular polarizations are quite different. For the linear polarization, HH exhibits a stronger angular dependence; VV reduces in the forward region and enhances notably in the backward region; and HV decreases but VH increases. For circular polarizations, the cross-polarized LR increases in the backward region but decreases in the forward region, and the copolarized RR enhances on the whole scattering plane. With the increasing surface roughness, the scattering coefficient becomes more evenly distributed over the entire scattering plane.

This paper examines the polarized backscattering of spatially anisotropic rough surfaces. To better explore the physical mechanisms that control the azimuthal dependence of the backscattering from anisotropic surfaces, the effects of surface roughness [correlation length and root-mean-square (rms) height], dielectric constant, and radar parameters from anisotropic surfaces are studied. The advanced integral equation model (AIEM) is used to simulate both co- and cross-polarized backscattering coefficients, including the single and multiple scattering. Numerical results suggest that the multiple scattering exhibits a stronger azimuthal dependence for HH than VV polarization, especially more so at a larger incident angle. For weakly anisotropic surface, the azimuthal variation of backscattering tends to be a sinusoidal-like pattern. However, with the enhancement of anisotropy, such a scattering pattern is distorted, and the sharp dip appears at up/down direction. As the rms height and dielectric constant increase, the scattering is enhanced on the whole. The HH/VV ratio at lower dielectric constant is greater than that at higher one. In comparison, scattering shows stronger dependence on anisotropy at lower dielectric constant, especially at a larger incident angle. As an application example, we compare the model predictions with reported measurements from two different sites. Preliminary results are quite encouraging, and thus, the analysis presented in this paper is potentially useful to predict and interpret backscattering from crop field surface, where strong anisotropic surfaces commonly present due to plowing or raking practice.

We analyze the error caused by `stop-and-go' assumption on Moon-based SAR image focusing. In Moon-based SAR imaging of earth, the extremely long propagation time delay and curved trajectory violate the `stop-and-go' assumption. The separation of transmitting and receiving antenna in Moon-based SAR is not regarded as monostatic, but quasi-bistatic mode. The Doppler parameters should be estimated under this condition. Preliminary analysis shows that error caused by stop-and-go assumption can seriously affect the focusing of Moon-based SAR imaging, and should be corrected or compensated.

This paper presents bistatic scattering from rough surface with an inhomogeneous layer where the permittivity is a transitionary function of depth. For reference, both homogeneous and inhomogeneous media are discussed in co- and cross polarization with various background dielectric constant and transition rate. Results show that the angular curves become more slowly in azimuthal direction in bistatic scattering, where the minimum dips are smearing out compared to those in homogeneous layer. The azimuthal dip appearing in the case of homogeneous surface starts to vanish for inhomogeneous surface, and a valley-like azimuthal curve is observed in both HH and VV polarizations, though the valley angles are different. To a great extent, the dynamic range of the scattering coefficient is reduced, particularly for VV polarization. Moreover, as for inhomogeneous medium, the HV polarized scattering coefficients are greater than VH polarization, which is opposite to the homogeneous medium. But for larger background dielectric constant, HV- and VH- polarized scattering coefficients look more symmetrical and VH polarization is relatively larger.

This paper examines the quality of eight satellite soil moisture products at two typical spatial resolutions, including an intercomparison of SMAP passive, SMOS, JAXA AMSR2, LPRM AMSR2, ESA CCI and the Chinese FY3B soil moisture products at a coarse resolution of ~0.25°, and the newly released SMAP enhanced passive and JAXA AMSR2 soil moisture products at a medium resolution of ~0.1°. Insitu measurements from two representative dense networks, i.e., the Little Washita Watershed (LWW) in the United States and the REMEDHUS networks in Spain are used to compare and validate the eight soil moisture products. The results show that the SMAP passive and FY3B products outperform the other products with the lowest unbiased root mean square (ubRMSE) values of 0.027 m3m-3and 0.025 m3m-3in the LWW and REMEDHUS network regions respectively. SMOS slightly underestimates soil moisture with a dry bias, but it correlates well with insitu data with an average correlation value of 0.77. The JAXA product performs much better at 0.25° than at 0.1°, but both of them underestimate soil moisture (bias>-0.05 m3m-3) at most time. The SMAP enhanced passive soil moisture well captures the temporal variation of ground measurements with a correlation coefficient larger than 0.8, and is generally superior to the JAXA product. The LPRM shows much larger amplitude and temporal variation than the ground soil moisture with a wet bias larger than 0.09 m3m-3. The ESA CCI product shows satisfactory performance with acceptable error metrics (ubRMSE3m-3), revealing the effectiveness of merging active and passive soil moisture products. The good performance of SMAP and FY3B demonstrates the potential in integrating them into the existing long-term ESA CCI product to form a more complete product.

In this letter, we investigate volume-scattering effects on microwave emission from a foam-covered ocean surface by a comparative study of emissivities and polarization indexes of a numerical radiative transfer model and an incoherent emission model. The matrix doubling method is applied in the numerical model, which fully accounts for the multiple scattering in the foam layer and at the interfaces. The incoherent emission model considers incoherent interactions between air-foam and foam-seawater interfaces but ignores the volume scattering in the foam layer. Model analyses show that foam volume scattering affects both the magnitude and polarization index of emission from the foam-covered ocean surface. Foam volume scattering reduces the rate of increase in emissivity with increasing foam water fraction. Results also show that polarization index is more sensitive to volume scattering at a large observation angle and at a low foam water fraction. Comparison between model predictions and experimental measurements accentuate the need to account for foam-scattering effects in interpreting emission measurements from the foam-covered ocean surface.

This letter explores the optimal bistatic radar configurations for bare soil moisture retrieval at L-band using a global sensitivity analysis method, the extended Fourier amplitude sensitivity test (EFAST) algorithm. Complete sets of bistatic scattering, covering a wide range of geometric parameters and ground surface conditions, are simulated by the well-established advanced integral equation model. The sensitivity of radar bistatic signals to soil moisture and surface roughness, and the interactions among the parameters are quantified using the EFAST algorithm. The results show that in bistatic scattering, VV polarization has notably higher sensitivity to soil moisture than HH polarization, particularly at large incident angles. For VV polarization, as incident angle increases, the sensitivity zone of soil moisture expands and shifts toward the forward direction, specifically at small azimuth scattering angles and large scattering angles, thereby becoming promising configurations for soil moisture retrieval. For HH polarization, in contrast, the sensitive zone gradually moves to the backward direction as incident angle increases, and an intermediate incident angle (e.g., 40°) is recommended for retrieving soil moisture by considering both sensitivity strength and parameter interaction effects.

The newest soil moisture-dedicated satellite, the Soil Moisture Active Passive (SMAP) mission, provides global maps of soil moisture using concurrent L-band radar and radiometer acquisitions. To support the ongoing validation activities of SMAP soil moisture products, in this paper, we examined the retrieval accuracy of four SMAP soil moisture products by using well-calibrated and dense in situ measurements from the Little Washita Watershed network, one of the SMAP core validation sites with intensive ground sampling. The four SMAP products include the active (3 km), passive (36 km), active-passive (9 km), and the enhanced passive product which is a newly released soil moisture data set with a grid resolution of 9 km. Efforts on identifying the possible error sources of these products were also made for the purpose of improving the SMAP soil moisture algorithms. The results show that the passive and active-passive products can well capture the temporal dynamic of ground soil moisture with overall unbiased root-mean-square error (ubRMSE) values of 0.032 and 0.041 m3· m-3, respectively, which generally meet their mission requirement of 0.04 m3· m-3. In contrast, some irregular fluctuations exist in the active product, leading to an overall wet bias, which makes its accuracy a little poorer than its expected retrieval accuracy of 0.06 m3· m-3. The new enhanced passive product shows the lowest ubRMSE value of 0.026 m3· m-3though it underestimates in situ measurements with a bias of 0.059 m3· m-3, revealing its great potential to substitute the active-passive product to provide global soil moisture measurements at a medium resolution of 9 km. The underestimation of SMAP surface temperature data may be one of the reasons that contribute to the dry bias of SMAP passive, active-passive, and enhanced passive products. The microwave polarization difference index and HV-polarized backscatter show good response to in situ soil moisture and may be considered in SMAP algorithms to further improve the accuracy of soil moisture retrievals. We expect that our findings can be fed back to improve the SMAP soil moisture algorithms and thus promote the application of SMAP soil moisture products in terrestrial water, energy, and carbon cycles.

A semi-empirical model is proposed by merging an improved directional spectrum into the advanced integral equation method (AIEM) in this paper. By establishing a new angular spreading function (ASF), this improved directional spectrum provides a better description of wave directionality, especially over wavenumber range from short-gravity waves to capillary waves. Based on this scattering model, the features of L-band ocean surface backscatter are studied. A unique negative upwind-crosswind (NUC) asymmetry of L-band ocean backscatter over a low wind speed range that was recently observed is interpreted and simulated first. The model is validated against Aquarius/SAC-D observations at the L band and the geophysical model functions (GMFs) at higher frequency band. The simulation results have good agreements with observations and the GMFs for different incidence angles, azimuth angles and wind speeds.

Numerous works show the great potential of microwave remote sensing in assessing biophysical variables of rice plants. However, most studies are focusing on the backscattering character of rice canopy. Comparatively, much lesser attention has been devoted to modeling bistatic scattering behavior of rice canopy. Therefore, this study aims to improve the understanding of bistatic scattering response of rice canopy, and thus help design a bistatic radar system for better monitoring rice growth and yield estimation. Firstly, a bistatic scattering of a cluster of rice plant was modeled by solving the vector radiative transfer equations. Then, the concept of antenna array is applied to account for the inter-cluster wave interactions to obtain the total scattered field and bistatic scattering coefficient. The results show that the coherent scattering of clusters is found to be a function of vegetation growth stages. Scattering coefficient corrections need to be considered, especially at initial growth stages, and in the forward scattering direction.

It has been shown that the global positioning system (GPS) signal scattering for the like-wise polarization (right-hand to right-hand circular polarization (RR)) yielded 20-30 dB lower level than that for the counter-wise one (left-hand to right-hand circular polarization (LR)), thus the counter-wise polarization was preferred in application to remote sensing by GPS signal scattering. In this paper we investigate the bistatic scattering from both double-layered and single sastrugi surfaces at C-band (5 GHz) at different azimuthal angles. Comparisons of bistatic scattering are made between like-wise type (LL, RR) and counter-wise type (RL, LR). Numerical simulations show that the conclusions of these past theoretical and numerical tests are not always valid for randomly sastrugi surfaces. The like-wise polarized scattering is comparable to the counter-wise one, and is even larger than counter-wise at some azimuthal angles.

We investigate the ionospheric effect on the global change observation lunar-based synthetic aperture radar (GCOLB-SAR), which offers an unprecedented temporal resolution and spatial coverage. However, the GCOLB-SAR is confronted with a great challenge: an extra-long distance electromagnetic wave propagating through atmospheric and ionospheric effects. The ionospheric effect for the satellite-borne SAR is no longer applicable for the GCOLB-SAR due to its ultra-long synthetic aperture time. In this paper, we consider the temporal-spatial variation effects of the ionosphere on the imaging of GCOLB-SAR. For the purpose of the study, an L-band GCOLB-SAR observation is considered. The signal model based on curved trajectory for GCOLB-SAR is derived and the temporal-spatial variation background ionosphere is analyzed. Preliminary analysis shows that the range shift is on the order of hundreds of meters, while the azimuth shift is of tens of meters. Such geometric deviations apparently should be corrected or compensated for GCOLB-SAR.

This paper presents a multiple scattering in the AIEM model. The derivation is mathematically intricate, though, the final expression is compact and easy for numerical implementation, which only involves series of two-dimensional integration. Some of special cases in backscattering are derived and compared with known analytical model to partly validate the update AIEM model in cross polarization. Finally, comparisons with numerical simulation and field measurements are made to show the model performance and accuracy, particularly in cross-polarized backscattering.

Recently the bistatic radar systems have seen increasing attention and development for the advantages in remote sensing of ocean surfaces. With the considerable merits of spatial diversity, bistatic systems supplement the retrieval of ocean physical parameters with conventional monostatic systems. For instance, an emerging bistatic radar technique, global navigation satellite signal reflectometry (GNSS-R), has been developed and utilized in retrieval of high wind speed. Moreover, bistatic phenomena such as the Brewster effect can reveal target properties that are not revealed clearly in monostatic scattering [1]. Therefore, for better application of ocean microwave remote sensing, understanding of bistatic scattering from the ocean surface is crucial and meaningful.

The synergy of active and passive microwave measurements have attracted increasing attention in recently years. In this study, we investigate the relationship and covariation of the SMAP radar backscatter and radiometer reflectivity as a function of surface roughness and vegetation. Two radar-derived indices, namely the radar vegetation index (RVI) and radar roughness index (RRI) are adopted to account for the contributions from vegetation and surface roughness respectively. The results show RVI distinguishes vegetation density well in sparse to densely vegetated regions, while significantly overestimates the biomass over some dry desert regions due to possible soil volume scattering effects. RRI well captures the negative covariation of active and passive measurements in bare and sparsely vegetated surfaces, while becomes ineffective in densely vegetated areas due to the reduced contribution from soil surfaces.

This paper presents a comprehensive reappraisal of the scattering, both backscattering and bistatic scattering, and emission of rough soil surface predicted by a well-established theoretical model, the advanced integral equation model (AIEM). Extensive numerical data simulated by approximate and numerically exact models as well as experimental datasets of well-characterized bare soil surfaces were used to evaluate the performance of AIEM in predicting the scattering coefficient and microwave emissivity over a wide range of geometric parameters and ground surface conditions. The results show that AIEM predictions are generally in good consistency with both of numerical simulations and experiment measurements in terms of angular, frequency and polarization dependences, except for some deviations in a few cases (e.g. at large incident angles and dry soil conditions). Possible explanations for the discrepancy between model prediction and data are given, together with suggestions for model usage and refinements.

