ALBERT

Description: With respect to recent advances in remote sensing technologies, the spatial resolution of airborne and spaceborne sensors is getting finer, which enables us to precisely analyze even small objects on the Earth. This fact has made the research area of developing efficient approaches to extract spatial and contextual information highly active. Among the existing approaches, morphological profile and attribute profile (AP) have gained great attention due to their ability to classify remote sensing data. This paper proposes a novel approach that makes it possible to precisely extract spatial and contextual information from remote sensing images. The proposed approach is based on extinction filters, which are used here for the first time in the remote sensing community. Then, the approach is carried out on two well-known high-resolution panchromatic data sets captured over Rome, Italy, and Reykjavik, Iceland. In order to prove the capabilities of the proposed approach, the obtained results are compared with the results from one of the strongest approaches in the literature, i.e., APs, using different points of view such as classification accuracies, simplification rate, and complexity analysis. Results indicate that the proposed approach can significantly outperform its alternative in terms of classification accuracies. In addition, based on our implementation, profiles can be generated in a very short processing time. It should be noted that the proposed approach is fully automatic.

Description: In recent years, new algorithms have been proposed to retrieve maximum available information in synthetic aperture radar (SAR) interferometric stacks with focus on distributed scatterers. The key step in these algorithms is to optimally estimate single-master (SM) wrapped phases for each pixel from all possible interferometric combinations, preserving useful information and filtering noise. In this paper, we propose a new method for SM-phase estimation based on the integer least squares principle. We model the SM-phase estimation problem in a linear form by introducing additional integer ambiguities and use a bootstrap estimator for joint estimation of SM-phases and the integer unknowns. In addition, a full error propagation scheme is introduced in order to evaluate the precision of the final SM-phase estimates. The main advantages of the proposed method are the flexibility to be applied on any (connected) subset of interferograms and the quality description via the provision of a full covariance matrix of the estimates. Results from both synthetic experiments and a case study over the Torfajökull volcano in Iceland demonstrate that the proposed method can efficiently filter noise from wrapped multibaseline interferometric stacks, resulting in doubling the number of detected coherent pixels with respect to conventional persistent scatterer interferometry.

Description: Different aspects of the operational constraints of remote sensing inverse problems are thoroughly investigated by simulation studies, using a deterministic method, namely regularized total least squares (RTLS). For demonstration purposes, water vapor profiles retrievals from simulated Suomi NPP Cross-track Infrared Souder (CrIS) hyperspectral measurements are considered. Synthetic CrIS radiances are generated using a line-by-line radiative transfer model (GENSPECT) with ∼424 realistic radiosonde profiles and US 1976 standard atmosphere as inputs. These results are also compared with those from a prevalent stochastic method. Our findings show that the stochastic method, even with additional deterministic constraints (truncated singular value decomposition) applied on top of it, is often unable to produce useful retrieval results, i.e., posterior error is more than the a priori error. In contrast, RTLS is able to produce deterministically unique results according to the available information content in the measurements, which could result in a paradigm shift in operational satellite inversion.

Description: Recently, the Cassini RADAR has been used as a sounder to probe the depth and constrain the composition of hydrocarbon seas on Saturn's largest moon, Titan. Altimetry waveforms from observations over the seas are generally composed of two main reflections: the first from the surface of the liquid and the second from the seafloor. The time interval between these two peaks is a measure of sea depth, and the attenuation from the propagation through the liquid is a measure of the dielectric properties, which is a sensitive property of liquid composition. Radar measurements are affected by uncertainties that can include saturation effects, possible receiver distortion, and processing artifacts, in addition to thermal noise and speckle. To rigorously treat these problems, we simulate the Ku-band altimetry echo received from Titan's seas using a two-layer model, where the surface is represented by a specular reflection and the seafloor is modeled using a facet-based synthetic surface. The simulation accounts for the thermal noise, speckle, analog-to-digital conversion, and block adaptive quantization and allows for possible receiver saturation. We use a Monte Carlo method to compare simulated and observed waveforms and retrieve the probability distributions of depth, surface/subsurface intensity ratio, and subsurface roughness for the individual double-peaked waveform of Ligeia Mare acquired by the Cassini spacecraft in May 2013. This new analysis provides an update to the Ku-band attenuation and results in a new estimate for its loss tangent and composition. We also demonstrate the ability to retrieve bathymetric information from saturated altimetry echoes acquired over Ontario Lacus in December 2008.

Description: This paper proposes the wrapped staring spotlight (WSS) SAR imaging mode, which is a new method to extend the azimuth steering capability for phased array SAR to achieve either an improved azimuth geometric or radiometric resolution. It investigates the utility of steering directions with main lobe gains that are smaller than that of the grating lobes and exposes how these directions can be exploited. Furthermore, two methods are proposed to reduce the speckle and the image noise at once, i.e., the Look-Normalized Pattern Correction and the $Omega$ - weighting. Based on two example TerraSAR-X WSS acquisitions, the image performance of extended and point targets is discussed.

Description: Remote sensing systems equipped with multispectral and hyperspectral sensors are able to capture images of the surface of the Earth at different wavelengths. In these systems, hyperspectral sensors typically provide images with a high spectral resolution but a reduced spatial resolution, while on the contrary, multispectral sensors are able to produce images with a rich spatial resolution but a poor spectral resolution. Due to this reason, different fusion algorithms have been proposed during the last years in order to obtain remotely sensed images with enriched spatial and spectral resolutions by wisely combining the data acquired for the same scene by multispectral and hyperspectral sensors. However, the algorithms so far proposed that are able to obtain fused images with a good spatial and spectral quality require a formidable amount of computationally complex operations that cannot be executed in parallel, which clearly prevent the utilization of these algorithms in applications under real-time constraints in which high-performance parallel-based computing systems are normally required for accelerating the overall process. On the other hand, there are other state-of-the-art algorithms that are capable of fusing these images with a lower computational effort but at the cost of decreasing the quality of the resultant fused image. In this paper, a new algorithm named computationally efficient algorithm for fusing multispectral and hyperspectral images (CoEf-MHI) is proposed in order to obtain a high-quality image from hyperspectral and multispectral images of the same scene with a low computational effort. The proposed CoEf-MHI algorithm is based on incorporating the spatial details of the multispectral image into the hyperspectral image, without introducing spectral distortions. To achieve this goal, the CoEf-MHI algorithm first spatially upsamples, by means of a bilinear interpolation, the input hyperspectral image to the spatial resolution of the input multispe- tral image, and then, it independently refines each pixel of the resulting image by linearly combining the multispectral and hyperspectral pixels in its neighborhood. The simulations performed in this work with different images demonstrate that our proposal is much more efficient than state-of-the-art approaches, being this efficiency understood as the ratio between the quality of the fused image and the computational effort required to obtain such image.

Description: The standard ocean wind product from the Advanced Scatterometer (ASCAT) is retrieved on a 12.5-km grid. Ultrahigh-resolution (UHR) processing enables ASCAT wind retrieval on a high-resolution 1.25-km grid. Ideally, such a high-resolution grid allows for improved analysis of winds with high spatial variability, such as those in near-coastal regions and storms. This paper provides an analysis and validation of ASCAT UHR wind estimates to evaluate its spatial resolution and accuracy. This is done via a comparison with two other sources: buoy-measured winds in coastal regions and winds estimated from synthetic aperture radar (SAR) data over the open ocean. Near-coastal ocean measurements may be contaminated by nearby land, introducing a wind speed bias in the retrieved winds. To enable near-coastal UHR wind retrieval, we use a land contribution ratio (LCR) approach to discard ASCAT measurements with high land contamination before UHR processing and wind retrieval. Through a comparison with near-coastal buoy winds, we find that the LCR approach is appropriate for precisely controlling the tradeoff between land contamination and spatial coverage near land. We find that the resolution of the UHR data is improved over the 25-km wind product to 10 km, and likely down to 4 km in some cases. In comparing SAR and UHR winds, we find that both products have common fine-scale features and have derivative fields that match well and that the UHR product matches better the expected spectral properties of ocean winds.

Description: Copolar correlation coefficient is one of the essential polarimetric variables because it enhances the ability of weather radars to discriminate among different hydrometeor types as well as nonmeteorological scatterers. Theoretical values of the copolar correlation coefficient $(|rho_mathrm{hv}(0)|)$ are between zero and one, but due to statistical errors, estimates of $|rho_mathrm{hv}(0)|$ can take values that are outside this interval, in which case the estimates are deemed unusable. Additionally, even if a $|rho_mathrm{hv}(0)|$ estimate is valid, it can introduce errors in echo classification if not sufficiently accurate. Therefore, it is imperative to keep the number of invalid estimates at a minimum while maintaining the acceptable accuracy and precision of measurements. Herein, a technique to improve copolar correlation coefficient estimation is presented. The technique produces estimates with reduced bias which results in an increased number of valid estimates as well as improved accuracy of $|rho_mathrm{hv}(0)|$ fields. This is achieved through a combination of previously proposed $|rho_mathrm{hv}(0)|$ estimators with several novel estimators presented in this paper.

Description: The long time series of nighttime light (NTL) data collected by DMSP/OLS sensors provides a unique and valuable resource to study changes in human activities. However, its full time series potential has not been fully explored, mainly due to inconsistencies in the temporal signal. Previous studies have tried to resolve this issue in order to generate a consistent NTL time series. However, due to geographic limitations with the algorithms, these approaches cannot generate a coherent NTL time series globally. The purpose of this study is to develop a methodology to create a consistent NTL time series that can be applied globally. Our method is based on a novel sampling strategy to identify pseudoinvariant features. We select data points along a ridgeline—the densest part of a density plot generated between the reference image and the target image—and then use those data points to derive calibration models to minimize inconsistencies in the NTL time series. Results show that the algorithm successfully calibrates DMSP/OLS annual composites and generates a consistent NTL time series. Evaluation of the results shows that the calibrated NTL time series significantly reduces the differences between two images within the same year and increases the correlations between the NTL time series and gross domestic product as well as with energy consumption, and outperforms the Elvidge et al. (2014) method. The methodology is simple, robust, and easy to implement. The quality-enhanced NTL time series can be used in a myriad of applications that require a consistent data set over time.

Description: In this paper, a theoretical and experimental analysis of polarimetric synthetic aperture radar (SAR) architectures is undertaken for sea oil slick observation purposes. Reference is made to the conventional full-polarimetric (FP) SAR that is here contrasted with new-generation polarimetric SAR architectures, known as compact-polarimetric (CP) SAR. Two CP modes are considered, i.e., the hybrid-polarity and $pi/4$ modes, whose measurements are emulated from actual L- and C-band FP SAR data. Polarimetric sea surface scattering is predicted according to an extended version of the Bragg scattering model (X-Bragg) in order to point out the differences exhibited between FP and CP SAR architectures and among CP SAR modes. Theoretical predictions are then contrasted with experiments undertaken on actual polarimetric SAR data collected over well-known oil slicks and weak-damping surfactants. Results confirm model prediction, showing that differences mainly apply when polarimetric features are estimated over slick-free sea surface using different SAR architectures, with the $pi/4$ mode behaving closer to FP SAR. Although CP SAR architectures measure only a subset of the FP information content, they represent an interesting operational alternative for both detecting oil slicks and discriminating them from weak-damping surfactants.

Description: Detection of changes caused by major events—such as earthquakes, volcanic eruptions, and floods—from interferometric synthetic aperture radar (SAR) data is challenging because of the coupled effects with temporal decorrelation caused by natural phenomena, including rain, snow, wind, and seasonal changes. The coupled effect of major events and natural phenomena sometimes leads to misinterpretation of interferometric coherence maps and often degrades the performance of change detection algorithms. To differentiate decorrelation sources caused by natural changes from those caused by an event of interest, we formulated a temporal decorrelation model that accounts for the random motion of canopy elements, temporally correlated dielectric changes, and temporally uncorrelated dielectric changes of canopy and ground. The model parameters are extracted from the interferometric pairs associated with natural changes in canopy and ground using the proposed temporal decorrelation model. In addition, the cumulative distribution functions of the temporally uncorrelated model parameters, which are associated with natural changes in canopy and ground, are estimated from interferometric pairs acquired before the event. Model parameters are also extracted from interferometric SAR data acquired across the event and compared with the cumulative probabilities of natural changes in order to calculate the probability of a major event. Subsequently, pixels with cumulative probabilities greater than 75% are marked as changed due to the event. A case study for detecting volcanic ash during the eruption of the Shinmoedake volcano in January 2011 was carried out using L-band Advanced Land Observation Satellite PALSAR data.

Description: In remote sensing data exploitation, the spectral mixture analysis technique is generally used to detect the land cover materials and their corresponding proportions present in the observed scene. Traditionally, a fixed endmember spectral signature for each land cover material is used to perform the unmixing task. In the literature, some scholars have proposed performing the unmixing by taking the spectral variability into consideration. Among these spectral-variability-based unmixing approaches, multiple-endmember spectral mixture analysis (MESMA) is probably the most widely used method. However, when the number of land cover materials is large, the computational load of the MESMA method could be very heavy. In this paper, a sparse multiple-endmember spectral mixture model (SMESMM) is proposed to handle this problem. This model treats the spectral mixture procedure as a linear block sparse inverse problem. The SMESMM is first solved using a block sparse algorithm to obtain an initial block sparse solution. Then, MESMA is used to resolve the mixed pixel using the selected land cover materials, which correspond to the nonzero blocks in the solution obtained in the first step. The block sparse solution obtained in the first step can help to determine how many and which land cover materials are involved in the considered mixed pixel. This can largely decrease the number of possible candidate models for the MESMA method when the number of land cover materials is large. Experimental results on simulated and real hyperspectral data demonstrate the efficacy of the proposed method.

Description: The next generation of weather radars, which may also support other missions, is likely to be based on phased arrays that will utilize simultaneous receive beams to achieve the required update times. Some of the disadvantages of using simultaneous receive beams, such as higher two-way antenna radiation pattern sidelobes, can be mitigated by using adaptive beamforming. A majority of the existing adaptive beamforming algorithms are designed for point targets, and direct application to distributed scatterers (e.g., hydrometeors) can lead to significantly biased estimates of key radar variables. This paper presents an adaptive beamspace processing algorithm specifically designed for weather-surveillance radar applications. Through both simulated and real data, it is shown that the proposed adaptive beamspace processing algorithm can produce accurate and calibrated estimates of radar variables while also automatically rejecting interference signals.

Description: In this paper, we present a new lossy compression method for hyperspectral images that aims to optimally compress in both spatial and spectral domains and simultaneously minimizes the effect of the compression on linear spectral unmixing performance. To achieve this, a nonnegative Tucker decomposition is applied. This decomposition is a function of three dimension parameters. By employing a link between this decomposition and the linear spectral mixing model, an optimization problem is defined to find the optimal parameters by minimizing the root-mean-square error between the abundance matrices of the original and reconstructed data sets. The resulting optimization problem is solved by a particle swarm optimization algorithm. An approximate method for fast estimation of the free parameters is introduced as well. Our simulation results show that, in comparison with well-known state-of-the-art lossy compression methods, an improved compression and spectral unmixing performance of the reconstructed hyperspectral image is obtained. It is noteworthy to mention that the superiority of our method becomes more apparent as the compression ratio grows.

Description: Automatic ship detection on spaceborne optical images is a challenging task, which has attracted wide attention due to its extensive potential applications in maritime security and traffic control. Although some optical image ship detection methods have been proposed in recent years, there are still three obstacles in this task: 1) the inference of clouds and strong waves; 2) difficulties in detecting both inshore and offshore ships; and 3) high computational expenses. In this paper, we propose a novel ship detection method called SVD Networks (SVDNet), which is fast, robust, and structurally compact. SVDNet is designed based on the recent popular convolutional neural networks and the singular value decompensation algorithm. It provides a simple but efficient way to adaptively learn features from remote sensing images. We evaluate our method on some spaceborne optical images of GaoFen-1 and Venezuelan Remote Sensing Satellites. The experimental results demonstrate that our method achieves high detection robustness and a desirable time performance in response to all of the above three problems.