In this letter, we investigate the circularly polarized bistatic scattering from sastrugi snow surfaces at the L-band (1.575 GHz) at different azimuthal angles. Comparisons of circularly polarized bistatic scattering are made between the likewise (left-hand circularly polarized transmitting and left-hand circularly polarized receiving or right-hand circularly polarized transmitting and right-hand circularly polarized receiving) and the counterwise (RL or LR). Numerical simulations show that the counterwise configurations are only preferred, in the sense of maximum received power, for randomly sastrugi surface, but not always so when the surface structure is double-layered, at which the likewise polarized scattering strength is comparable to, and at certain azimuthal angles it is even larger than, that of the counterwise polarized scattering. Physical mechanism of such a behavior is explicated by the phenomena of local Fresnel reflections. Results from this letter offer physical insights for bistatic sensing of layered snow surface. It is suggested that for microwave remote sensing of snow surface by bistatic signals (e.g., global positioning system), both transmitting and receiving polarizations in likewise and counterwise be adopted.

We investigate the magnitude of Sun-glint through modeling the Soil Moisture Active Passive (SMAP) brightness temperature (BT). Model results show that the specular reflection of Sun-glint in the L-band can spread over a wide range of view angles due to the roughness and undulation of the land surface and therefore affect SMAP radiometer observations. Due to SMAP's low incidence angle (40°), Sun-glint in the specular direction is never observed, and only the noncoherent component of Sun-glint has influence on SMAP observations. Sun-glint is particularly an issue over wet soil surfaces at low solar zenith angles (SZAs), and caution has to be taken for the terrain effect even for high SZAs, because the local solar incident angle can be significantly changed by the terrain slope and then the specular reflection of Sun-glint can be viewed by SMAP. Model results also show that BT in V-pol is less contaminated by Sun-glint than that in H-pol. During an intense solar radio burst, it was found that the land surface BT in H-pol increased by 50 K in the forward scattering direction from the SMAP observation. This is roughly equivalent to 1 K increment by every 100 solar flux units on a dry soil and/or dense vegetation. When the solar activity is quiet in 2015, the Sun-glint from both wet land and ocean surfaces can reach up to 10 K in the SMAP L1B BT product. This paper suggests that BT observations around the solar specular direction should be masked for soil moisture retrieval at the L-band.

In this paper, an attempt is made to investigate the angular signatures of bistatic scattering, in the azimuthal direction, from rough surfaces, with the aim of deepening our understanding of the bistatic scattering behaviors and exploring its potential applications. Three distinct angular features, dip angle, scattering strength, and angular width, as a function of the surface roughness and dielectric constant, are identified. Brewster's scattering, and its role in angular behavior, is examined at limited extent. Results reveal that the angular features strongly correlate with the surface parameters and scattering geometry. For small scattering angle, dip angle and width are independent of surface roughness. Comparatively, for larger incident and scattering angles, beyond 50°, the dip angle and scattering strength are sensitive, simultaneously, to rms height and dielectric constant, while the dip width only responses to rms height. Dips, induced by Brewster's scattering effect, not only shift in the polar direction, but also in the azimuthal direction, and are strongly dependent on surface parameters and bistatic geometry. Increasing the surface roughness or, equivalently, the incident angle tends to promote the disappearance of dips. The main contributions of this paper can be summarized as follows: 1) quantitative description of dip features, including angle, scattering strength, and angular width; 2) comprehensive characterization of the dip features and their dependence on surface parameters and bistatic geometries; and 3) limited investigation of the behavior of Brewster's scattering-induced dip.

Sediments on exposed intertidal flats are very dynamic and perform vital ecosystem functions. This paper proposes a new classification scheme for mud and sand flats on intertidal flats using fully polarimetric synthetic aperture radar (SAR) data. Freeman-Durden (FD) and Cloude-Pottier (CP) polarimetric decomposition components as well as double bounce eigenvalue relative difference (DERD) are introduced into the feature sets instead of the original intensity polarimetric channels. Classification is carried out using the random forest (RF) theory, and the results are evaluated using confusion matrices, kappa coefficients, and RF variable importance indices. Three study sites with different environmental conditions are chosen to demonstrate the effectiveness of the proposed classification chain. To further assess the performance of the proposed feature set, we set different feature combinations and process with the same processing chain. Results show that the DERD parameter can detail the sediment mappings on exposed intertidal flats and is a useful SAR feature to distinguish mud and sand flats efficiently. The combined FD and CP components have the ability to describe the polarimetric characteristics of sediments more correctly than the commonly used original intensity channels. Meanwhile, the RF theory shows great potential in distinguishing sediments in intertidal zones accurately and time efficiently.

In this paper, a physically based numerical electromagnetic approach, by solving Maxwell's equations, is developed to investigate the scattering and the emission from double-layered media, where both the top and bottom interfaces are random sastrugi surfaces with a random horizontal shift between the two corresponding negative-slope facets of both interfaces. Numerical simulations are illustrated for bistatic scattering and four Stokes parameters at L-, C-, X-, and K-bands. Results indicate that the bistatic scattering and Stokes parameters are asymmetric for the double-sastrugi media, whereas for the other two structures, sastrugi surface alone and sastrugi surface with a planar bottom boundary, their scattering and the first two Stokes parameters are even symmetric relative to the azimuthal angle of 90°, and the third and fourth Stokes parameters are odd symmetric. In particular, the Stokes results no longer have the strong coherent fluctuations in angular variations shown in periodic double-layered structures because the random double-layered structure eliminates the coherent interference. It is interesting to observe that the internal total reflection may cancel if the coupled interactions between the two sastrugi interfaces are very strong, which results in decreasing scattering at the L-band, and its Stokes parameters are similar to those of the sastrugi alone. Numerical results also reveal that the maxima of cross-polarized specular scattering from double-sastrugi structures can be observed at cross incidence; however, they are at a deep dip for the latter two statistical symmetric structures. The sastrugi-sastrugi structure, being geometrically anisotropic, is capable of generating strong cross-polarized scattering and, subsequently, significant amounts of the third and fourth Stokes. The azimuthal patterns, at a viewing angle of 55°, of four Stokes parameters, although more complex, are feature rich where two striking extrema of third and fourth Stokes are presented. Compared with the L-band, the C-band presents strong azimuthal dependence of four Stokes. The simulation results offer deeper insights into the scattering and emission process in sastrugi surface and may lead to better retrieval of surface parameters from radar or radiometric measurements.

Theoretical modeling plays a significant role as forward and inverse problem in active and passive microwave remote sensing. Understanding the validity and limitations of the models is essential for model refinements and, perhaps more importantly, model applications. Motivated by these, this paper presents a comprehensive analysis of the scattering, both backscattering and bistatic scattering, and emission of rough soil surface predicted by the advanced integral equation model (AIEM), a well-established theoretical model. Numerically simulated data, covering a wide range of surface parameters, and in situ measurement data set of well-characterized bare soil surfaces were used to evaluate the performance of AIEM in predicting the scattering coefficient and microwave emissivity over a wide range of geometric parameters and ground surface conditions. The results show that the AIEM predictions are generally in good consistency with both numerical simulations and experiment measurements in terms of angular, frequency, and polarization dependences, except for some deviations in a few cases (e.g., at large incident angles and dry soil conditions). Extensive comparison confirms the effectiveness and practicability of AIEM for both scattering and emission of rough soil surface. Possible explanations for the discrepancy between the model prediction and data are given, together with suggestions for model usage and refinements.

In this paper, we study the polarimetric properties of bistatic scattering from a double-layered random surface in which both interfaces are randomly irregular, and a horizontal shift exhibits between the two interfaces. Such a surface structure is commonly known as Sastrugi surface found in vast area of Greenland. A physically based numerical approach solving the Maxwell's equations is developed to study the polarimetric scattering for which the surface integral equations governing the surface fields are formulated and solved to compute the scattered fields from which the scattering matrix and coherency matrix can be obtained. We then examine the properties of coherency matrix including the reflection symmetry, rotational symmetry, and azimuthal symmetry, along with polarimetric statistics. Results from this study prove useful for the polarimetric remote sensing of snow in Greenland.

Forest ecosystem plays an important role in regulating the global climate by serving as the primary carbon pool for atmospheric carbon dioxide. Deforestation and forest degradation could pose a serious threat on the global emission of greenhouse gases, and this has been declared as the critical issue for United Nation's REDD programme. Under the threat of global warming and climate change, it is critical to quantify the mechanism of carbon flux in order to locate the so-called missing carbon sink and to identify potential strategies for mitigation. With complex ecosystem functions and uncertainty of climate change, both the micro-and macroscopic behavior of carbon flow and its influencing factors have to be systematically studied. There are consequently significant national and international efforts to develop a carbon monitoring system, such as the National Forest Carbon Monitoring, Accounting and Reporting System developed by Canada government, and National Carbon Accounting System by Australian government. These systems aims to tracking and forecasting land based emissions and removals of greenhouse gases from land-use changes, livestock and crop production, and disturbance events such as deforestation, afforestation, and natural disturbances. Based on synergic analysis of data from detailed forest inventory, remote sensing, and ecosystem modeling, the accounting results are used to monitor forest and carbon cycles, and are reported internationally.

This paper presents a numerical study of microwave brightness temperature of residual foam cover on ocean surface. The foam layer is modeled using Kelvin's Tetrakaidecahedron structure. Absorption and scattering of the foam layer, i.e. the layer phase matrix, are calculated using a multilevel UV method to accelerate method of moment (MoM) solution of Maxwell's equations. Matrix doubling method, which accounts for interactions between the foam volume and interfaces, is then utilized to determine the total emission from the foam layer. The calculated foam layer absorption rates at a frequency range of 1.5GHz to 36.5GHz are presented and analyzed for both H and V polarizations. Brightness temperatures at 1.5GHz for both planar and rough interface with different water fractions in the foam layer are simulated and analyzed. Effects of foam-boundary interactions on foam layer emission are discussed.

SAR is a complex system that integrates two major parts: data collector and image formatter [1-2]. In the phase of data collection, radar transmits electromagnetic waves toward the target and receives the scattered waves. The transmitted signal can be modulated into certain types, commonly linearly frequency modulated with pulse or continuous waveform. The process involves signal transmission from generator, through various types of guided device, to antenna, by which the signal is radiated into free space, and then undergoes propagation. The measured scattered signal been made in bistatic or monostatic configurations is essentially in time-frequency (delay time - Doppler frequency) domain. The role of image formatter is then to map the time-frequency data into spatial domain where the targets are located. The mapping from the data domain to image domain, and eventually, into target or object domain must minimize both geometric and radiometric distortions. Essentially, two models that define the SAR operational process: physical model and system model. This paper concentrates on the physical process of a SAR system from wave scattering to imaging. System simulation based on the stationary (frequency modulation continuous wave) FMCW is developed and implemented for both point target and extended target. To further validate the simulation and thus our physical understanding of the imaging chain, measurements at aniconic chamber with two mental spheres and two dielectric spheres displaced with varying spacing were conducted. Good agreement between the simulated by extend target model and real measured SAR images is obtained.

This paper presents bistatic scattering response of rice canopy over growth stages based on a three-layer microwave scattering model, which is developed using the iterative solution of the vector radiative transfer equations up to the second order, and the dense medium phase and amplitude correction theory (DM-PACT) is used to improve the phase matrix taking coherent effects into account. To validate the model, ground-based measurements of backscattering coefficients over an entire rice-growth stage is used. Then, the bistatic scattering response of rice canopy is analyzed in respect of canopy components, including plant height, structure, soil moisture, and surface roughness, as well as their interactions with sensor configurations, such as frequency, polarization, and incident angle. The sensitivity analysis was carried out to demonstrate how the model reacts to changes of each input. The results show that bistatic scattering coefficient varies greatly over different stages, and is strongly dependent on rice plants' structure, including the size, shape, and orientation, especially stems.

In this paper, we investigated the bistatic scattering of soil moisture and surface roughness of bare soil surfaces. We generated database by using an advanced integral equation model (AIEM). For better understanding the bistatic scattering characteristics of bare soil surfaces, we adopted single polarized simulations and combination of dual angular simulations. To explore the sensitivity of bistatic scattering to soil moisture, we applied a defined sensitivity index. The results shown that the scattering coefficients of VV polarization are more sensitive to soil moisture compared with the HH polarized scattering coefficients. Moreover, the forward direction is the most sensitive to soil moisture in all cases. Besides, the dual angular observations show good sensitivity to soil moisture, especially when the differences of the two incident angles are large.

In this paper, we evaluate the polarimetric calibration method that relies only on the characteristics of the calibration targets and requires no assumptions regarding scene statistics and radar characteristics. It allows the correction of polarimetric distortions of the radar system using only three types of calibration target: trihedral, dihedral, the 22.5° and 45°-rotated dihedral corner reflectors. The method is simple, accurate and yet robust in terms of field deployment in some cases. Evaluation and testing of the algorithm was made at microwave anechoic chamber measurements, numerical simulation, and airborne L-band Pi-SAR data. It is found that the amplitude error can be achieved to as small as 0.5dB, while the phase error is within 5° in high resolution of Pi-SAR system. Due to the sensitivity of 22.5°-rotated dihedral to rotation angle, it is not recommended to use the 22.5° -rotated dihedral for polarimetric calibration.

This paper investigates the bistatic scattering from a surface that features randomly corrugated with irregular grooves. The total scattered field is decomposed into coherent and incoherent components to analyze their respective contributions. It was found that for randomly corrugated surfaces with irregular ridges or grooves, the coherent scattering is profound at several scattering angles with strong main lobes. Numerical simulation shows that the coherent lobe's angular shift away from specular direction is quasi-linearly dependent on the ridge density, suggesting that the coherent scattering pattern substantially contain the surface geometric information. It is expected our study offers deeper understanding of coherent imaging of rough surface and helps designing a novel imaging system for random targets.

The SMAP (soil moisture active passive) which is one of the satellites that specifically designed for soil moisture monitoring, was launched on 31 January 2015. Recently, the SMAP radiometer soil moisture product has been released to the public. It is very urgent to evaluate the reliability of this product before it can be widely used in hydrometeorological studies. In the study, we carried out an initial evaluation of SMAP radiometer soil moisture product against in-situ measurements from three networks. The three networks cover different land surface conditions, including two dense networks established in United States and Finland, and one sparse network set up in Romania. The results show that the SMAP soil moisture product agrees very well with the in-situ measurements although it sometimes exhibits dry or wet bias at different network regions. The overall ubRMSE of SMAP product is 0.036 m3m-3, well within the mission requirement of 0.04 m3m-3. Considering the algorithms are still under refinement, it can be reasonably expected that applications such as climate modeling and flood forecasting will benefit from the SMAP passive soil moisture product.