Description: A vicarious calibration with reference to characterized surface tarps was conducted to determine the first radiometric characteristics of KOMPSAT-3. The 6S radiative transfer model was also used by inputting various initial parameters, such as the spectral response function of KOMPSAT-3, and atmospheric and geometric conditions. Moderate-Resolution Imaging Spectroradiometer atmospheric products, such as aerosol optical depth, precipitable water, and total ozone, were used as input parameters to interpret solar radiation reflection, scattering, and absorption effects. In the first field campaign, the radiometric coefficients from each of the spectral bands were estimated by calculating the predicted radiance at sensor level and the digital number (DN) of KOMPSAT-3 based on a linear least squares fit over a range of target reflectance levels. The second field campaign measurements were also used to upgrade the KOMPSAT-3 DNs to radiance coefficients. The root-mean-square error differences between simulated radiance and measured radiance during the second field campaign for “sensor-to-itself” calibration were 2.072 W/m 2 sr (blue), 6.80 W/m 2 sr (green), 7.512 W/m 2 sr (NIR), and 5.712 W/m 2 sr (red), respectively. This highlights that radiometric calibration with tarps is a reliable method. Furthermore, the gain ratio between the first and the second one was < 5%, indicating reasonable radiometric calibration results. Additionally, cross-validation of KOMPSAT-3 with radiometrically well-calibrated Landsat-8 was performed over bright desert. Although the difference between the vicarious calibration with surface tarps and cross-validation with Landsat-8 was significant, reasonable results were obtained under close geometrical conditions, despite inherent vicarious calibration error.

Description: Anomalies usually refer to targets with a spot of pixels (even subpixels) that stand out from their neighboring background clutter pixels in hyperspectral imagery (HSI). Compared to backgrounds, anomalies have two main characteristics. One is the spectral anomaly, i.e., their spectral signatures are different from those associated to their surrounding backgrounds; another is the spatial anomaly, i.e., anomalies occur as few pixels (even subpixels) embedded in the local homogeneous backgrounds. However, most of the existing anomaly detection algorithms for HSI only employed the spectral anomaly. If the two characteristics are exploited in a detection method simultaneously, better performance may be achieved. The third-order (two modes for space and one mode for spectra) tensor representation of HSI has been proved to be an effective tool to describe the spatial and spectral information equivalently; therefore, tensor representation is convenient for exhibiting the two characteristics of anomalies simultaneously. In this paper, a new anomaly detection method based on tensor decomposition is proposed and divided into three steps. Three factor matrices and a core tensor are first estimated from the third-order tensor that is constructed from the HSI data cube by using the Tucker decomposition, and their major and minor principal components (PCs) are more likely to correspond to the spectral signatures of the backgrounds and the anomalies, respectively. In the second step, a reconstruction-error-based method is presented to find the first largest PCs along each mode to eliminate the spectral signatures of the backgrounds as much as possible, and thus, the remaining data may be modeled as the spectral signatures of the anomalies with a Gaussian noise. Finally, a CFAR test is implemented to detect the anomalies from the remaining data. Experiments with simulated, synthetic, and real HSI data sets reveal that the proposed method outperforms those spectral-anomaly-based met- ods with better detection probability and less false alarm rate.

Description: For the purpose of large-area topographic mapping, this study proposes an imaging system based on a multicamera array unmanned aerial system (UAS) comprised of five small-format digital cameras with a total field of view of 127 $^{circ}$ . The five digital cameras are aligned in a row along the across-track direction with overlap between two neighboring cameras. The suggested system has higher data acquisition efficiency than the single-camera UAS imaging system. For topographic mapping purposes, we develop a modified projective transformation method to stitch all five raw images into one sensor geometry. In this method, the transformation coefficients are obtained by on-the-job multicamera self-calibration, including interior and relative orientations. During the stitching process, two systematic errors are detected and corrected. In the end, a large-format digital image can be produced for each trigger event independently. The photogrammetric collinearity condition is evaluated using several external accuracy assessments, such as conventional aerial triangulation, stereoplotting, and digital surface model generation procedures. From the accuracy assessment results, we conclude that the presented raw image stitching method can be used to construct a one sensor geometry from a multicamera array and is feasible for 3-D mapping applications.

Description: Morphological operators (MOs) and their enhancements such as morphological profiles (MPs) are subject to a lively scientific contemplation since they are found to be beneficial for, for example, classification of very high spatial resolution panchromatic, multi-, and hyperspectral imagery. They account for spatial structures with differing magnitudes and, thus, provide a comprehensive multilevel description of an image. In this paper, we introduce the concept of object-based MPs (OMPs) to also encode shape-related, topological, and hierarchical properties of image objects in an exhaustive way. Thereby, we seek to benefit from the so-called object-based image analysis framework by partitioning the original image into objects with a segmentation algorithm on multiple scales. The obtained spatial entities (i.e., objects) are used to aggregate multiple sequences obtained with MOs according to statistical measures of central tendency. This strategy is followed to simultaneously preserve and characterize shape properties of objects and enable both the topological and hierarchical decompositions of an image with respect to the progressive application of MOs. Subsequently, supervised classification models are learned by considering this additionally encoded information. Experimental results are obtained with a random forest classifier with heuristically tuned hyperparameters and a wrapper-based feature selection scheme. We evaluated the results for two test sites of panchromatic WorldView-II imagery, which was acquired over an urban environment. In this setting, the proposed OMPs allow for significant improvements with respect to classification accuracy compared to standard MPs (i.e., obtained by paired sequences of erosion, dilation, opening, closing, opening by top-hat, and closing by top-hat operations).

Description: Airborne co-polarization and cross-polarization observations of ocean surface normalized radar cross section (NRCS) were conducted over the North Atlantic during January and February 2015. Observations were made using the University of Massachusetts' Imaging Wind and Rain Airborne Profiler (IWRAP) radar system and a prototype antenna for the next-generation European scatterometer aboard MetOp-SG. Both were installed on a National Oceanic and Atmospheric Administration (NOAA) WP-3D research aircraft to characterize the wind response of the ocean-surface cross-polarization NRCS. During the flights, numerous constant-roll-angle circle maneuvers were performed at several different angles to collect NRCS measurements over a range of incidence angles. Surface winds at speeds between 8 and 34 ms −1 were observed at incidence angles from 20° to 60° at all polarization combinations. The majority of measurements fell between 8 and 20 ms −1 . Wind-direction dependence similar to copolarized NRCS was observed in the cross-polarized (VH) NRCS. The amplitude of the VH NRCS with respect to direction is less than that of copolarized NRCS at all wind speeds. Incidence angle dependence was also observed in the VH NRCS at all wind speeds. As a function of wind speed, the mean VH NRCS $(A_{0})$ has a similar shape to the VV NRCS. The VH NRCS appears to not saturate at most incidence angles, unlike the VV and HH NRCS. VH and HH geophysical model functions (GMFs) were developed as functions of wind speed, incidence angle, and wind-relative azimuth for the wind speeds and incidence angles observed.

Description: Designing a kernel function with good discriminating ability and a highly application-adaptive kernelized classifier is the key of many kernel methods. However, not many kernel functions combining directly the bitemporal images' information are designed specifically for change detection tasks. In addition, extreme learning machine (ELM) has not found wide applications in change detection tasks, even though it is a potential kernel method possessing outstanding approximation and generalization capabilities as well as great classification accuracy and efficiency. Therefore, an approach relying on a difference correlation kernel (DCK) and a multistage ELM (MS-ELM) is proposed in this paper for synthetic aperture radar (SAR) image change detection. First, a DCK function is constructed specifically for change detection by measuring the “distance” between any two pixels. The DCK function depicts the cross-time similarities between couples of bitemporal image patches at any cyclic shifts with a kernel correlation operation and the high-order spatial distances between two differently located pixels with an algebraic subtraction. The DCK function possesses strong noise immunity and good identification of changed areas simultaneously. Second, an MS-ELM classifier is constructed for obtaining the change detection result. In MS-ELM, the hidden nodes and weights between the hidden and output layers are updated stage by stage by improving the kernel functions that compose them. Each stage of the MS-ELM is a standard kernel-ELM, and the DCK function is utilized in the first stage. The regenerative kernel functions incorporate the output spatial-neighborhood information of the previous stage for enhancing remarkably the MS-ELM's discriminating ability and noise resistance. The converged result at the last stage of MS-ELM is the final change detection result. Experiments on real SAR image change detection demonstrate the effectiveness of the DCK function- and the MS-ELM algorithm, particularly its good identification of changed areas and strong robustness against noise in SAR images.

Description: Global Navigation Satellite Systems Reflectometry (GNSS-R) methods sense ocean roughness by cross correlating scattered GNSS signals with a locally generated replica of the transmitted signal. The resulting delay–Doppler map (DDM) is related to surface slope statistics through established scattering models. DDM samples are correlated in time and between delay and Doppler coordinates, limiting the number of independent samples available to reduce measurement error. Performance predictions for future GNSS-R missions depend on a model with sufficient fidelity to represent these statistics. A previously developed model for the correlation in time and a new model for the correlation between delays are used to create a GNSS-R signal simulator. A change of variables reduces these models to the numerically efficient form of a single integral and a convolution. Independent normally distributed white noise is passed through a filter bank implementing these models to generate an ensemble of synthetic noisy measurements having realistic correlation in time and between delay bins. Correlation between Doppler bins, however, is not represented by this model. The output of this simulator is compared to 1-D (delay-only) DDMs collected during a 2009 airborne experiment in the North Atlantic, with winds from 5 to 25 m/s. Good agreement is found in the variance, time correlation, and covariance matrix. The probability density functions show reasonable agreement. A bias between the synthetic and observed data was found to result from a bias in the wind/roughness retrieval. Agreement was worse for the low-wind (5.8 m/s) example, perhaps due to a component of specular reflection. One application of this simulator is in generating synthetic DDMs, maintaining accurate representation of statistics following nonlinear processing (e.g., incoherent averaging). The simulator presents a numerically efficient method for generating large statistically significant ensembles of DDMs under ide- tical conditions.

Description: With the emergence of huge volumes of high-resolution remote sensing images produced by all sorts of satellites and airborne sensors, processing and analysis of these images require effective retrieval techniques. To alleviate the dramatic variation of the retrieval accuracy among queries caused by the single image feature algorithms, we developed a novel graph-based learning method for effectively retrieving remote sensing images. The method utilizes a three-layer framework that integrates the strengths of query expansion and fusion of holistic and local features. In the first layer, two retrieval image sets are obtained by, respectively, using the retrieval methods based on holistic and local features, and the top-ranked and common images from both of the top candidate lists subsequently form graph anchors. In the second layer, the graph anchors as an expansion query retrieve six image sets from the image database using each individual feature. In the third layer, the images in the six image sets are evaluated for generating positive and negative data, and SimpleMKL is applied to learn suitable query-dependent fusion weights for achieving the final image retrieval result. Extensive experiments were performed on the UC Merced Land Use–Land Cover data set. The source code has been available at our website. Compared with other related methods, the retrieval precision is significantly enhanced without sacrificing the scalability of our approach.

Description: The National Aeronautics and Space Administration's (NASA) Soil Moisture Active and Passive (SMAP) mission, which was launched on January 31, 2015, is providing global measurements of soil moisture and freeze/thaw state. The SMAP radiometer operates within the protected Earth Exploration Satellite Service passive frequency allocation of 1400–1427 MHz. However, unauthorized in-band transmitters and out-of-band emissions from transmitters operating at frequencies adjacent to this allocated spectrum are known to cause interference to microwave radiometry in this band. Because measurement corruption by these terrestrial transmissions, which is referred to as radio-frequency interference (RFI), threatens mission success, the SMAP radiometer includes special flight hardware to enable the detection and filtering of RFI. Results from the first year of SMAP data show the presence of RFI with frequent occurrence over Asia and Europe. During the calibration/validation stage of the mission, the RFI detection and mitigation algorithms were modified to provide enhanced performance. Analysis of the L1B_TB products indicates good algorithmic performance with respect to RFI detection and removal. However, some regions of the globe (e.g., Japan) continue to experience complete data loss. This paper summarizes updates to the SMAP RFI processing algorithms based on prelaunch tests and on-orbit measurements, as well as RFI information obtained in SMAP's first year on orbit.

Description: Vertical profiles of vertical air motion and raindrop size distributions (DSDs) within stratiform rain are estimated using two collocated vertically pointing radars (VPRs) operating at 3 and 35 GHz. Different raindrop backscattering cross sections occur at 3 and 35 GHz with Rayleigh scattering occurring for all raindrops at 3 GHz and Mie scattering occurring for larger raindrops at 35 GHz. This frequency-dependent backscattering cross section causes differently shaped reflectivity-weighted Doppler velocity spectra leading to radar transmit frequency-dependent radar moments of intrinsic reflectivity factor, mean Doppler velocity, and spectrum variance. The retrieval method described herein uses four radar moments as inputs to retrieve four outputs at each height within a precipitation column. The inputs include 3-GHz VPR mean Doppler velocity and unattenuated reflectivity factor and 35-GHz VPR mean Doppler velocity and spectrum variance. The outputs include vertical air motion and three parameters of a gamma-shaped DSD. To account for different VPR sample volumes, radar observations were accumulated over 45 s and over several range gates to represent time–space scales larger than either VPR sample volumes. Observed variability over this common time–space scale is used to estimate retrieval uncertainties. The retrieved air motions and DSD parameters compare well against retrievals from a collocated 449-MHz VPR that estimated air motions from Bragg scattering signals and DSD parameters from Rayleigh scattering signals.

Description: In this paper, a framework for multiscale and multifeature normalized cut (MMNCut) segmentation is proposed for high spatial resolution (HSR) remote sensing images. Normalized cuts (NCuts), as a widely used segmentation method for natural images, can obtain a globally optimized segmentation result corresponding to the optimized partitions of a graph. However, it is difficult to apply the traditional NCuts directly to HSR images because of the huge computational complexity and the diversity of the characteristics of the land covers. In order to solve these problems, the proposed MMNCuts builds a multiscale graph based on superpixels, which can provide powerful grouping cues to guide the segmentation. Generated by different algorithms with varying parameters, superpixels can capture diverse and multiscale visual patterns of HSR images. In addition, the newly constructed graph integrates the multiscale information by considering various connection relationships. Meanwhile, the successful integration of the multifeature cues, including the spectral information, texture information, and structure information, from a large number of superpixels, helps to enhance the expression ability of the graph. Computationally, this leads to a much more efficient algorithm than the traditional NCuts, and in effect, the proposed method achieves a significantly better performance than the traditional approaches. The experimental results with three HSR image data sets demonstrate that the proposed MMNCut algorithm shows a competitive performance in both qualitative and quantitative evaluations when compared with the other state-of-the-art segmentation algorithms for HSR images.

Description: Spectral unmixing is an important technique for remotely sensed hyperspectral data exploitation. It amounts to identifying a set of pure spectral signatures, which are called endmembers , and their corresponding fractional , draftrulesabundances in each pixel of the hyperspectral image. Over the last years, different algorithms have been developed for each of the three main steps of the spectral unmixing chain: 1) estimation of the number of endmembers in a scene; 2) identification of the spectral signatures of the endmembers; and 3) estimation of the fractional abundance of each endmember in each pixel of the scene. However, few algorithms can perform all the stages involved in the hyperspectral unmixing process. Such algorithms are highly desirable to avoid the propagation of errors within the chain. In this paper, we develop a new algorithm, which is termed robust collaborative nonnegative matrix factorization (R-CoNMF), that can perform the three steps of the hyperspectral unmixing chain. In comparison with other conventional methods, R-CoNMF starts with an overestimated number of endmembers and removes the redundant endmembers by means of collaborative regularization. Our experimental results indicate that the proposed method provides better or competitive performance when compared with other widely used methods.