Remote sensing of soil moisture in a bistatic mode has attracted increasing attention in recent years. This paper investigates the bistatic radar response of soil moisture and surface roughness at L-band by using the advanced integral equation model (AIEM). To better explore the potential of bistatic scattering for soil moisture sensing, both polarized and angular scattering coefficients, and their combinations, were evaluated using a defined sensitivity index. Results show that sensitivity is enhanced in a bistatic mode compared to the monostatic case. Using a combination of dual polarized and angular data suppresses an undesired impact of the surface correlation function. Among the combinations, the dual angular observation reduces the influence of surface roughness, preserves a good sensitivity to soil moisture, and thus seems to be a good candidate for sensing soil moisture in bistatic configuration.

This letter investigates the bistatic radar response of soil moisture and surface roughness of bare soil surfaces at L-band using the advanced integral equation model (AIEM). It focuses on the use of bistatic geometries away from the specular region. To better explore the potential of bistatic scattering for soil moisture sensing, both polarized and angular scattering coefficients, and their combinations, are evaluated using a defined sensitivity index. The results show that sensitivity is enhanced in a bistatic mode compared with the monostatic case. Using a combination of dual polarized and angular data suppresses an undesired impact of the surface correlation function. Moreover, the forward region is preferred to soil moisture sensing regardless of the surface correlation function. Among the combinations investigated, the dual angular observation reduces the influence of surface roughness, preserves a good sensitivity to soil moisture response, and thus seems to be a good candidate for soil moisture sensing in bistatic configuration.

This paper presents an electromagnetic modeling of temporal variations of microwave backscatter from rice canopy based on radiative transfer theory to understand the complex microwave scattering mechanisms of rice crops at different growth stages. Model validation is made by a comparison with field measurements from four independent campaigns by ground-based scatterometers and spaceborne synthetic aperture radar (SAR). Then, the frequency responses and angular dependence are examined, followed by an analysis of parameter sensitivity and uncertainty to microwave backscatter. The validation shows promising results for the physically based model in assessing radar response of rice plant in its full-growth stage. It is found that radar scattering of rice canopy is highly dependent on the growing stages of rice plants, in addition to radar configurations, such as frequency, polarization, and incident angle. Moreover, backscattering of rice canopy is more dependent on stems than on leaves, and as rice plants grow, increasing stems and leaves tend to promote HH- and VH-polarized scattering, while first promoting and then reducing VV-polarized scattering. Moreover, the greatest uncertainty takes place at the late growth stage for copolarization and at the early stages for cross-polarization. The main contributions of this paper can be summarized as follows: 1) the establishment of a rice canopy microwave backscattering model using radiative transfer theory adapted to all the stages in the phenological cycle of rice; 2) the validation of the performance of the proposed model with numerous independent measurements from spaceborne SAR and ground-based scatterometers from different countries and regions; 3) the investigation of the interactions between microwave backscatter signatures and rice canopy growth variables over growth stages; and 4) the identification of the dominant influential parameters of the rice scattering model and determination model uncertainty.

This paper investigates the bistatic scattering from randomly corrugated surfaces with irregular grooves. For a given incident wave, the surface fields and the scattered field were computed by the method of moments in which the rooftop basis function was used to account for fast phase changes due to steep surface slopes. The total scattered field is decomposed into coherent and incoherent components to analyze their respective contributions. We found that, for randomly corrugated surfaces with irregular ridges or grooves, the coherent scattering is profound at several scattering angles with strong main lobes, whose beamwidths are strongly correlated with the ridge density. The numerical simulation has shown that the lobe angular shift away from the specular direction is quasi-linearly dependent on the ridge density, suggesting that the coherent scattering pattern substantially contains the surface geometric information. We expect this paper to offer deeper understanding of coherent imaging of rough surface and to help in designing a novel imaging system.

The Soil Moisture Active Passive (SMAP) mission, which is the newest L-band satellite that is specifically designed for soil moisture monitoring, was launched on January 31, 2015. A beta quality version of the SMAP radiometer soil moisture product was recently released to the public. It is crucial to evaluate the reliability of this product before it can be routinely used in hydrometeorological studies at a global scale. In this paper, we carried out a preliminary evaluation of the SMAP radiometer soil moisture product against in situ measurements collected from three networks that cover different climatic and land surface conditions, including two dense networks established in the U.S. and Finland, and one sparse network set up in Romania. Results show that the SMAP soil moisture product is in good agreement with the in situ measurements, although it exhibits dry or wet bias at different network regions. It well reproduces the temporal evolution and anomalies of the observed soil moisture with a favorable correlation greater than 0.7. The overall ubRMSE (unbiased root mean square error) of SMAP product is 0.036 m3· m-3, well within the mission requirement of 0.04 m3· m-3. The error sources of SMAP soil moisture product may be associated with the parameterization of vegetation and surface roughness but still needs to be tested and confirmed in more extent. Considering that the algorithms are still under refinement, it can be reasonably expected that hydrometeorological applications will benefit from the SMAP radiometer soil moisture product.

This paper investigates the algorithms that are used to predict the full polarimetric bistatic normalized radar cross-section of rough surfaces. These include small perturbation method (SPM), the physical optics (PO) approach, the small slope approximation (SSA) and the integration equation method (IEM) and its derivatives improved IEM and advanced IEM. The methods are then compared to ground truth values obtained from multiple Monte Carlo runs a numerical Method of Moments (MOM) code using the same surface statistics. Effects of using band-limited exponential instead of a true exponential correlation function for surface statistics are also explored. Mean L1-norm error values integrated over the hemisphere are given between AIEM and MOM and SSA and MOM.

The advancement of synthetic aperture radar (SAR) technology with high-resolution and quad-polarization data demands better and efficient polarimetric SAR (PolSAR) speckle-filtering algorithms. Two requirements on PolSAR speckle filtering are proposed: 1) speckle filtering should be applied to distributed media only, and strong hard targets should be kept unfiltered; and 2) scattering mechanism preservation should be taken into consideration, in addition to speckle reduction. The purpose of this paper is twofold: 1) to propose an effective algorithm that is an extension of the improved sigma filter developed for single-polarization SAR; and 2) to investigate speckle characteristics and the need for speckle filtering for very high resolution (decimeter) PolSAR data. The proposed filter was specifically developed to account for the aforementioned two requirements. Its effectiveness is demonstrated with Jet Propulsion Laboratory airborne synthetic aperture radar data, and comparisons are made with a boxcar filter, the refined Lee filter, and a Wishart-based nonlocal filter. For very high resolution PolSAR systems, such as the German Aerospace Center F-SAR and Japanese Pi-SAR2, with decimeter spatial resolution, we found that the complex Wishart distribution is still valid to describe PolSAR speckle characteristics of distributed media and that speckle filtering may be needed depending on the size of objects to be analyzed. F-SAR X-band data with 25-cm resolution is used for illustration.

We have performed simulations of random rough surface scattering using 3-D numerical solution of Maxwell equations (NMM3D) using surface size up to 32 × 32 squared wavelengths. The rough surfaces are characterized by exponential correlation functions. The simulation results of crossand copolarization backscattering coefficients were in good agreement with experimental measurements of bare soils at L-band. Because in numerical solutions of Maxwell equations the electric fields of the scattered wave are calculated for each realization, scattering matrices can be simulated by NMM3D, and such simulations are performed in this paper. For a given RMS height, correlation length, soil permittivity, and incident angle, we calculated the radar scattering matrix up to 958 independent realizations. For each realization, the components of the scattering matrix, namely, SHH, SVV, SHV, and SVH, are calculated. Using the simulated scattering matrices, we calculate the polarimetric speckle statistics (amplitude and phase difference), followed by a comparison with theoretical distributions. For fully developed speckle from the homogeneous rough surface, the results are examined and validated to ensure the simulated data quality as far as polarimetric properties are concerned. By taking ensemble averages, we calculate the coherency matrix from which the eigenvalues, entropy, anisotropy, and alpha angle in coherent target decomposition are then calculated. In particular, characterization of polarimetric descriptors for rough surface is presented. Issues of scattering symmetry characteristics are also discussed.

This study aims at quantifying vegetation fractional cover (VFC) by incorporating multi-resolution satellite images, including Formosat-2(RSI), SPOT(HRV/HRG), Landsat (MSS/TM) and Terra/Aqua(MODIS), to investigate long-term and seasonal vegetation dynamics in Taiwan. We used 40-year NDVI records for derivation of VFC, with field campaigns routinely conducted to calibrate the critical NDVI threshold. Given different sensor capabilities in terms of their spatial and spectral properties, translation and infusion of NDVIs was used to assure NDVI coherence and to determine the fraction of vegetation cover at different spatio-temporal scales. Based on the proposed method, a bimodal sequence of intra-annual VFC which corresponds to the dual-cropping agriculture pattern was observed. Compared to seasonal VFC variation (78~90%), decadal VFC reveals moderate oscillations (81~86%), which were strongly linked with landuse changes and several major disturbances. This time-series mapping of VFC can be used to examine vegetation dynamics and its response associated with short-term and long-term anthropogenic/natural events.

Radar interferometry has been widely applied in measuring the terrain height and its changes. The information about surface can be derived from phase interferograms. However, the inherent phase noise reduces the accuracy and reliability of that information. Hence, the minimization of phase noise is essential prior to the retrieval of surface information that is embedded in an interferometric phase. This paper presents a refined filter based on the Lee adaptive complex filter and the improved sigma filter that was originally developed for amplitude image filtering. The basic idea is to adaptively filter the interferometric phase according to the local noise level to minimize the loss of signal for a particular pattern of fringes, including such extreme cases as involving broken fringes, following the removal of undesired pixels. Ultimately, the goals are to preserve the fringe pattern, to reduce phase bias and deviation, to reduce the number of residues, and to minimize the phase error. The preservation of the fringe pattern is particularly of concern in areas of high frequency of fringe and large phase gradient corresponding to steep terrains. The proposed refined filter was validated using both simulated data and real interferometric data. Results demonstrate that the filtering performance is better than that of commonly used filters.

Radar interferometry has been widely applied in measuring the surface height. The information about surface can be derived from phase interferograms. However, phase noise reduces the accuracy and reliability of that information. Hence, the minimization of phase noise is essential to the retrieval of surface information. This work presents a refined filter that is based on the Lee adaptive INSAR filter and the sigma filter. The basic idea is to filter adaptively the interferometric phase according to the local noise level to minimize the loss of signal for a particular shape of fringes, including in such extreme cases as involve broken fringes, following the elimination of unreliable pixels of phase noise. The goal is to reduce the phase deviation and the number of residues, and minimize the phase error. The proposed filter was inspected herein using both simulated data and real interferometer data. Results reveal that the filtering performance is better than that of commonly used filters.

Knowing the temporal features of soil moisture dynamics is essential for proper water resource management, fertilization management, and crop production. This paper proposes a recurrent dynamic learning neural network (RDLNN) to estimate soil moisture evolution by rainfall forcing. Long-term measurements of rainfall and soil moisture content were gathered. Soil moisture contents estimated from daily and/or hourly precipitation by RDLNN, were compared with ground measurements. Experimental results suggested that RDLNN is a promising tool for estimating soil moisture from hourly precipitation.

In this letter, spatial information through fractal measures is adopted to combine with the spectral information to improve the land cover classification. The spectral features alone and later combined with texture features, using MODIS/ASTER airborne simulator imagery, were fed into a neural classifier. Classification performance was evaluated by a confusion matrix measured by overall accuracy and kappa coefficient. In particular, a parallel differential box-counting (DBC) (PDBC) algorithm for fractal estimation was implemented on a multicore PC. The computation efficiency was ensured through the use of PDBC algorithm which is much faster than that of the original DBC. Furthermore, multicore processors offer great potential for speeding up the computation by partitioning the load among the cores. Multithreading technique is adopted to fully explore its multicore capability. Experimental results demonstrate that the proposed approach provides substantial improvements in classification accuracy while requiring much less computation time without extra hardware resources.

Earth is an integrated, complex system with strong coupling among atmosphere, hydrosphere, biosphere, and lithosphere processes. The US Geological Survey has reported globally more than 17 earthquakes per year with a magnitude 7 and higher in the last 18 years. The remote sensing community is actively and quickly moving toward more advanced methodologies, linking remote sensing with in situ measurements and ancillary data for more precise mapping, faster analysis, and more effective forecasting and data delivery to the user community. The International Charter BSpace and Major Disasters was established to enable such collaboration in sensor tasking during times of crisis and is often activated in response to calls for assistance from authorized users. Insight is provided from a US perspective into sensor support for Charter activations and other disaster events.

This study design a high precision active radar calibrator (ARC) with programmable echo points which served as control points for both geometric and radiometric calibrations or other applications. The echo delay and Doppler shift of ARC system are fully digitally controlled. Specification and performance of the digital ARC were tested on C-band ERS-2 SAR image. From the experimental results, the FPGA circuit can produce several calibration points of desired range or azimuth direction. In InSAR application, the echo points of ARC system can estimate squint angle difference in two co-registration images. Compared to analog time delay ARC system, the current ARC generates delay error less than 10ns, greatly reduced from 30-50 ns that analog time delay has. In addition, the stability of the time delay and output power in the FPGA based ARC are both much better than the traditional analog delay ARC.

In this paper, we propose a terrain categorization algorithm for single-polarization high-resolution SAR imagery, and coded the terrain categories by colors that resemble Pauli decomposition color coding based on scattering mechanisms of PolSAR data. The proposed algorithm emphasizes preserving high resolution and making information dissemination easier for single-pol SAR data.