Description: Due to the nonlinear property of large uniaxial anisotropic scatterers, many iterative optimization methods have a high risk of being trapped in local minima. In this paper, a frequency-hopping subspace-based optimization method (SOM) is proposed to reconstruct the relative permittivity distribution of 2-D large uniaxial anisotropic scatterers with transverse electrical (TE) illumination. This hybrid method utilizes the results obtained at lower frequency to provide good initial guesses for higher frequency reconstruction, which reduces the occurrence of local minima for the inversion at the higher frequency. For lower frequency, it can only obtain coarse resolution image although it is unable to show the details of the scatterers. However, this coarse image provides a priori information for the reconstruction at higher frequencies to get finer resolution. Numerical examples demonstrate that the proposed hybrid method can effectively rebuild large uniaxial anisotropic scatterers (six wavelengths) with higher stability compared with conventional SOM that uses only single-frequency data.

Description: Simplex volume is the most commonly used parameter for endmember extraction. However, when outliers exist in the image, the maximum-volume-criterion (MVC)-based methods tend to extract them as endmembers. Those outlier endmembers could be either physically meaningless or not representative enough for prevalent land covers. This is the biggest bottleneck preventing MVC-based methods from being extended from theoretical analysis to practical applications. This is mainly due to the limitation of the simplex volume formula itself, which is only determined by simplex vertices and completely ignoring the statistics of the data cloud. Usually, the simplex with vertices containing outliers has a larger volume than the one with vertices only containing true endmembers; thus, outliers are more favorably extracted as endmembers. Usually, the outliers are distributed in the direction of low information content. When extracted endmembers contain outliers, the overall information content (OIC) of the data cloud projected onto the endmember subspace will be definitely reduced. Motivated by this fact, we present the concept of statistical volume and develop a new endmember extraction method, which is named statistical volume analysis (SVA). The algorithm simultaneously utilizes the geometrical property of the simplex and the statistical characteristic of the projected data in the endmember subspace. Therefore, SVA not only can find a simplex with a large volume but also can get a large OIC of the projected data. Experiments with both simulated and real data show that SVA can compete with state-of-the-art methods in extracting endmembers of prevalent land covers. Moreover, it is capable of avoiding extracting outliers as endmembers.

Description: The Ingara X-band fully polarimetric medium-grazing-angle sea-clutter data set was collected in the Australian maritime environment over an angular range of 360° in azimuth and from 15° to 45° in grazing. This paper reports further analysis of this data set, focusing on understanding the azimuth variation to enable improved simulation accuracy and extraction of relevant geophysical parameters. This includes some original properties of the co- and cross-polarized normalized radar cross section as a function of the scattering geometry and sea surface parameters. We also assess the performances and limitations of recent sea surface scattering models in the light of this rich data set.

Description: Persistent scatterer interferometry (PSI) is in operational use for spaceborne synthetic aperture radar (SAR)-based deformation analysis. A limitation inherently associated with PSI is that, by definition, a persistent scatterer (PS) is a single dominant scatterer. Therefore, pixels containing signal contributions from multiple scatterers, as in the case of a layover, are typically rejected in the PSI processing, which in turn limits deformation retrieval. SAR tomography has the ability to resolve layovers. This paper investigates the added value that can be achieved by operationally combining SAR tomography with a PSI approach toward the objective of improving deformation sampling in layover-affected urban areas. Different tomographic phase models are implemented and compared as regards their suitability in resolving layovers. Single-look beamforming-based tomographic inversion and a generalized likelihood ratio test (GLRT)-based detection strategy are used to detect single and double scatterers. The quantity of the detected scatterers is weighed against their quality as defined in terms of the phase deviation between the single-look complex (SLC) measurements and the tomographic model fit. The gain in deformation sampling that can be derived with tomography relative to a PSI-based analysis is quantitatively assessed, and alongside the quality of the scatterers obtained with tomography is compared with the quality of the PSs identified with a PSI approach. The experiments are performed on an interferometric stack of 50 TerraSAR-X stripmap images. The results obtained show that, although there is a tradeoff between the quantity and the quality of the detected scatterers, the tested SAR tomography approach leads to an improvement in deformation sampling in layover-affected areas.

Description: Hyperspectral imaging offers new opportunities for pattern recognition tasks in the remote sensing community through its improved discrimination in the spectral domain. However, such advanced image processing also brings new challenges due to the high data dimensionality in both the spatial and spectral domains. To relieve this issue, in this paper, we present a novel multidomain subspace (MDS) feature representation and classification method for hyperspectral images. The proposed method is based on a patch alignment framework. In order to optimally combine the feature representations from the various domains and simultaneously enhance the subspace discriminability, we incorporate the supervised label information into each domain and further generalize the framework to a multidomain version. Furthermore, we develop an iterative approach to alternately optimize the MDS objective function by considering it as two subconvex optimizations. The classification performance on three standard hyperspectral remote sensing images confirms the superiority of the proposed MDS algorithm over the state-of-the-art subspace learning methods.

Description: This paper deals with synthetic aperture radar (SAR) raw data simulation for ocean scenes featuring surface waves and currents, an issue which has proven to be of great necessity in preparing for future oceanic SAR missions. In this paper, the inverse Omega-K (IOK) algorithm, which is originally designed for SAR raw data simulation of stationary land scenes, is extended to ocean scenes. To realize such a generalization, endeavors are made in two aspects. First, specially aimed at ocean dynamics of ocean waves and currents, the 2-D spectrum of the SAR signal is derived. Second, to account for the spatial variation of ocean-motion parameters, we adopt a strategy called batch processing, whose basic feature is that a single implementation of the IOK algorithm will simultaneously simulate a collection of ocean-surface backscattering elements that have the same radial velocity. For the proposed simulator, the velocity bunching effect is embodied via making the long-wave radial orbital velocities physically enter the range equation of the SAR raw signal, instead of superimposing this effect onto the reflectivity map. The spread of the facet velocities within one resolution cell enters the raw data through a random perturbation of the local long-wave orbital velocity. The proposed simulator is not only rather accurate due to the fact that, in deriving the range frequency mapping function, no Taylor expansion is made on the range equation, but also much more efficient than its time-domain counterpart. Effectiveness of the proposed simulator is validated by using simulation results.

Description: Boreal forests are characterized by a rather homogeneous stand structure that allows by means of a single allometric equation to estimate biomass from forest height with sufficient accuracy and, therefore, to use this equation for quantitative biomass classifications. In this paper, interferometric TanDEM-X DEM data are used to estimate forest height over boreal forest systems. The accuracy of the height inversion is evaluated for single- and dual-baseline scenarios, under summer and winter conditions. Then, an allometric equation is used to transfer forest height to biomass. For this, two forest sites, boreal (Krycklan) and hemi-boreal (Remningstorp) in north and southern Sweden, respectively, are investigated. A performance analysis is carried out to estimate the maximum number of biomass classes obtained, depending on the height estimation accuracy. For summer acquisitions, four biomass classes can be obtained, with a maximum biomass class of > 200 Mg/ha. For winter acquisitions or when a mixed summer—winter approach is applied, five biomass classes, up to 220 Mg/ha, can be obtained. This classification shows a good agreement with CORINE, an existing land cover classification, and can improve it by adding quantitative forest biomass classes with a high spatial resolution of 16 $times$ 16 m.

Description: The Alpha Jet Atmospheric eXperiment (AJAX) is a project to measure the atmospheric profiles of greenhouse gases (GHGs) and ozone (O 3 ) regularly over California and Nevada. Airborne instruments measuring GHGs and O 3 are installed in a wing pod of an Alpha Jet aircraft and operated from the National Aeronautics and Space Administration Ames Research Center at Moffett Field, CA. The instruments yield precise and accurate in situ vertical profiles of atmospheric carbon dioxide (CO 2 ), methane (CH 4 ), and O 3 . Measurements of vertical profiles of GHGs and O 3 over Railroad Valley, NV have been conducted directly under the Greenhouse gases Observing SATellite (GOSAT) over passes on a monthly basis as part of the AJAX project since June 2011. The purpose of this work is to calculate aircraft-based dry-air mole fractions of the GHGs for the validation of GOSAT data products. This study expands and improves our previous comparisons by evaluating three algorithms against 24 months of in situ data collected over a Gain-M target. We used three different algorithms: Atmospheric CO 2 Observations from Space (ACOS v3.4r3), Remote Sensing of Greenhouse Gases for Carbon Cycle Modeling (RemoteC v2.3.5FP), and National Institute for Environmental Studies (NIES v2.11). We find that the CO 2 average differences of ACOS and RemoteC from AJAX are 0.26% and 0.24%, respectively. The difference between NIES and AJAX is 0.96%, which is higher than that of ACOS and RemoteC. The CH 4 average differences for RemoteC and NIES are 2.1% and 1.7%, respectively.

Description: A retrieval algorithm is presented for the Level 2 ocean surface wind speed data product of the Cyclone Global Navigation Satellite System (CYGNSS) mission. The algorithm is based on the approach described by Clarizia et al ., 2014. The approach is applied to the specific orbital measurement geometry, antenna, and receiver hardware characteristics of the CYGNSS mission. Several additional processing steps have also been added to improve the performance. A best weighted estimator is used to optimally combine two different partially correlated estimates of the winds by taking their weighted average. The optimal weighting dynamically adjusts for variations in the signal-to-noise ratio of the observations that result from changes in the measurement geometry. Variations in the incidence angle of the measurements are accounted for by the use of a 2-D geophysical model function that depends on both wind speed and incidence angle. Variations in the propagation time and signal Doppler shift at different measurement geometries affect the instantaneous spatial resolution of the measurements, and these effects are compensated by a variable temporal integration of the data. In addition to a detailed description of the algorithm itself, the root-mean-square wind speed retrieval error is characterized as a function of the measurement geometry and the wind speed using a detailed mission end-to-end simulator.

Description: Compact polarimetric (CP) synthetic aperture radar (SAR) systems are attracting increasing interest as they possess the ability to construct quad-polarimetric information with advantages of simpler system, lower data rate, and wider swath, compared to a fully polarimetric (FP) system. In order to get undistorted polarimetric information, calibration of CP measurements has to be performed. Freeman and Chen both proposed methods for CP SAR calibration using merely calibrators, which require at least three calibrators at a range line and would require many more calibrators when monitoring the change of the system across the swath is necessary. It has been proved that FP SAR calibration methods exploring scattering characteristics of natural distributed targets (DTs) can reduce the requirement of the number of deployed calibrators. However, it is much more complicated in the case of CP SAR calibration, as only two channels of CP measurements are available. This paper proposes a numerical CP calibration method using natural DTs and only one corner reflector (CR) for a reciprocal CP SAR system. The validity of the method for three typical CP modes is confirmed by simulations with Advanced Land Observing Satellite Phased Array L-band SAR data. Through sensitivity analysis, a dihedral at 0 $^{circ}$ is selected as the best CR for the calibration algorithm.

Description: The aim of this paper was to estimate soil moisture in agricultural crop fields from fully polarimetric L-band synthetic aperture radar (SAR) data through the polarimetric decomposition of the SAR coherency matrix. A nonnegative-eigenvalue-decomposition scheme, together with an adaptive volume scattering model, is extended to an adaptive model-based decomposition (MBD) (Adaptive MBD) model for soil moisture retrieval. The Adaptive MBD can ensure nonnegative decomposed scattering components and allows two parameters (i.e., the mean orientation angle and a degree of randomness) to be determined to characterize the volume scattering. Its performance was tested using airborne SAR data and coincident ground measurements collected over agricultural fields in southeastern Australia and compared with previous MBD methods (i.e., the Freeman three-component decomposition using the extended Bragg model, the Yamaguchi three-component decomposition, and an iterative generalized hybrid decomposition). The results obtained with the newly proposed decomposition scheme agreed well with expectations based on observed plant structure and biomass levels. The new method was superior in tracking soil moisture dynamics with respect to previous decomposition methods in our study area, with root-mean-square error of soil moisture estimations being 0.10 and 0.14 m 3 /m 3 , respectively, for surface and double-bounce components. However, large variability in the achieved soil moisture accuracy was observed, depending on the presence of row structures in the underlying soil surface.

Description: Due to the complementary properties of different features, multiple feature fusion has a large potential for hyperspectral imagery classification. At the meantime, hashing is promising in representing a high-dimensional float-type feature with extremely low bit binary codes while maintaining the performance. In this paper, we study the possibility of using hashing to fuse multiple features for hyperspectral imagery classification. For this purpose, we propose a multiple feature fusion framework to evaluate the performance of using different hashing methods. For comparison and completeness, we also have an extensive comparison to five subspace-based dimension reduction methods and six fusion-based methods which are popular solutions to deal with multiple features in hyperspectral image classification. Experimental results on four benchmark hyperspectral data sets demonstrate that using hashing to fuse multiple features can achieve comparable or better performance with the traditional subspace-based dimension reduction methods and fusion-based methods. Moreover, the binary features obtained by using hashing need much less storage and are faster to compute distances with the help of machine instructions.

Description: Radiometric correction of radar images is essential to produce accurate estimates of biophysical parameters related to forest structure and biomass. We present a new algorithm to correct radiometry for 1) terrain topography and 2) variations of canopy reflectivity with viewing and tree-terrain geometry. This algorithm is applicable to radar images spanning a wide range of incidence angles over terrain with significant topography and can also take into account aircraft attitude, antenna steering angle, and target geometry. The approach includes elements of both homomorphic and heteromorphic terrain corrections to correct for topographic effects and is followed by an additional radiometric correction to compensate for variations of canopy reflectivity with viewing and tree-terrain geometry. The latter correction is based on lookup tables and enables derivation of biophysical parameters irrespective of viewing geometry and terrain topography. We evaluate the performance of the new algorithm with airborne radar data and show that it performs better than classical homomorphic methods followed by cosine-based corrections.

Description: Based on the construction of the Chinese Meridian Project, two mesosphere, stratosphere, and troposphere (MST) radars were designed and located in Beijing and Chongyang, Hubei province, China, to investigate the atmospheric dynamics in the troposphere, stratosphere, and mesosphere. The official names of the two radio systems are Beijing MST radar and Wuhan MST radar. They are the monostatic radars with an active phased antenna array consisting of 576 Yagi antennas and operated by a 53.8-MHz frequency. The three-element Yagi antennas are arranged in a square grid of 96-m side length to form the aperture of 9216 m 2 . This antenna array arrangement forms the five symmetric radar beams of vertex, east, west, south, and north. The beamwidth is 3.2°, the maximum directive gain is 34.8 dB, and the total transmitting peak power is ∼172 kW. The zenith angle of the oblique beams is adjustable between 0 and 20° with a 1° step. The average power aperture product of the radars is $3.2times 10^{8} text{Wm}^{2}$ . There are three operating modes of the MST radars, including the low, middle, and high modes, applied to observe the troposphere, stratosphere, and mesosphere, respectively. Thus, these two coherent pulse Doppler MST radars have the ability to study the features of the midlatitude atmospheric turbulence and wind field vector from the troposphere to the lower thermosphere with high spatiotemporal resolution. In this paper, the antenna array, system hardware, and signal processing methods, as well as the typical observation results of the troposphere, stratosphere, and mesosphere, are introduced.