The one of the greatest contributions of Professor Adrian K. Fung to the microwave remote sensing is the development of the integral equation model that has significantly improved and advanced our ability in modeling microwave surface emission and scattering signals. The surface emission and scattering model is one of the essential components in many applications of the microwave remote sensing of geophysical properties in the complex earth terrain. It is a direct component in monitoring soil moisture in the bare surfaces. The surface emission and scattering model serves as the boundary condition in studying snow, vegetation, and atmospheric properties. It has been found that the underground surface emission and scattering signals have a great impact on snow water equivalence and vegetation properties retrieval. How the multi-frequency and polarization measurements to be used correctly in deriving the surface geophysical and atmospheric properties is an important research issue in microwave remote sensing. It is heavily depending upon our understanding on the ability of modeling surface emission and scattering signals. In honor of Professor Adrian K. Fung, this paper demonstrates how the integral equation model has been used in the developing the retrieval algorithms for several geophysical parameters, including soil moisture and vegetation properties.

In this letter, we investigated an inversion technique to estimate snow water equivalence (SWE) under Advanced Microwave Scanning Radiometer for Earth Observing System (AMSR-E) sensor configurations. Through our numerical simulations by the advanced integral equation model (AIEM), we found that the ground surface emission signals at 18.7 and 36.5 GHz were highly correlated regardless of the ground surface properties (dielectric and roughness properties) and can be well described by a linear function. It leads to a new development for describing the relationship between snow emission signals observed at 18.7 and 36.5 GHz as a linear function. The intercept (A) and slope (B) of this linear equation depend only on snow properties and can be estimated from the observations directly. This development provides a new technique that separates the snowpack and ground surface emission signals. With the parameterized snow emission model from a simulated database that was derived using a multiscattering microwave emission model (dense medium radiative transfer model-AIEM-matrix doubling) over dry snow covers, we developed an algorithm to estimate the SWE using the microwave radiometer measurements. Evaluations on this technique using both the model simulated data and the field experimental data with the airborne Polarimetric Scanning Radiometer data from National Aeronautics and Space Administration Cold Land Processes Experiment 2003 showed promising results, with root-mean-square errors of 32.8 and 31.85 mm, respectively. This newly developed inversion method has the advantages over the AMSR-E SWE baseline algorithm when applied to high-resolution airborne observations.

Trees play an important role in maintaining environmental conditions suitable for life on the earth. To classify the tree type is very important for the forest maintenance. With the advent of high spatial resolution remote sensing sensors, our ability has greatly increased for tree type identification. Considering the amount of data in need of processing and the high computational costs required by image processing algorithms, conventional computing environments are simply impractical. Therefore, it is necessary to develop techniques and models for efficiently processing large volume of remote sensing images. In this study, a cluster computing environment was adopted to speed up the computation time. The test image was first partitioned into hundreds of manageable sub-images. Scheduled by the head node, the sub-images were then distributed to compute nodes for processing. A distributed K-mean clustering algorithm with undetermined number of class was applied to each compute node. A promising result was obtained. Compared to the field investigations, tree types of the test site were properly identified. In addition, great improvement in computation time was obtained. The distributed K-mean clustering algorithm implemented on our cluster computing environment performed much faster than stand-alone alternatives. By adding more compute nodes to our cluster computing environment, further improvement in computation time is expected.

This paper addresses the feasibility of using RADARSAT-2 fine Quad-Pol mode to monitor coastal environment and young tree growth. It will be shown that interferometric coherence may not be high enough for the height estimation of young trees at C-band, but polarimetric sensitivity could be used for tree and crop classification. For coastal environment, we found that polarimetric signature of oyster farm reveals the effect of double bounce scattering and the orientation angle effects.

In this study, we develop a combined active/passive technique to estimate surface soil moisture with the focus on the short vegetated surfaces. We first simulated a database for both active and passive signals under SMAP's sensor configurations using the radiative transfer model with a wide range of conditions for surface soil moisture, roughness and vegetation properties that we considered as the random orientated disks and cylinders. Using this database, we developed 1) the techniques to estimate surface backscattering and emission components and 2) the technique to estimate soil moisture with the estimated surface backscattering and emission components. We will demonstrate these techniques with the model simulated data and its validation with the airborne PALS image data from the soil moisture SGP'99 and SMEX'02 experiments.

The effects of soil surface roughness play a significant role in the microwave emission from the surface. Therefore, a good parameterization of the effects is a prerequisite for retrieving surface soil moisture information. With recent physical model developments, the advanced integral equation model (AIEM) has been proven to provide accurate representation over a wide range of surface-roughness conditions. We evaluated the capability of the AIEM model in simulating multiangular surface emission signals in comparison with a field experiment data set. A simplified multiangular surface emission model was developed based on simulated database using the AIEM model. Based on the parameterized model, an inversion procedure was developed using dual-polarization microwave brightness temperatures to retrieve soil moisture. Two data sets were used to test the inversion algorithm, and the accuracies in root-mean-square error were about 4% for incidence angles from 20° to 50°. This new simple model is suitable for soil moisture retrieval from future L-band satellite data.

This paper presents an overview of the current status and data applications of FORMOSAT-2, Taiwanese's first earth observation satellite mission. Highlights of its contributions to monitoring of global natural disasters and earth environmental changes will be illustrated. The FORMOSAT-2 satellite successfully complements existing high spatial resolution imaging satellites such as SPOT-5, IKONOS, and QuickBird, among others, with its unique capability of daily revisits worldwide. The FORMOSAT-2 follow-up program to ensure data continuity to the user community is briefly introduced.

In this paper we present a new supervised classification method, referred to as the k-way tree semi-greedy (KTSG) classifier, for the classification of multisource remote sensing images. The generalized positive Boolean function (GPBF) classifier scheme is recently proposed based on minimum classification error (MCE) criteria to improve classification performance. It makes use of MCE criteria to apply positive and negative samples as training parameters. Unfortunately, the classification performance of GPBF is limited when the number of classes increases. This is occurred in training phase by the unbalanced numbers of positive and negative samples caused by the use of a large number of classes. The proposed KTSG overcomes this drawback by modifying the scheme from the perception of pattern-node based semi-greedy (bottom-up scheme used in GPBF) to the conception of region-based semi-greedy (also known as the top-down scheme in KTSG). It is organized by a k-way tree in which every node is composed of a set of k-dimensional positive and negative labeled samples as represented as a percentage, i.e. the corresponding ratio of number of a specific (positive) class samples to the total number of the other (negative) classes. It iteratively divides the d-dimensional hyperplane into 2dsubspaces according to the centroids of the labeled (training) samples of all classes. The statistical ratios between different classes are then compared as a basis for stopping the new subspace separation and identifying which subspace belongs to which class. By delivering both positive and negative samples of different classes to KTSG learning modules, KTSG outperforms GPBF and traditional classifiers in terms of classification accuracies. The effectiveness of the proposed KTSG is evaluated by fusing MODIS/ASTER airborne simulator (MASTER) hyperspectral images and airborne synthetic aperture radar (AIRSAR) images for land cover classification during the Pacrim II campaign.

Greedy modular eigenspaces (GME) has been developed for the band selection of hyperspectral images (HSI). GME attempts to greedily select uncorrelated feature sets from HSI. Unfortunately, GME is hard to find the optimal set by greedy operations except by exhaustive iterations. The long execution time has been the major drawback in practice. Accordingly, finding an optimal (or near-optimal) solution is very expensive. In this study we present a novel parallel mechanism, referred to as parallel particle swarm optimization (PPSO) band selection, to overcome this disadvantage. It makes use of a new particle swarm optimization scheme, a well-known method to solve the optimization problems, to develop an effective parallel feature extraction for HSI. The proposed PPSO improves the computational speed by using parallel computing techniques which include the compute unified device architecture (CUDA) of graphics processor unit (GPU), the message passing interface (MPI) and the open multi-processing (OpenMP) applications. These parallel implementations can fully utilize the significant parallelism of proposed PPSO to create a set of near-optimal GME modules on each parallel node. The experimental results demonstrated that PPSO can significantly improve the computational loads and provide a more reliable quality of solution compared to GME. The effectiveness of the proposed PPSO is evaluated by MODIS/ASTER airborne simulator (MASTER) HSI for band selection during the Pacrim II campaign.

The orientation angle of scattering media affects the polarimetric radar signatures. This paper investigates the effect of orientation compensation on polarimetric target decompositions including Pauli decomposition, Freeman and Durden decomposition and Yamaguchi decomposition. The Cloude and Pottier decomposition is excluded, because entropy, anisotropy and alpha angle are rotational invariant. We will show that after the orientation compensation, the volume scattering power is consistently decreased, while the double bounce power has increased. The surface scattering power is relatively unchanged, and the helicity power is rotational invariant. All these characteristics can be explained by the compensation effect on the nine elements of the coherency matrix. This analysis reveals that, contrary to the general perception, the 4-component component decomposition by Yamaguichi et al. does not use complete information of the coherency matrix. Only six quantities are included — one more than the Freeman/Durden decomposition under the assumption of refection symmetry.

In this paper, we study the electromagnetic scattering properties of 2-D rough surface with large slope and large height. The ridges on the surface have heights of about 20cm. In microwave remote sensing of land, these heights are larger than wavelength. By using a tapered incident wave, the surface fields are solved by using numerical solution of Maxwell equations. Method of Moment (MOM) is used to solve the surface integral equations and rooftop basis function and Galerkin's method are used. The bistatic reflection and transmission are then calculated from the surface fields. Then the reflectivity from sastrugi over layered snow is calculated by solving multilayer radiative transfer (RT) equation with bistatic reflection and transmission coefficients as the boundary condition. We compare the electromagnetic scattering properties between Sastrugi rough surface and smooth surface. We show in this paper that for the sastrugi case, transmission angle can be larger than incident angle when incident from air to snow. This results in total internal reflection when the second layer of snow beneath sastrugi has a smaller permissivity and larger reflectivity than smooth surface.

This study demonstrates a new algorithm development for estimating bare surface soil moisture using dual-polarization L-band backscattering measurements. Through our analyses on the numerically simulated surface backscattering database by Advanced Integral Equation Model (AIEM) with a wide range of soil moisture and surface roughness conditions, we found that the relative difference of the overall surface roughness parameters at the different co-polarizations can be well estimated through a roughness index. This new finding leads to an algorithm on estimation of bare surface soil moisture. We will demonstrate the theory and techniques of this algorithm through the AIEM simulated database and validate it with two field ground scatterometer experimental data. The results indicate that bare surface soil moisture can be estimated quite well with only co-polarized backscattering signals. It provides a solid support for Soil Moisture Active and Passive mission (SMAP).

Hyperspectral images with hundreds of narrow spectral channels are currently available and instruments with thousands of spectral bands are under development. It is necessary to develop techniques and models for efficiently processing large volume of remote sensing images. Multi-core processors present an opportunity for speeding up the computation by partitioning the load among the cores. As multi-core processor systems become more and more widespread, the demand of efficient parallel algorithms also propagates into the field of remote sensing images processing. Classification of land cover types in a hyperspectral image is demonstrated in this study. A dynamic learning neural network (DLNN) is utilized as a supervised classifier. To get better classification accuracy, texture information is extracted and combined with the spectral information. Fractal dimension, the texture information applied, is estimate by a differential box-counting (DBC) technique. The original DBC is inefficient because the fractal dimension is evaluated sequentially. In this study, a parallel DBC is proposed and implemented on a multi-core PC to improve its efficiency. To fully explore its multi-core capability, multi-threading technique is adopted. Experimental results reveal that the improvement in computation time of the parallel DBC is depended on the ratio of window size M and grid size s. In addition, further improvement provided by multi-threading techniques is linearly proportion to the number of cores.

Based on a nonquadratic-optimization method originally proposed for spotlight-mode SAR image reconstruction, a modification for stripmap-mode SAR images is presented in this letter. This is done by mathematically reformulating the projection kernel and numerically putting it into a form that is suitable for optimization. The performance was evaluated by measures of the target contrast and 3-dB beamwidth using Radarsat-1 data. Results were analyzed and compared with those using minimum-variance and multiple-signal-classification methods. Results demonstrate that the target's features are effectively enhanced and that the dominant scattering centers are well separated using the proposed method. In addition, the image fuzziness is greatly reduced, and the image fidelity is well preserved. The effectiveness of the modification is thus validated.

The Lee sigma filter was developed in 1983 based on the simple concept of two-sigma probability, and it was reasonably effective in speckle filtering. However, deficiencies were discovered in producing biased estimation and in blurring and depressing strong reflected targets. The advancement of synthetic aperture radar (SAR) technology with high-resolution data of large dimensions demands better and efficient speckle filtering algorithms. In this paper, we extend and improve the Lee sigma filter by eliminating these deficiencies. The bias problem is solved by redefining the sigma range based on the speckle probability density functions. To mitigate the problems of blurring and depressing strong reflective scatterers, a target signature preservation technique is developed. In addition, we incorporate the minimum-mean-square-error estimator for adaptive speckle reduction. Simulated SAR data are used to quantitatively evaluate the characteristics of this improved sigma filter and to validate its effectiveness. The proposed algorithm is applied to spaceborne and airborne SAR data to demonstrate its overall speckle filtering characteristics as compared with other algorithms. This improved sigma filter remains simple in concept and is computationally efficient but without the deficiencies of the original Lee sigma filter.

Worldwide, medium- to short-term earthquake prediction is becoming ever more essential for safeguarding man due to an un-abating population increase, but hitherto there have been no verifiable methods of reliable earthquake prediction developed. This dilemma is a result of previous and still current approaches to earthquake prediction which are squarely based on the measurement of crustal movements, observable only after a tectonic stress-change discharge (earthquake) has occurred. During the past decades it was proved and shown that it is not possible to derive reliable models for earthquake predictions from crustal movement measurements alone, and that an entirely new approach must be taken and rigorously pursued over years and decades to come. In support of this conclusion, there have been reported throughout the history of man anecdotal historical up to scientifically verifiable earthquake precursor or "seismo-genic" signatures of various kind - biological, geological, geo-chemical and especially a rather large plethora of diverse electromagnetic ones on ground, in air and space, denoted as "seismo-electromagnetic" signatures. Taiwan is one of the few regions where those phenomena may best be observed. In this overview a systematic analysis of main historical records, a summary of pertinent "seimogenic" as well as observed "seismo-electromagnetic" effects and modern ground-based to air- and space-borne metrological signature investigations are presented placing major emphasis on ongoing studies in Taiwan.

In this paper we present a parallel classification learning method, referred to as parallel minimum classification error (PMCE) learning, for supervised classification of multisource remote sensing images. The approach is based on the positive Boolean function (PBF) classifier scheme. The PBF implements the minimum classification error (MCE) as a criterion to improve classification performance. By evenly distributing both positive and negative samples of MCE learning modules to different PMCE learning nodes, PMCE outperforms the original one in terms of execution time. It fully utilizes the significant parallelism embedded in MCE learning of PBF to create a set of PMCE learning nodes implemented by using the message passing interface (MPI) library and the open multi-processing (OpenMP) application programming interface. A sophisticated hierarchical structure of hybrid PMCE, which combines cluster based MPI with multicore-based OpenMP, is proposed to demonstrate the flexibility of implementation of the proposed scheme. The effectiveness of the proposed PMCE is evaluated by fusing MODIS/ASTER airborne simulator (MASTER) hyperspectral images and the Airborne Synthetic Aperture Radar (AIRSAR) images for land cover classification during the Pacrim II campaign. The experimental results demonstrated that PMCE can improve the computational speed of PBF classification significantly.

In this paper we present a parallel band selection approach, referred to asparallelsimulatedannealingbandselection(PSABS), for hyperspectral imagery. The approach is based on thesimulatedannealingbandselection(SABS) scheme. The SABS algorithm is originally designed to group highly correlated hyperspectral bands into a smaller subset of band modules regardless of the original order in terms of wavelengths. SABS selects sets of non-correlated hyperspectral bands based onsimulatedannealing(SA) algorithm and utilizes the inherent separability of different classes in hyperspectral images to reduce dimensionality. In order to be effective, the proposed PSABS is introduced to improve the computational speed by using parallel computing techniques. It allowsmultipleMarkovchains(MMC) to be traced simultaneously and fully utilizes the significant parallelism embedded in SABS to create a set of PSABS modules on each parallel node implemented by themessagepassinginterface(MPI) cluster-based library and theopenmulti-processing(OpenMP) multicore-based application programming interface. The effectiveness of the proposed PSABS is evaluated byMODIS/ASTERairbornesimulator(MASTER) hyperspectral images for hyperspectral band selection during the PACRIM II campaign. The experimental results demonstrated that PSABS can significantly improve the computational loads and provide a more reliable quality of solution compared to the original SABS method.

The advancement of SAR technology with high resolution and multiple polarization data demands better and efficient speckle filtering algorithms. In this paper, we developed an effective speckle filtering algorithm for dual-pol and fully polarimetric high resolution data. The proposed algorithm is effective and computational efficient based on the improved sigma filter [1]. ALOS/PALSAR and JPL AIRSAR data were used for demonstration.

The purpose of this study is to evaluate whether the NASA/JPL dual frequency airborne system, Passive Active L-band and S-band (PALS), can provide a reliable soil moisture measurements so that they can be integrated to provide soil moisture data at the scales of the spaceborne coarse resolutions. Through evaluations of the AIEM simulated the random rough surface emissivities with a wide range of soil moisture and roughness conditions at both L-band and S-band, it was found that the bare surface emission signals at above two frequencies are essentially close to identical regardless of the surface soil moisture and roughness properties when the effect of soil properties (temperature and moisture) in the vertical profile on emission signals is minor. This makes it possible to further estimate the vegetation components in the omega-tau model at each frequency and polarization and to carry out the corrections for the vegetation effects without assumption on polarization dependence of the vegetation effects. Thus, the surface soil moisture can be inferred by the estimated surface emission signals. We will show the evaluation and validation of this technique with the airborne PALS measurements obtained during SMEX'02 soil moisture experiment with the intensive ground soil moisture measurements.

It is very difficult to perform change detection in SAR images, because speckle noise contaminates the images in nature. Speckle, which results from coherent energy imaging, is indeed a chaotic phenomenon. As a result, an SAR signal can be modeled by a spatial chaotic system and characterized by its fractal dimension. The differential box-counting (DBC) technique is adopted to estimate fractal dimension in this paper. Based on the spatial chaotic model (SCM), a simplified SAR image change detection procedure is proposed. Observations provided by SAR sensors are uncertain due to changing illumination conditions at different acquiring time. Besides, the selection of window size M and grid size s in DBC provides an additional degree of uncertainty. Both the uncertainty involved in the measurements and the uncertainty involved in the selection of M and s motivate us of integrating type-2 fuzzy sets with the SCM to achieve a better performance. The proposed approach is applied to multitemporal polarimetric SAR images for change detections as demonstrations. The change detection results of using the original SCM method and the proposed approach are compared. The effects of misregistration for different change detection approaches are also presented. Simulation results suggest that the proposed approach is more tolerant to misregistration and offers better results of detecting changes when speckle noise is present.

Coastline detection in SAR images suffers from the presence of speckle effect and the strong signal return from a wind-roughened and/or wave-modulated sea. It has already been recognized that ocean areas in SAR images are almost always much more homogeneous in grey levels than land areas. Therefore, features reflecting the roughness of an image can be very useful for ocean-land separation. To represent its geometric property, an SAR signal is modeled by a spatial chaotic model (SCM) and characterized by its fractal dimension. The differential box-counting (DBC) technique is adopted to estimate fractal dimension in this paper. Observations provided by SAR sensors are uncertain due to changing illumination conditions at different locations. Besides, the selection of window sizeMand grid size s in DBC provides an additional degree of uncertainty. Both the uncertainty involved in the measurements and the uncertainty involved in the selection ofMand s motivate us of integrating type-2 fuzzy sets with the SCM to achieve an appropriate selection of the threshold for ocean-land segmentation. The proposed approach is applied to an SAR image for coastline detection as a demonstration. The final result shows that the coastline detected coincide very well with the true situation when it is overlaid on the original image. Besides, the detected coastline also agrees very well with the terrestrial measurements.

In this paper, we report on the polarimetric active and passive microwave remote signatures for exponential correlation function surfaces. Applications are in soil moisture problems at the frequencies L, C and X band. We use the same physical parameters of rms heights and correlation lengths at the three frequencies. Results for 2D case with rms height up to 2 wavelengths at X band are shown. The hybrid UV-SMCG method for RWG basis is also used to accelerate MoM solution. Comparisons are made with SPM, KA and AIEM predictions. We also compare backscattering between horizontal and vertical polarization cases at different rms heights with exponential correlation. At small rms height, the backscattering for vertical polarization case is larger than that for horizontal polarization case. On the other hand, at large rms height, the backscattering for horizontal polarization case is larger.

Rough surface scattering effects are important problem for solving emission in microwave remote sensing. Stochastically the surface length should be infinite to simulate random rough surface scattering which is possible for analytical method but not for numerical method. Generally there are two approximations to simulate real life: the tapered wave method and periodic boundary condition. Both are valid for random rough surface scattering if convergence is shown with increase of surface length for the tapered wave and if the period is large enough for periodic boundary condition. In this paper we compare the results for the four Stokes parameters between periodic boundary conditions and the tapered wave approach.

Four radiative transfer equations for Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) bands 11, 12, 13, and 14 are built involving six unknowns (average atmospheric temperature, land surface temperature, and four band emissivities), which is a typical ill-posed problem. The extra equations can be built by using linear or nonlinear relationship between neighbor band emissivities because the emissivity of every land surface type is almost constant for bands 11, 12, 13, and 14. The neural network (NN) can make full use of potential information between band emissivities through training data because the NN simultaneously owns function approximation, classification, optimization computation, and self-study ability. The training database can be built through simulation by MODTRAN4 or can be obtained from the reliable measured data. The average accuracy of the land surface temperature is about 0.24 K, and the average accuracy of emissivity in bands 11, 12, 13, and 14 is under 0.005 for test data. The retrieval result by the NN is, on average, higher by about 0.7 K than the ASTER standard product (AST08), and the application and comparison indicated that the retrieval result is better than the ASTER standard data product. To further evaluate self-study of the NN, the ASTER standard products are assumed as measured data. After using AST09, AST08, and AST05 (ASTER Standard Data Product) as the compensating training data, the average relative error of the land surface temperature is under 0.1 K relative to the AST08 product, and the average relative error of the emissivity in bands 11, 12, 13, and 14 is under 0.001 relative to AST05, which indicates that the NN owns a powerful self-study ability and is capable of suiting more conditions if more reliable and high-accuracy ASTER standard products can be compensated.

In this paper, we study the bistatic scattering using an advanced integral equation model (AIEM). By keeping all the surface current terms in the Kirchhoff surface fields, the bistatic scattering coefficients are obtained. For simplification, the complete Kirchhoff field did not cast into the derivation of the complementary field. We compare varied updated versions of IEM-based models with the small perturbation model, geometrical optics model, and Kirchhoff approximation standard models at respective regions of validity. The results indicate that the new AIEM provides much more accurate predictions for bistatic scattering.

Radar Polarimetry, Radar Interferometry and Polarimetric SAR Interferometry represent the current culmination in active ‘Microwave Remote Sensing’ technology, but we still need to progress very considerably in order to reach the limits of physical realizability. Whereas with radar polarimetry the textural fine-structure, target orientation, symmetries and material constituents can be recovered with considerable improvement above that of standard ‘amplitude-only’ radar; by implementing ‘radar interferometry’ the spatial (in depth) structure can be explored. With Polarimetric Interferometric Synthetic Aperture Radar (POL-IN-SAR) imaging, it is possible to recover such co-registered textural and spatial information from POL-IN-SAR digital image data sets simultaneously, including the extraction of Digital Elevation Maps (DEM) from either Polarimetric (scattering matrix) or Interferometric (dual antenna) SAR systems. Simultaneous Polarimetric-plus- Interferometric SAR Imaging offers the additional benefit of obtaining co-registered textural-plus-spatial three-dimensional POL-IN-DEM information, which when applied to Repeat-Pass Image-Overlay Interferometry provides differential background validation and environmental stress-change information with highly improved accuracy. However, hitherto only single-polarization-channel repeat-pass IN-SAR was developed and considered; and therefore the aim of this paper is to scrutinize and determine why and for what specific problems fully polarimetric differential RP-POL-IN-SAR imaging is required.

In this paper, we present a parallel computing technique, referred to as parallel positive Boolean function (PPBF), for supervised classification of multispectral images. The approach is based on the generalized positive Boolean function (GPBF) scheme, which has been successfully applied in multispectral image classification. The GPBF classifier is developed from a stack filter. The stack filter is defined as the class of all nonlinear digital filters. Each stack filter corresponding to a GPBF possesses the weak superposition property and the ordering property. In order for the GPBF to be effective, the proposed PPBF is performed to improve the computational speed by using parallel cluster computing techniques. It creates a set of stack filters in each parallel node implemented by message passing interface (MPI). The proposed PPBF technique reduces the structure complexity of original GPBF. The effectiveness of the proposed PPBF is evaluated by fusing Systeme Pour l’Observation de la Terre (SPOT) images and digital elevation model (DEM) information for land cover classification during the post 921 Earthquake period in Taiwan. The experimental results demonstrated that PPBF not only significantly improves the computational loads of GPBF classification, but also substantially improves the precision of classification compared to conventional classification.

Microwave scattering from snow is difficult to model due to the complexity and heterogeneity of natural snow. In this paper, we developed a multi-layer, multi-scattering model based on recent theoretical advances in snow and surface modeling. In the proposed multi-layer model, Matrix Doubling method is used to account for scattering from each snow layer; and Advanced Integral Equation Model (AIEM) is incorporated into the model to describe surface scattering. Comparisons were made between the model predictions and field observations from truck-mounted L- and Ku-band scatterometers (frequencies are 1.25 GHz and 15.5 GHz) at Local-Scale Observation Site (LSOS) of NASA Cold- land Processes Field Experiment (CLPX) during Third Intensive Observation Period (IOP3). It was found that model predictions were in good agreement with field observations with proper particle size selected. Analysis on scatterer shape, multiple scattering and snow stratification effects were also made based on model simulations.

The IEM model proposed in 1992, for rough surface scattering has been extensively applied for microwave remote sensing of terrain. Model validation has been made by experimental measurements and by numerical simulations. However, estimation accuracy was verified mostly for like-polarizations in the monostatic configuration. Since then, much effort has been devoted to improving the model performance. A recent AIEM(Advance IEM) incorporated these improvements and demonstrated significant enhancement in accuracy. Analysis was also extended to bistatic scattering for a wide range of surface parameters. In this study, we recognized that both theoretically and experimentally, the modified Stokes vector (Kennaugh matrix), including polarization correlation terms, are very useful to retrieve surface parameters such as wind field. Consequently, we extend the AIEM model to include all polarization correlation terms in order to understand the dependence of four Stokes terms on geophysical surface parameters, such as roughness, correlation length, and dielectric constant. Sensitivity analysis is also performed. Another useful aspect is in the physical interpretation of polarimetric returns as a function of azimuthal angle. Potential applications to soil moisture estimation will be illustrated and discussed.

The IEM model proposed in 1992, for rough surface scattering has been extensively applied for microwave remote sensing of terrain. Model validation has been made by experimental measurements and by numerical simulations. However, estimation accuracy was verified mostly for like- polarizations in the monostatic configuration. Since then, much effort has been devoted to improving the model performance. A recent AIEM(Advance IEM) incorporated these improvements and demonstrated significant enhancement in accuracy. Analysis was also extended to bistatic scattering for a wide range of surface parameters. In this study, we recognized that both theoretically and experimentally, the modified Stokes vector (Kennaugh matrix), including polarization correlation terms, are very useful to retrieve surface parameters such as wind field. Consequently, we extend the AIEM model to include all polarization correlation terms in order to understand the dependence of four Stokes terms on geophysical surface parameters, such as roughness, correlation length, and dielectric constant. Sensitivity analysis is also performed. Another useful aspect is in the physical interpretation of polarimetric returns as a function of azimuthal angle. Potential applications to soil moisture estimation will be illustrated and discussed.