Description: Cameron's coherent decomposition is reconsidered in this paper for incoherent target decomposition. Characterizations of the minimum and maximum symmetric scattering components are considered in a unified model and projected into the Pauli basis to derive a new scattering vector model for the representation of coherent target scattering. Some roll-invariant parameters can be derived from the defined scattering vector model, and the classification method that Cameron developed for operational use of his decomposition can be performed using these parameters. The proposed model is compared with Touzi's target scattering vector model (TSVM), and it is revealed that the parameters of the TSVM are not unique and are invariant to the orientation angle of the polarization ellipse of the maximum copolarization backscattering return, which is descriptive of the polarization, whereas the extracted parameters from the proposed scattering vector are invariant to the orientation of the symmetry axis of the maximum symmetric component of the scatterer, which is descriptive of the target. From the implementation using Uninhabited Aerial Vehicle Synthetic Aperture Radar and Experimental Synthetic Aperture Radar images, the advantages and efficiency of the proposed scattering vector model are demonstrated.

Description: The Advanced Technology Microwave Sounder (ATMS) on board the Suomi National Polar-orbiting Partnership (NPP) satellite is a total power radiometer and scans across the track with a range of ±52.77° from nadir. It has 22 channels and measures the microwave radiation at either quasi-vertical or quasi-horizontal polarization from the Earth's atmosphere. The ATMS scanning reflector is made of beryllium coated with gold and can have an emission due to the surface roughness. During prelaunch phase, an estimate of the reflector emissivity was not explored. In this paper, a new methodology is developed to assess the antenna emission from the ATMS pitch-over observations. It is found that the antenna emission is significant and dominates the scan-angle-dependent features in the ATMS antenna temperatures. Retrieved emissivity from K- to G-bands ranges from 0.002 to 0.006. An error model was also developed to assess the impact of antenna emissivity to calibration accuracy of antenna temperature products. Simulation results show that the calibration error is scene temperature dependent and can be as large as 2.5 K for space view.

Description: High-resolution ice concentration maps are of great interest for ship navigation and ice hazard forecasting. In this case study, a convolutional neural network (CNN) has been used to estimate ice concentration using synthetic aperture radar (SAR) scenes captured during the melt season. These dual-pol RADARSAT-2 satellite images are used as input, and the ice concentration is the direct output from the CNN. With no feature extraction or segmentation postprocessing, the absolute mean errors of the generated ice concentration maps are less than 10% on average when compared with manual interpretation of the ice state by ice experts. The CNN is demonstrated to produce ice concentration maps with more detail than produced operationally. Reasonable ice concentration estimations are made in melt regions and in regions of low ice concentration.

Description: Underground mines are characterized by a network of intersecting tunnels and sharp turns, an environment which is particularly challenging for radiofrequency based positioning systems due to extreme multipath, non-line-of-sight propagation, and poor anchor geometry. Such systems typically require a dense grid of devices to enable 3-D positioning. Moreover, the precise position of each anchor node needs to be precisely surveyed, a particularly challenging task in underground environments. Magneto-inductive (MI) positioning, which provides 3-D position and orientation from a single transmitter and penetrates thick layers of soil and rock without loss, is a more promising approach, but so far has only been investigated in simple point-to-point contexts. In this paper, we develop a novel MI positioning approach to cover an extended underground 3-D space with unknown geometry using a rapidly deployable anchor network. The key to our approach is that the position of only a single anchor needs to be accurately surveyed—the positions of all secondary anchors are determined using an iterative refinement process using measurements obtained from receivers within the network. This avoids the particularly challenging and time-intensive task in an underground environment of accurately surveying the positions of all of the transmitters. We also demonstrate how measurements obtained from multiple transmitters can be fused to improve localization accuracy. We validate the proposed approach in a man-made cave and show that, with a portable system that took 5 min to deploy, we were able to provide accurate through-the-earth location capability to nodes placed along a suite of tunnels.

Description: Hyperspectral remote sensing images, which are characterized by their high dimensionality, provide us with the capability to accurately identify objects on the ground. They can also be used to identify subclasses of objects. However, these subclasses are usually embedded in different subspaces due to the complex distribution of pixels in the feature space. In the literature, few hyperspectral image classification methods can take both the subclass and subspace into consideration at the same time. Motivated by the fact that natural DNA can distinguish biological subspecies (subclasses in hyperspectral images) using critical DNA fragments (subspaces in hyperspectral images), a semisupervised subspace-based DNA encoding and matching classifier for hyperspectral remote sensing imagery (SSDNA) is proposed in this paper. First, in the process of DNA encoding, the hyperspectral remote sensing image is transformed into a DNA cube, in which the first-order spectral curve of the hyperspectral remote sensing image is utilized in order to take the gradient information of the spectral curve into consideration. Second, in the process of DNA optimization, evolutionary algorithms are used to obtain the best DNA library of the typical objects, which includes the following: 1) A multicenter individual representation is designed in order to consider the existence of subclasses in the hyperspectral remote sensing image; 2) the unlabeled samples are utilized in the process of population initialization and fitness calculation to enhance the diversity of the population and the generalization of the classification performance; and 3) the different classes are embedded in different subspaces. A semisupervised technique is used to extract the subspaces, including the global subspace for all the classes and the local subspace for each class. Three hyperspectral data sets were tested and confirm that SSDNA performs better than the other supervised or semisupervised classifiers.

Description: In this paper, we propose a spectral–spatial feature based classification (SSFC) framework that jointly uses dimension reduction and deep learning techniques for spectral and spatial feature extraction, respectively. In this framework, a balanced local discriminant embedding algorithm is proposed for spectral feature extraction from high-dimensional hyperspectral data sets. In the meantime, convolutional neural network is utilized to automatically find spatial-related features at high levels. Then, the fusion feature is extracted by stacking spectral and spatial features together. Finally, the multiple-feature-based classifier is trained for image classification. Experimental results on well-known hyperspectral data sets show that the proposed SSFC method outperforms other commonly used methods for hyperspectral image classification.

Description: We present a method to retrieve wind speeds in hurricanes from spaceborne passive microwave radiometer data. Brightness temperature $(T_{B})$ observations acquired at the 6.9-GHz horizontal polarization channel by the AMSR-E and AMSR2 onboard the Earth Observing System Aqua and Global Change Observation Mission-Water 1 satellites are selected for wind retrieval due to the fact that the signal at this frequency is sensitive to high wind speeds but less sensitive to rain scatter than those acquired at other higher frequency channels. The AMSR-E and AMSR2 observations of 53 hurricanes between 2002 and 2014 are collected and collocated with stepped-frequency microwave radiometer (SFMR) measurements. Based on the small slope approximation/small perturbation method model and an ocean surface roughness spectrum, the wind speeds are retrieved from the $T_{B}$ data and validated against the SFMR measurements. The statistical comparison of the entire data set shows that the bias and root-mean-square error (RMSE) of the retrieved wind speeds are 1.11 and 4.34 m/s, respectively, which suggests that the proposed method can obtain high wind speeds under hurricane conditions. Two case studies show that the wind speed retrieval bias and RMSE are 1.08 and 3.93 m/s for Hurricane Earl and 0.09 and 3.23 m/s for Hurricane Edouard, respectively. The retrieved wind speeds from the AMSR-E and AMSR2 continuous three-day observations clearly show the process of hurricane intensification and weakening.

Description: The electromagnetic scattering by a dielectric cylinder of finite length is important in many applications, particularly for microwave remote sensing of vegetated terrain. Yet, not only a unified analytical solution is still elusive in providing the scattering cross sections but also other important aspects, such as the phase of the scattering amplitude function, energy conservation, and reciprocity relation, have been scantly touched in the literature. The treatment of a dielectric inhomogeneous cylinder of finite length brings forth new challenges. Taking on such challenges is the focus of this paper. The main plan of attack is to extend the virtual partition method, which is a T-matrix-based semi-analytical model, that we have previously proposed to treat scattering from a homogeneous dielectric cylinder of finite length, to the inhomogeneous cases. The effectiveness of the proposed method is validated numerically, including: 1) high-fidelity prediction of the copolarized and cross-polarized cross sections for arbitrary bistatic scattering configuration; 2) high-fidelity predictions of the phase of the scattering amplitude function; 3) verification of energy conservation; and 4) verification of the reciprocity theorem.