Earthquakes or torrential rains often lead to landslides. The Chichi earthquake in 1999 struck Nantou County of central Taiwan and caused civilian casualties of more than 2,400. The earthquake also turned the disaster areas into the sites of tens of thousands of landslides and rock avalanches. Added with the annual April–June Mei-Yu rain season and July–September typhoon season, Nantou County is further victimized by these elemental factors because it is one of the most affected areas during those seasons. With respect to the worsening scenario, this project attempts an in-depth look at the most damaged regions, using the high resolution SPOT satellite images taken in Septembers of 1999, 2002 and 2005. Landslide prediction is of spatial and temporal concerns, since these targeted sites cannot be thoroughly understood without knowledge of its geological past. This study thus examines the prospective changes in the entire measures of area of landslide sites in Nantou using Markov’s statistics model. The purpose is to develop a comprehensive methodology for conducting a random prediction of spatial and environmental factors in response to the transitory processes of the landslide areas. With the application of transition matrices derived from Markov’s model in the changes in future landslide coverage, it is estimated that the ratio of landslide will gradually stabilize from 1.02% in 2002 and 0.93% in 2005 to 0.82% in 2032. And from the analysis of Logit model, it is clear that the landslides are closely related to cumulated rainfall, altitude, slope gradients and geological formations, and in particular slope gradients having the greatest effect.

Because of its all-weather and all-time characteristics, SAR images are particularly effective to monitor disaster events. When a disaster occurs, the image acquired from a SAR sensor changes dramatically. As a result, damage estimation for natural and human-made disasters from SAR images can be achieved by applying a change detection technique. Theoretically, SAR signals can be characterized as a chaotic phenomenon because that the scattering signals within a resolution cell are summed up coherently. Accordingly, SAR signal can be represented by a spatial chaotic model and characterized by its fractal dimension. In this paper, based on the spatial chaotic model, a simplified SAR image change detection procedure is proposed. The proposed method is then applied to estimate the flood-damage area caused by flooding events. Experimental results reveal that the proposed method is an effective and efficient tool for damage estimation from SAR images.

The use of remote sensing images from various sensors is supposed to be able to improve classification accuracies. In this paper, a multiple classifiers system is adopted to fully utilize the complementary information among different data sources. A weighting policy may be applied to fuse knowledge acquired by classifiers according to their classification performances. Based on the past researches, there are some kinds of complex relationship among the classifiers' outputs. It is believe that the classification accuracy will be further improved if these relationships could be modeled properly. Therefore, a neural decision maker is proposed to express their relationships and to determine their weights among classifiers' outputs. Another type of the multisource classifier, neural networks approach, is also introduced. The classification performances of utilizing various multisource classifiers, i.e. neural network approach, multiple classifiers systems weighted by y the conventional Bagging and Boosting algorithms and the proposed method, to the application of multisource remote sensing images classification/ data fusion are demonstrated and compared. Experimental results show that both the neural networks approach and multiple classifiers system can dramatically improve the classification accuracy. In addition, the classification performance of the proposed method is better than that of using neural networks approach. Moreover, the proposed method outperforms the multiple classifiers systems weighted by the conventional Bagging and/ or Boosting algorithms.

The prevailing complex geological and ecological conditions of Taiwan have drawn considerable attention from various geo-ecological communities because of their vulnerability to produce various natural hazards at different scales. Located in the tropical/subtropical zone of the Pacific Rim, its ecological and rugged mountainous properties are environmentally sensitive making monitoring and observations especially difficult because of the high population density. For example, in terms of natural hazard mitigation tectonically active regions are used for analyzing the cause of abundant risk events, such as earthquakes, landslides and land subsidence. In fact Taiwan is well suited as a test site for studying those geologically disastrous processes. Implementing novel techniques of space remote sensing has proved to be an effective means in recent years for greatly improving our understanding of these phenomena. In this paper we report on the monitoring of such events using multi-modal polarimetric and/or interferometric SAR images at C and L band from ERS, JERS-1, RADARSAT-1, ENVISAT, and from the recent ALOS satellite. For crustal and surface deformation, we used radar image pairs with long temporal baselines and large areas of coverage for investigating deformation over Western Taiwan. Pre-seismic and co-seismic deformation patterns are spatial-temporally analyzed. The other topic deals with the coastline changes observed from a sequence of ERS-1/2 SAR images within the years of 1996 to 2005. Waterlines were extracted using multi-scale procedures of edge detection and were corrected with tidal motion data. Substantial analyses were carried out in conjunction with ground surveys and lidar mapping. The topographic feature changes due to large scale landslides triggered by torrential rains were also monitored. In addition, the SAR interferograms were used to analyze the deposition changes along the riverbeds and riverbanks for short-intervals using optimal baselines. Summary and remarks on the implementation of such multi-modal polarimetric and/or interferometric SAR imagery for environmental monitoring are provided.

The prevailing complex geological and ecological conditions of Taiwan have drawn considerable attention from various geo-ecological communities because of their vulnerability to produce various natural hazards at different scales. Located in the tropical/subtropical zone of the Pacific Rim, its ecological and rugged mountainous properties are environmentally sensitive making monitoring and observations especially difficult because of the high population density. For example, in terms of natural hazard mitigation tectonically active regions are used for analyzing the cause of abundant risk events, such as earthquakes, landslides and land subsidence. In fact Taiwan is well suited as a test site for studying those geologically disastrous processes. Implementing novel techniques of space remote sensing has proved to be an effective means in recent years for greatly improving our understanding of these phenomena. In this paper we report on the monitoring of such events using multi-modal polarimetric and/or interferometric SAR images at C and L band from ERS, JERS-1, RADARSAT-1, ENVISAT, and from the recent ALOS satellite. For crustal and surface deformation, we used radar image pairs with long temporal baselines and large areas of coverage for investigating deformation over Western Taiwan. Pre-seismic and co-seismic deformation patterns are spatial-temporally analyzed. The other topic deals with the coastline changes observed from a sequence of ERS-1/2 SAR images within the years of 1996 to 2005. Waterlines were extracted using multi-scale procedures of edge detection and were corrected with tidal motion data. Substantial analyses were carried out in conjunction with ground surveys and lidar mapping. The topographic feature changes due to large scale landslides triggered by torrential rains were also monitored. In addition, the SAR interferograms were used to analyze the deposition changes along the riverbeds and riverbanks for short-intervals using optimal baselines. Summary and remarks on the implementation of such multi-modal polarimetric and/or interferometric SAR imagery for environmental monitoring are provided.

In microwave remote sensing of land surfaces, the surfaces with exponential correlation functions have become a common choice in recent years. The use of Gaussian correlation function is not appropriate for land surfaces because the computed backscattering coefficients are many decibels below that of measurements. However, the numerical simulations of Maxwell equations were performed using Gaussian correlation functions. In the past, regimes of validity were established for analytic methods using numerical 2-D simulations. However, these past efforts of numerical tests to establish regimes of validity were for Gaussian correlation functions. The conclusions of these past numerical tests are not valid for exponential correlation functions. In this paper, we report on the polarimetric active and passive microwave remote signatures for exponential correlation function surfaces. Comparisons are made with analytic theory such as small perturbation method (SPM), Kirchhoff approximation (KA) and Advanced Integral Equation Model (AIEM). We particularly emphasize on the case of moderate rms heights of the order of 1 wavelength. This is particularly important for X band scattering from land surfaces. Numerical results are illustrated for bistatic scattering and emissivities as functions of frequencies, incidence and scattering angles and polarization for cases of interests in microwave remote sensing. We also compare backscattering between horizontal and vertical polarization cases at different rms heights with exponential correlation function. At small rms height or small slope, the backscattering for vertical polarization case is larger than that for horizontal polarization case. On the other hand, at large rms height or large slope, the backscattering for horizontal polarization case is larger.

The rough surfaces in Greenland are exhibited as sastrugi. The roughness heights are less than 8 cm for much of the year except in late winter and spring, when they increase to 25 cm or less. Roughness profiles were also related to snow and firn ventilation. WindSat, launched in January 2003, was the first spaceborne polarimetric radiometer to measure all 4 elements of Stokes vector, viz., the vertical polarized brightness temperatures, the horizontal polarized brightness temperatures, and the real and imaginary part of the cross-correlations of the vertical and horizontal polarizations. It was shown by Tsang (1984, 1990) that azimuthal asymmetry will create nonzero third and fourth Stokes parameter in passive microwave remote sensing. Thus the third and fourth Stokes parameters contain information of the azimuthal structure. Usually the third and fourth Stokes parameters are quite small between 0.5 K to 1 K over land and less than plusmn2.5 K over ocean. However, measurements of third and Stokes parameters over Greenland show surprising values of 10 K for the third Stokes parameter and between -10 K and 20 K for the fourth Stokes parameter. In this paper, we use physically based electromagnetic model to study the passive polarimetric remote sensing of snow in Greenland by consider the scattering and emission from a random rough surface over multi-layered media. We consider the random rough surface varied in only one horizontal direction so that azimuthal asymmetry exists in the 3-D problem. Dyadic Green's functions of multilayered medium (Tsang et al., 2000) is used to formulate the surface integral equation so that the polarization dependence of emission and scattering is accounted for systematically. The surface integral equations are solved by using the method of moments in conjunction with fast numerical algorithms such as the multilevel UV method. Numerical results of brightness temperatures are illustrated for all four Stokes parameters to demonstrate the signatures of sastrugi in passive microwave remote sensing. To account for the large third and fourth Stokes parameters, we also consider the case of anisotropic scatterers in volume scattering. Full multiple volume scattering are studied with numerical solutions of the radiative transfer equations for non-spherical scatterers with preferred orientation.

In this paper, a novel technique is proposed for a supervised classification of multisource images for the purpose of landslide hazard assessment. The method, known as the generalized positive Boolean function (GPBF), is developed for land cover classification based on the fusion of remotely sensed images of the same scene collected from multiple sources. It presents a framework for data fusion of multisource remotely sensed images, which consists of two approaches, referred to as the band generation process (BGP) and the positive Boolean function (PBF) classifier. The PBF classifier developed from a stack filter has been successfully applied in hyperspectral image classification. For the PBF to be effective for multispectral images, a multiple adaptation BGP is introduced to create a new set of additional bands especially accommodated to landslide classes. These bands include nonlinear normalized difference vegetation index data and morphological information in the form of digital elevation model (DEM)-derived slope values that originate from multiple sources. The performance of the proposed method is evaluated by fusing Systeme Pour l'Observation de la Terre images and DEM information for land cover classification during the post 921 Earthquake period in Taiwan. Experimental results demonstrate the proposed GPBF multiclassification approach is suitable for land cover classification in Earth remote sensing and improves the precision of image classification compared to conventional classifiers

The 19 articles in this special issue focus on remote sensing for major disaster prevention, monitoring, and assessment. Topics include earthquakes and landslides, tsunami, hurricanes and typhoons, floods and fires, as well as papers with a broader focus, highlighting innovative tools and procedures to exploit Earth observation data.

In this paper, predictions from the small-slope approximation (SSA), the advanced integral equation model (AIEM), and the reduced third-order local curvature approximation (RLCA3) theories of rough surface scattering are compared with each other and with predictions obtained from numerical simulations for dielectric surfaces with Gaussian and exponential correlation functions. A discussion of the results obtained is provided, along with plans for continued investigations.

It is very difficult to detect changes from SAR images because of two major difficulties associated with SAR, which are the removal of speckle noise and the registration of information between images. Speckle is a chaotic phenomenon because that the scattering signals within a resolution cell are summed up coherently. Therefore, SAR signal can be modeled by a spatial chaotic system and characterized by its fractal dimension. Then, simplified procedures for SAR image change detection are proposed because that the process of image despeckling is unnecessary. The proposed approach is applied to multitemporal polarimetric SAR images for change detections. The experimental results of using a simple image difference (DI) technique, the principal component analysis (PCA), and the proposed (Fractal) approach are compared. The effects of misregistration for different approaches are also presented. Simulation results reveal that misregistration affects less and less as SNR is increased. When SNR is low, by using DI or PCA methods, the overall performance of change detection is degraded by spurious differences due to misregistration. On the contrary, Fractal method can tolerate misregistration effect at low SNR. In addition, when the difference between changed classes is small, it is fail to detect changes by using of either DI or PCA method. In contrast, the Fractal method can still effectively detect land cover changes.

A novel method, which employs both a matrix doubling algorithm and the first-order solution of a radiative transfer (RT) equation for modeling microwave backscattering from forest, is presented in the paper. The method is based on the assumption that a forest canopy can be divided into a number of distinct horizontal vegetation layers over a dielectric half-space rough surface. The scattering phase matrix of each layer is calculated by either matrix doubling to account for the multiple-scattering effect or first-order solution of an RT equation, depending on the scattering characteristics of the layer. The first-order solution of the RT equation is used for the trunk layer while the matrix doubling technique is applied to both the crown layer and understory. The advanced integral equation model and reflectivity matrix are used to calculate the noncoherent and coherent surface boundary conditions. Comparisons between model predictions and field measurements on radar backscattering coefficients for a walnut orchard showed a good agreement at both L-band and X-band and for all three polarizations. Comparative analyses of model predictions for backscattering from a forest medium calculated using the combined model, first-order RT model, and the standard matrix doubling model were also presented. Understory effects, that can significantly change the weight of each scattering mechanism, were also evaluated by using the combined method.

A physically based bare-surface soil moisture inversion technique for application with passive microwave satellite measurements, including the Advanced Microwave-Scanning Radiometer-Earth Observing System, Special Sensor Microwave/Imager, Scanning Multichannel Microwave Radiometer, and Tropical Rainfall Measuring Mission Microwave Imager, was developed in this paper. The inversion technique is based on the concept of a simple parameterized surface emission model, the Qpmodel, which was developed using advanced integral equation model simulations of microwave emission. Through evaluation of the relationship between roughness parameters Qpat different polarizations, it was found that they could be described by a linear function. Using this relationship and the surface emissivities measured from two polarizations, the effect of the surface roughness is cancelled out. In other words, this approach consisted in adding different weights on the v and h polarization measurements so as to minimize the surface roughness effects. This method leads to a dual-polarization inversion technique for the estimation of the surface dielectric properties directly from the emissivity measurements. For validation, we compared the soil moisture estimates, derived from ground radiometer measurements at C- to Ka-band obtained from the Institute National de Recherches Agronomiques' field experimental data in 1993 and the Beltsville Agricultural Research Center's field experimental data at C- and X-band obtained in 1979-1982, with the field in situ soil moisture measurements. The accuracies [root-mean-square error (rmse)] are higher than 4% for the available experimental data at the incidence angles of 50deg and 60deg. The newly developed inversion technique should be very useful in monitoring global soil moisture properties using the currently available satellite instruments that commonly have incidence angles between 50deg and 55deg

This study describes a semi-empirical bare surface emission model for AMSR-E. Through evaluation of a bare surface emission database generated by the Advanced Integral Equation Model (AIEM) for a wide range of surface dielectric and roughness properties under AMSR-E, we developed a new semi-empirical multi-frequency-polarization surface emission model - the Qp model. This model relates the effects of the surface roughness on the emission signals through the roughness variable Qp at different polarization - p (v or h). The Qp can be simply described as a single surface roughness property of the random surface slope - S. The comparison of the simulations by the Qp and AIEM models indicated that the error is extremely small, its magnitude is only as 10 -3 . The evaluation of this model with the field experimental data also showed a very good agreement. We will show its validation with the field ground radiometer measurement and its application in estimation of soil moisture.

Abstract : Surface can be characterized in terms of its material (dielectric) and geometric properties. The dielectric properties of the surface are expressed primarily by its moisture content, while the roughness describes the geometric characteristics of surface. Various techniques for information retrieval from remotely sensed data have been proposed in a number of recent studies. Some of them are based on an empirical relationship between the measured return signals and the ground truth. Because of their development from a limited number of observations, these models are generally valid only for the conditions under which those measured data were taken. These models also appear that no dependence on the roughness parameter, l-correlation length. In this work, the potential of using the polarimetric SAR data over surface scatterers in order to invert surface parameters is investigated. The model-based and image-based inversion schemes are investigated and compared; the former is doing retrieval from a dynamic learning neural network trained with the Advanced Integral Equation Model, while the latter is schemed from a decomposition of coherency matrix. In model based approach, only the surface scattering term of total return is used in order to remove the vegetation effects. The image based approach accounts for non-zero cross-polarized, backscattering as well as depolarization by three polarimetric parameters, namely the scattering entropy(H), the scattering anisotropy(A), and the alpha angle(alpha). The features of these two schemes are discussed in terms of numerical aspects and physical implications of the surface parameters being inverted by using experimental E-SAR L-band data . We also show the performances of inversion and discuss the advantages and drawbacks of both schemes.

Taiwan Island is of mountainous with high density of population concentrated on a narrow belt of western plain. Human activities are pushing to move toward hillside and even mountains after an overdeveloping of flat plain. These include local community, agricultural zone, golf course, and road network, among others. The competition of land use with nature leads to landscape change to dramatic degree. The complex geological setting of is prone to landslide and soil erosion triggered by torrent storms if soil and water conversation are not well cared. The 921 Great Earthquake further made the soil even more vulnerable to slide and collapse. This has been seen from floods in 2004 which caused large scale damages in middle Taiwan. Hence, monitoring of hillside change becomes critical for effective and efficient land management. These changes are of spatial and temporal variability. In this paper, we analyze the statistical properties of land cover changes detecting by SPOT images from two years of continuous observations and monitoring, with ancillary data from base maps, land use maps, GIS data, and DTM. All the data handling and analysis are through a GIS system. To assess the detection accuracy, a total of well distributed 206 samples from the whole island are selected for verification. Those changed areas are then analyzed in terms of occurrence frequency associated with location and time. Spatially, the occurrence frequency is high for slope between 15%-30% with altitude of around 1500m. The occurrence frequency was also dependent on the geological risk sites and strongly related to landslide-prone sites. Preliminary, it was found that the correlation between space and time is weak. Longer observation in time may be necessary and is undergoing.

Taiwan Island poses very rough and steep terrain along the Central Range to east coast with only about 30% of plain area along the west coast. It suffers from strikes from typhoon and flush storms almost every year. For example, the Mindulle typhoon in July 2004 carried heavy rainfalls and raged viciously to the several counties in Middle Taiwan. The aim of this paper was to understand and document the caused landslides by means of a model using satellite images in order to assess the potential risk, to build up the attribution properties, and to compare with the past analysis result. It is hoped that we may improve our understanding about the status and trend of the catchments, drainage network, and watersheds. From the analysis results, it was found that the model to extract landslide is workable, and when the accumulation rainfall is over 1600mm, the new developed landslide was obviously added. The tatal landslide is 23,748.37 hectares, only 11 %( 2607.23 hectares) belong to mountainside range, the others were part of forest area. In terms of the scale by numbers of landslide, mostly were fallen between 0.1 to 0.5 hectares, or about 33.4% in total; if terms of area, areas between 2 to 10 hectares had 33% of the total; gradient at most number is 53% between 30 with 45 degree. It was also found that most landslides were in between 45 to 60 degrees and 25% of them were between 1,500 to 2,000 meters in altitude. It was worth mentioning that Nantou County, the 921 great earthquake site, wrap up a total of 13,270 hectares, increasing 1,792 hectares from previous report.

This study demonstrates the capability of estimating soil moisture using multi-polarization L-band backscattering coefficients. It shows an algorithm development for estimation of bare surface soil moisture and roughness. It is found that the surface rms height can be estimated quite well with only co- polarized backscattering signals. For estimation of the surface dielectric properties, the accuracy can be significantly improved by using all three polarization signals.

Abstract – This study demonstrates a technique of estimating soil moisture using the combined passive/active L-band microwave measurements. It shows 1) evaluation of the small albedo assumption for using dual polarization passive measurements, 2) development of a synthesized technique to estimate soil moisture, and 3) evaluation with ground soil moisture measurements from the SMEX02 experiment data. I.

This study develops a parameterized bare surface emission model for the applications in analyses of the passive microwave satellite measurements from the Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E). We first evaluated the capability of the advanced integral equation model (AIEM) in simulating wide-band and high-incidence surface emission signals in comparison with INRA's field experimental data obtained in 1993. The evaluation results showed a very good agreement. With the confirmed confidence, we generated a bare surface emission database for a wide range of surface dielectric and roughness properties under AMSR-E sensor configurations using the AIEM model. Through the evaluations of the commonly used semiempirical models with both the AIEM simulated and the field experimental data, we developed a parameterized multifrequency-polarization surface emission model-the Qp model. This model relates the effects of the surface roughness on the emission signals through the roughness variable Qp at the polarization p. The Qp can be simply described as a single-surface roughness property-the ratio of the surface rms height and the correlation length. The comparison of the emissivity simulations by the Qp and AIEM models indicated that the absolute error is extremely small at the magnitude of 10/sup -3/. The newly developed surface emission model should be very useful in modeling, improving our understanding, analyses, and predictions of the AMSR-E measurements.

We evaluate the capability of the microwave emission model that including the Dense Media Radiative Transfer Model (DMRT) and AIEM for simulation of dry snow emissivity. We compared the model predictions with the ground experimental measurements. The comparison shows our snow microwave emission model agrees quite well with the measurements.

In this study, we used SPOT images over years of 2002 and 2003 each year covering two different seasons, to investigate the vegetation coverage over the whole Taiwan island. Attention was paid to evaluate the impacts of seasonal effects. Comparatively, the NDVI over the forested areas was stable, except the landslide sites. From this study, we can see that using the constellation of SPOT satellites to improve the observation frequency is essential to the useful monitoring of vegetation coverage.

Through analyses of the model simulated database, we developed a technique to estimate surface soil moisture under HYDROS radar sensor (L-band multipolarizations and 40deg incidence) configuration. This technique includes two steps. First, it decomposes the total backscattering signals into two components - the surface scattering components (the bare surface backscattering signals attenuated by the overlaying vegetation layer) and the sum of the direct volume scattering components and surface-volume interaction components at different polarizations. From the model simulated data-base, our decomposition technique works quit well in estimation of the surface scattering components with RMSEs of 0.12, 0.25, and 0.55 dB for VV, HH, and VH polarizations, respectively. Then, we use the decomposed surface backscattering signals to estimate the soil moisture and the combined surface roughness and vegetation attenuation correction factors with all three polarizations

The integral equation approach to modeling scattering from rough surfaces was introduced in 1992. At that time, it was noted that there was a need to find a transition reflection coefficient that could change its argument from the incident angle to the specular angle as frequency or roughness scale got large. One such reflection coefficient was published in 2001. In this letter, we would like to include this reflection coefficient in the integral equation model to interpret several multifrequency backscattering measurements from surfaces with surface parameters defined by the investigators who acquired the data.

An integral equation method (IEM) surface scattering model was examined in terms of its applicability to laboratory measurement and numerical simulations. New expressions for both single scattering and multiple scattering were obtained by rederiving the scattering coefficient to keep all the phase terms in the spectral representation of the Green's function. After quite intricate mathematical manipulations, a fairly compact form is obtained for the scattering coefficients. In addition, the Fresnel reflection coefficients used in the model were replaced by a transition function that takes surface roughness and permittivity into account. The results of comparisons with both the numerical simulations and measurements for the backscattering case indicate that the IEM is improved, becoming more accurate and practical to use.

This study demonstrates the capability of estimating soil moisture using repeat-pass L-band radiometer. It shows (1) evaluation of the effects of the surface roughness and vegetation in the repeat-pass measurements and (2) development of a technique to estimate soil moisture.

Presents a method, based on a fuzzy neural network, that uses fully polarimetric information for terrain and land-use classification of synthetic aperture radar (SAR) image. The proposed approach makes use of statistical properties of polarimetric data, and takes advantage of a fuzzy neural network. A distance measure, based on a complex Wishart distribution, is applied using the fuzzy c-means clustering algorithm, and the clustering result is then incorporated into the neural network. Instead of preselecting the polarization channels to form a feature vector, all elements of the polarimetric covariance matrix serve as the target feature vector as inputs to the neural network. It is thus expected that the neural network will include fully polarimetric backscattering information for image classification. With the generalization, adaptation, and other capabilities of the neural network, information contained in the covariance matrix, such as the amplitude, the phase difference, the degree of polarization, etc., can be fully explored. A test image, acquired by the Jet Propulsion Laboratory Airborne SAR (AIRSAR) system, is used to demonstrate the advantages of the proposed method. It is shown that the proposed approach can greatly enhance the adaptability and the flexibility giving fully polarimetric SAR for terrain cover classification. The integration of fuzzy c-means (FCM) and fast generalization dynamic learning neural network (DLNN) capabilities makes the proposed algorithm an attractive and alternative method for polarimetric SAR classification.

This paper presents a model of microwave emissions from rough surfaces. We derive a more complete expression of the single-scattering terms in the integral equation method (IEM) surface scattering model. The complementary components for the scattered fields are rederived, based on the removal of a simplifying assumption in the spectral representation of Green's function. In addition, new but compact expressions for the complementary field coefficients can be obtained after quite lengthy mathematical manipulations. Three-dimensional Monte Carlo simulations of surface emission from Gaussian rough surfaces were used to examine the validity of the model. The results based on the new version (advanced IEM) indicate that significant improvements for emissivity prediction may be obtained for a wide range of roughness scales, in particular in the intermediate roughness regions. It is also shown that the original IEM produces larger errors that lead to tens of Kelvins in brightness temperature, which are unacceptable for passive remote sensing.

The detection of the wake can provide substantial information about a ship, such as its size, direction, and speed of movement. In general though, ship-generated wakes in synthetic aperture radar images are associated with high sea clutter, which will cause some deterioration in the detection performance. Therefore, a wavelet correlator, based on an orthogonal basis function, is adopted. Three highpass images - horizontal, vertical, and diagonal direction - are generated for each resolution scale, followed by a process to correlate among the moduli of different scale modulus images formed from the three highpass images. The output of the correlation process is highly representative at the ship's wake edges. Comparisons with other methods indicate the superior performance of the present approach, in that not only can the wakes be detected, but their V-shaped pattern is well preserved. The second stage involves the application of the Radon transform technique to an estimation of the V-opening angle from the detected ship wakes. Ship-generated wake edges are found to be the local maxima in the wavelet transform method of several adjacent scales, and hence, the wake edge will be enhanced in the reconstructed data. The background noise is also greatly reduced. In particular, the process of spatial correlation is found to be critical. Compared to a direct Radon transform, the proposed scheme is demonstrated to be much more effective in terms of efficiency, as well as reliability, for ship wake detection in noisy backgrounds.

In this study, we evaluate the effect of the surface roughness on estimation of the relative soil moisture change in repeat-pass L-band radar measurements. It has found the surface roughness has a significant impact and a correction technique has been developed. I. INTRODUCTION During recent years, theoretical modeling and field experiments have established the fundamentals of active microwave remote sensing as an important tool in determining physical properties of soil. In attempt to use active microwave remote sensors in estimation of soil moisture, we are mainly facing two major problems: effects of surface roughness and vegetation cover. There are several algorithms developed for measurement of bare soil moisture quantitatively using dual or three polarization L-band SAR image data. A common idea beyond these algorithms is to separate the effects of the surface dielectric and roughness properties on the backscattering signals to present the model, which the inversion was based on, as a product of a dielectric function and a roughness function. They are first-order statistical inversion models. Depending on the data source, the selection of the surface roughness parameters and the backscattering measurements of the different polarizations or their linear combinations, the models have a great difference in terms of both the dielectric and roughness functions. The temporal variability of surface roughness is generally at much longer time scale than that of soil moisture, unless there was a human activity. Commonly, we can reasonably assume that the surface roughness is same at certain time interval. The change in SAR measurements between the repeat-passes, therefore, is resulted from the change of ground dielectric properties or soil moisture. Therefore, the repeat-pass measurements provide additional relative surface soil moisture change information and make it possible to directly estimate the relative moisture change and improving the accuracy of estimating the bare surface soil moisture. However, there has no quantitative algorithm being developed to estimate the relative soil moisture change using repeat-pass measurements. In this study, we evaluate 1) the effects of surface roughness in L-band repeat-pass measurements using IEM simulated data, 2) developing a quantitative algorithm to estimate relative soil moisture change, and 3) validating this technique with JPL/AIRSAR 92's experiment data over the little Washita test site.

This study demonstrates the capability of estimating the relative soil moisture change using repeat-pass L-band radiometer. It shows 1) evaluation of the effects of the surface roughness and vegetation in the repeat-pass measurements, 2) development of a technique to estimate the relative soil moisture change, and 3) validation with the ground soil moisture measurements from SGP99 experiment.

The original IEM surface scattering model used a simplified surface current estimate leading to relatively simple but accurate results for forward and backscattering configurations. Since then other estimates of the surface current based upon the same set of integral equations have appeared in the literature. A major reason for considering a more complex estimate is because in the original IEM model the phase in the Green's function was not included in the integration process over the surface current to find the scattered field. Thus, it is not applicable to multiple scattering calculations. Currently, there are three different modifications suggested by different investigators: (1) use of the phase factor in the Green's function of the upper medium for integration over surface current, (2) use of the phases in the Green's function in both the upper and lower medium for integration over surface current, and (3) in addition to (2) further modify the Fresnel reflection coefficient to be the sum of reflection coefficients evaluated at the incident and scattering angles divided by two. In this paper we want to compare model predictions based on the use of the above surface current estimates under backscattering and bistatic conditions.

The backscattering coefficient and emissivity of wet soil surface are studied with the 3-dimensional numerical simulations. The study is focused on the angular, frequency, and polarimetric dependence of the scattering and emission. The simulation results are compared with the experimental measurements from real-life soil surface for backscattering coefficients at L and C frequency bands and at the multi-incidence angles. The fairly good agreements are observed for the fixed physical surface roughness parameters at two frequency bands.

A generalized power law spectrum is proposed to describe the random rough surfaces in this paper. The parameters of the spectrum are related to the traditional physical parameters of root mean square (rms) height and correlation length. It can naturally reduce to the spectra of Gaussian and exponential correlation functions. The corresponding correlation functions are also derived. It can provide wider range of spectra to describe the random rough surfaces than other spectra. Based on the proposed spectrum, backscattering of soil surfaces is studied by using the integral equation model (IEM). The simulation results are compared with the experimental measurements of real soil surfaces at L, C, and X bands for the different roughness scales and moisture conditions. The reasonably good agreements between the measurements and the simulations are observed for all three-frequency bands and different incidence angles with the same sets of the physical roughness parameters.

Soil moisture is an important parameter for hydrological and climatic investigations. Future satellite missions with L-band passive microwave radiometers will significantly increase the capability of monitoring Earth's soil moisture globally. Understanding the effects of surface roughness on microwave emission and developing quantitative bare-surface soil moisture retrieval algorithms is one of the essential components in many applications of geophysical properties in the complex Earth terrain by microwave remote sensing. We explore the use of the integral equation model (IEM) for modeling microwave emission. This model was validated using a three-dimensional Monte Carlo model. The results indicate that the IEM model can be used to simulate the surface emission quite well for a wide range of surface roughness conditions with high confidence. Several important characteristics of the effects of surface roughness on radiometer emission signals at L-band 1.4 GHz that have not been adequately addressed in the current semiempirical surface effective reflectivity models are demonstrated by using IEM-simulated data. Using an IEM-simulated database for a wide range of surface soil moisture and roughness properties, we developed a parameterized surface effective reflectivity model with three typically used correlation functions and an inversion model that puts different weights on the polarization measurements to minimize surface roughness effects and to estimate the surface dielectric properties directly from dual-polarization measurements. The inversion technique was validated with four years (1979-1982) of ground microwave radiometer experiment data over several bare-surface test sites at Beltsville, Maryland. The accuracies in random-mean-square error are within or about 3% for incidence angles from 20/spl deg/ to 50/spl deg/.

L-band brightness predicted by the land surface process/radiobrightness (LSP/R) model for prairie grassland appears to be somewhat lower than expected. A crucial reason for the underestimate of the L-band brightness is that the soil surface was treated as smooth. In this paper, surface scattering of the soil determined by the IEM model is incorporated into the LSP/R model to examine its impact on the predicted L-band brightness. Eight sets of surface parameters, two correlation lengths (L) of 3 and 6 cm/spl times/4 root mean squared (RMS) heights (/spl sigma/) of 0.3, 0.6, 0.8, and 1.0 cm, are utilized to characterize the emission of the soil surface. It is found that H-polarized, L-band brightness is expectedly increased by different levels for all of the eight rough surface cases compared to the smooth surface case. The increase in the average of the H-polarized, L-band brightness is by as much as 13.2 K for the case with L=3 cm and /spl sigma/=1.0 cm. In addition, L-band's sensitivity to soil moisture is found to be approximately equal with and without the scattering effects. An increase in H-polarized, L-band brightness by about 12 K at the end of a 14-day simulation by the LSP/R model is in response to a decrease in soil moisture by 7% for all of the nine cases of concern (eight rough plus one smooth soil surfaces).

In the development of wave scattering models for randomly dielectric rough surfaces, it is usually assumed that the Fresnel reflection coefficients could be approximately evaluated at either the incident angle or the specular angle. However, these two considerations are only applicable to their respective regions of validity. A common question to ask is what are the conditions under which we would choose one or the other of these two approximations? Since these approximations are basically roughness-dependent, how can we handle the in-between cases where neither is appropriate? In this paper, a physical-based transition function that naturally connects these two approximations is proposed. The like-polarized backscattering coefficients are evaluated with the model and are compared with those calculated with a moment method simulation for both Gaussian and non-Gaussian correlated surfaces. It is found that the proposed transition function provides an excellent prediction for the backscattering coefficient in the frequency and angle trends.

The authors derive the multiple scattering expression within the framework of an IEM model for rough surface scattering. The complementary field coefficients are rederived based on a new surface slope expressions which are dependent on spatial variables. This leads to a more complete expression of the multiple scattering terms, thus allowing the authors to account for multiple effects more accurately. Numerical calculations and comparisons with numerical simulation are provided to demonstrate the results.

This paper presents the reconstruction of a wind field from three-beam scatterometer measurements under the framework of a neural network. A neural network is adopted to implement the inversion of a geophysical model function (GMF) that relates the scatterometer measurements of normalized radar cross section to surface wind speed and direction. To illustrate the functionality and applicability of the neural network, a set of wind fields generated by means of the Monte Carlo simulation are used. At each sample point of the wind field, the speed and direction are simulated. Then, a GMF CMOD4 is used to synthesize the normalized radar cross section at three pointing antennas according to the ERS-1 configuration. In such a case, the neural network is constructed to model the inverse transfer function. For inputs, a pixel-based and area-based scheme are considered. The network training is accomplished by mapping input-output pairs that are randomly selected from the database of simulated wind fields. The effectiveness of the neural network as an inverse transfer function is validated. Four data sets of ERS-1 scatterometer data over the western Pacific were selected for case study. Intercomparison with other methods concludes that the use of neural network has its indispensable advantages and better retrieval accuracy can be obtained.

A biophysically-based land-surface process/radiobrightness (LSP/R) model is integrated with a dynamic learning neural network (DLNN) to retrieve the land-surface parameters from its radiometric signatures. Predictions from the LSP/R model are used to train the DLNN and serve as the reference for evaluation of the DLNN retrievals. Both horizontally polarized and vertically polarized brightnesses at 1.4 GHz, 19 GHz, and 37 GHz for an incidence angle of 53/spl deg/ make up the input nodes of the DLNN. The corresponding output nodes are composed of land-surface parameters, canopy temperature and water content, and soil temperature and moisture (uppermost 5 mm). Under no-noise conditions, the maximum of the root mean-square (RMS) errors between the retrieved parameters of interest and their corresponding reference from the LSP/R model is smaller than 28 for a four-channel case with 19 GHz and 37 GHz brightnesses as the inputs of the DLNN. The maximum RMS error is reduced to within 0.5% if additional 1.4 GHz brightnesses are used (a six-channel case). This indicates that the DLNN produces negligible errors onto its retrievals. For the realization of the problem, two different levels of noises are added to the input nodes. The noises are assumed to be Gaussian distributed with standard deviations of 1 K and 2 K. The maximum RMS errors are increased to 9.3% and 10.3% for the 1 K-noise and 2 K-noise cases, respectively, for the four-channel ease. They are reduced to 6.0% and 9.1% for the 1 K-noise and 2 K-noise cases, respectively, for the six-channel case. This is an implication that 1.4 GHz is a better frequency in probing soil parameters than 19 GHz and 37 GHz.

Recently, neural networks have been increasingly applied to remote sensing imagery classification. The conventional neural network classifier performs learning from the representative information within a problem domain on a one-pixel-one-class basis; therefore, class mixture and the degree of membership of a pixel are generally not taken into account, often resulting in a poor classification accuracy. Based on the framework of a dynamic learning neural network (DL), this communications proposes a fuzzy version (FDL) based on two steps: network representation of fuzzy logic and assignment of membership. Comparisons between the DL and FDL are made by applying both neural networks to SAR image classification. Experimental results show that the FDL has faster convergence rate than that of DL. In addition, the separability between similar classes is improved. Moreover, the classification results match better with ground truth.

A practical method for extracting microwave backscatter for terrain-cover classification is presented. The test data are multifrequency (P, L, C bands) polarimetric SAR data acquired by JPL over an agricultural area called "Flevoland". The terrain covers include forest, water, bare soil, grass, and eight other types of crops. The radar response of crop types to frequency and polarization states were analyzed for classification based on three configurations: 1) multifrequency and single-polarization images; 2) single-frequency and multipolarization images; and 3) multifrequency and multipolarization images. A recently developed dynamic learning neural network was adopted as the classifier. Results show that using partial information, P-band multipolarization images and multiband hh polarization images have better classification accuracy, while with a full configuration, namely, multiband and multipolarization, gives the best discrimination capability. The overall accuracy using the proposed method can be as high as 95% with a total of thirteen cover types classified. Further reduction of the data volume by means of correlation analysis was conducted to single out the minimum data channels required. It was found that this method efficiently reduces the data volume while retaining highly acceptable classification accuracy.

The objective of the study is to examine the ease of applicability of three scattering models. This is done by considering the time taken to numerically evaluate these models and comparing their predictions as a function of surface roughness, frequency, incident angle and polarization with the moment method solution in two dimensions. In addition, the complexity of the analytic models in three dimensions and their analytic reduction to high and low frequency regions are also compared. The selected models are an integral equation model (IEM), a full wave model (FWM), and the phase perturbation model (PPM). It is noted that in three dimensions, the full-wave model requires an evaluation of a 10-fold integral, the phase perturbation model requires a 4- and 2-fold integral while the integral equation model is an algebraic equation in like polarization under single scattering conditions. In examining frequency dependence of IEM and PPM in two dimensions numerically, the same model expression is used for all frequency calculations, it is found that both the IEM and PPM agree with the moment method solution from low to high frequencies numerically.>

The neural network learning process is to adjust the network weights to adapt the selected training data. Based on the polynomial basis function (PBF) modeled neural network that is a modified multilayer perceptrons (MLP) network, a dynamic learning algorithm (DL) is proposed. The presented learning algorithm makes use of the Kalman filtering technique to update the network weights, in the sense that the stochastic characteristics of incoming data sets are implicitly incorporated into the network. The Kalman gains which represent the learning rates of the network weights updating are calculated by using the U-D factorization. By concatenating all of the network weights at each layer to form a long vector such that it can be updated without propagating back, the proposed algorithm improves the performance of convergence to which the backpropagation (BP) learning algorithm often suffers. Numerical illustrations are carried out using two categories of problems: multispectral imagery classification and surface parameters inversion. Results indicates the use of Kalman filtering algorithm not only substantially increases the convergence rate in the learning stage, but also enhances the separability for highly nonlinear boundaries problems, as compared to BP algorithm, suggesting that the proposed DL neural network provides a practical and potential tool for remote sensing applications.>

The properties of a surface bispectrum are found by generating a skewed surface on a digital computer and then evaluating its correlation function, bicoherence function, power spectrum, and bispectrum. The bispectrum is defined to be the Fourier transform of the bicoherence function. It is found that the surface bicoherence function and its first and second derivatives must all vanish at the origin. In general, the surface bispectrum is a complex function. Its real part is centrosymmetric, just like the surface spectrum, and its imaginary part is antisymmetric. A function with the above-stated properties is introduced to represent the imaginary part of the sea surface bispectrum. The unknown parameter in this function is calibrated using a data set from the FASINEX experiment. The sea surface backscattering model is based on an integral equation model which accounts for frequency, polarization, incident angle, azimuthal angle, and wind speed. It is found that the proposed bispectrum can be used to account for the up/down wind asymmetry.>

A surface scattering model based on an approximate solution of the integral equations for the surface tangential fields has been developed for non-Gaussian distributed, finitely conducting surfaces. It consists of two parts. One part is proportional to the surface roughness spectrum, and the other to the surface bispectrum. The bispectrum part comes into the model when the third-order surface statistics are included. While the second-order statistics account for the wind directional dependence, the third-order statistics account for the dependence on the sense of direction of the wind. Thus, it is the critical part for explaining the difference between upwind and downwind observations. In general, the bispectrum is a complex quantity and the asymmetric effect of the sea surface is represented by its imaginary part. The model characteristics, such as polarization and azimuthal dependence, are illustrated through numerical calculations. The predictions of the model are compared with field measurements, and excellent agreement is obtained.>

A backscattering model for scattering from a randomly rough dielectric surface is developed. Both like- and cross-polarized scattering coefficients are obtained. The like-polarized scattering coefficients contain single scattering terms and multiple scattering terms. The single scattering terms are shown to reduce to the first-order solutions derived from the small perturbation method when the roughness parameters satisfy the slightly rough conditions. When surface roughnesses are large but the surface slope is small, only a single scattering term corresponding to the standard Kirchhoff model is significant. If the surface slope is large, the multiple scattering term will also be significant. The cross-polarized backscattering coefficients satisfy reciprocity and contain only multiple scattering terms. The difference between vertical and horizontal scattering coefficients increases with the dielectric constant and is generally smaller than that predicted by the first-order small perturbation model. Good agreements are obtained between this model and measurements from statistically known surfaces.>