Description: The Advanced SCATterometer (ASCAT) measurement spatial response function (SRF) relates the weighted contribution of every location within the measurement footprint to the measured normalized radar cross section $sigma^{circ}$ . The SRF results from a combination of the antenna response and the onboard processing and is computed during ground processing by modeling in detail the measurement geometry, as this is required for an accurate $sigma^{circ}$ estimation. However, the computed SRF is not disseminated as part of the L1B data. For some applications of the L1B data, the SRF is additionally required. For these applications, an approximate description of the SRF is often sufficiently accurate, estimated from information contained in the L1B data, rather than from a full calculation based on the measurement geometry. This paper describes a parameterized model of the ASCAT SRF for each measurement. First, an SRF reference estimate that incorporates details on the ASCAT design and onboard measurement processing is created. A parameterized model is fit to the reference estimate. The parameterized SRF is computationally less demanding than the reference estimate and as such more useful for near-real-time processing. The two estimates are validated with the computed SRF used in ground processing and with the transponder data from calibration campaigns. Finally, to validate the SRF in a simple application, the land fraction (a measure of land contamination in near-coastal ocean measurements) is computed and compared to actual data for a sample region.

Description: Ferrous and highly conductive materials distort low-frequency magnetic fields and can significantly increase magnetoinductive positioning errors. In this paper, we use the image theory in order to formulate an analytical channel model for the magnetic field of a quasi-static magnetic dipole positioned above a perfectly conducting half-space. The proposed model can be used to compensate for the distorting effects that metallic reinforcement bars (rebars) within the floor impose on the magnetic field of a magnetoinductive transmitter node in an indoor single-story environment. Good agreement is observed between the analytical solution and numerical solutions obtained from 3-D finite-element simulations. Experimental results indicate that the image theory model shows improvement over the free-space dipole model in estimating positions in the distorted environment, typically reducing positioning errors by 22% in 90% of the cases and 26% in 40% of the cases. No prior information on the geometry of the metallic distorters was available, making this essentially a “blind” technique.

Description: Landscape heterogeneity is a common natural phenomenon but is seldom considered in current radiative transfer (RT) models for predicting the surface reflectance. This paper developed an analytical RT model for heterogeneous Agro-Forestry scenarios (RTAF) by dividing the scenario into nonboundary regions (NRs) and boundary regions (BRs). The scattering contribution of the NRs can be estimated from the scattering-by-arbitrarily-inclined-leaves-with-the-hot-spot-effect model as homogeneous canopies, whereas that of the BRs is calculated based on the bidirectional gap probability by considering the interactions and mutual shadowing effects among different patches. The multiangular airborne observations and discrete-anisotropic-RT model simulations were used to validate and evaluate the RTAF model over an agro-forestry scenario in the Heihe River Basin, China. The results suggest that the RTAF model can accurately simulate the hemispherical–directional reflectance factors (HDRFs) of the heterogeneous scenarios in the red and near-infrared (NIR) bands. The boundary effect can significantly influence the angular distribution of the HDRFs and consequently enlarge the HDRF variations between the backward and forward directions. Compared with the widely used dominant cover type (DCT) and spectral linear mixture (SLM) models, the RTAF model reduced the maximum relative error from 25.7% (SLM) and 23.0% (DCT) to 9.8% in the red band and from 19.6% (DCT) and 13.7% (SLM) to 8.7% in the NIR band. The RTAF model provides a promising way to improve the retrieval of biophysical parameters (e.g., leaf area index) from remote sensing data over heterogeneous agro-forestry scenarios.

Description: Hyperspectral images (HSIs) are inevitably corrupted by mixture noise during their acquisition process, in which various kinds of noise, e.g., Gaussian noise, impulse noise, dead lines, and stripes, may exist concurrently. In this paper, mixture noise removal is well illustrated by the task of recovering the low-rank and sparse components of a given matrix, which is constructed by stacking vectorized HSI patches from all the bands at the same position. Instead of applying a traditional nuclear norm, a nonconvex low-rank regularizer, i.e., weighted Schatten p -norm (WSN), is introduced to not only give better approximation to the original low-rank assumption but also to consider the importance of different rank components. The resulted nonconvex low-rank matrix approximation (LRMA) model falls into the applicable scope of an augmented Lagrangian method, and its WSN minimization subproblem can be efficiently solved by generalized iterated shrinkage algorithm. Moreover, the proposed model is integrated into an iterative regularization schema to produce final results, leading to a completed HSI restoration framework. Extensive experimental testing on simulated and real data shows, both qualitatively and quantitatively, that the proposed method has achieved highly competent objective performance compared with several state-of-the-art HSI restoration methods.

Description: On phased-array radars, scanning is done by stepping the beam from one direction to the next direction and dwelling long enough at each direction to achieve acceptable errors of estimates. Combining data from the three directions is suggested to obtain superresolution similar to that available on the national network of weather radar (Weather Surveillance Radar-1988 Doppler or WSR-88D). Spectral analysis of such data is addressed, and it is demonstrated that the Doppler spectra of simply concatenated time series have very strong sidebands due to the discontinuity of the signals from the three beam positions. This artifact degrades the performance of the spectral clutter filters and other methods that rely on spectral processing to enhance the weather signal. Special adjustments of the signals at each range location before concatenating (splicing) are proposed to mitigate the effects of discontinuities in time and thus improve clutter filtering. The adjustment is such that the total information contained in the signal can be preserved in subsequent processing. Spectral quality of the concatenated signals is quantified via results from simulations. Samples of spectra obtained with the National Weather Radar Testbed are presented to substantiate the predictions. A ground clutter detector/filter accepted by the National Weather Service is applied to the conditioned time series data, and the ensuing fields of reflectivity factor and Doppler velocity are compared to the fields from which clutter had not been removed.

Description: The mystery of the asteroid (4179) Toutatis was revealed by Chang'e-2 spacecraft during a close flyby on December 13, 2012. Optical imaging and navigation of the probe during the flyby were performed entirely under ground-based radio tracking and default sequence built on ground. This paper establishes a set of estimation algorithms of the relative trajectory between Chang'e-2 and Toutatis based on dynamics, optical, and radio constraints that are determined by the unique flyby mode. This study is the first time to precisely reproduce the core process of Chang'e-2's encounter with Toutatis based on several optical images. In addition to constructing a strict photogrammetric model, the shadowing effects caused by the illumination and the deviation of the center-of-mass (COM) from the center-of-figure (COF) in optical images are also considered. The spacecraft trajectory with regard to the COF of the body is estimated using images taken from 120 km or less. The formal one sigma uncertainty is (67, 20, and 11 m) in the principal axes frame of the position error ellipse, and the closest approaching distance between Chang'e-2 and Toutatis's COF is calculated as 1557 ± 11 m, which is more precise than previous results with an uncertainty of hundreds of meters. The spacecraft trajectory with regard to the COM of the body is estimated with an uncertainty of (211, 34, and 17 m), and the corresponding closest distance is estimated as 1451 ± 18 m based on the previously developed shape model of Toutatis. The algorithms and results in this study are important for evaluating the performance of this flyby mission and are also valuable for any similar optical navigation during a close approach. In addition, our results can help in precisely determining the axis of Toutatis and sizes of impact craters, which are critical for understanding the formation and evolution of Toutatis.

Description: A new segmentation approach for high-resolution remotely sensed imagery that combines the global edge and region information is developed from a new scheme to monitor the best conditions for each growing object to obtain the corresponding meaningful image object during multiscale analysis. The approach, which is an extension of the image object detection approach, includes new algorithms for determination of region-growing criteria, edge-guided image object detection, and assessment of edges. The method consists of two stages: In the first stage, edges are acquired from edge detection with embedded confidence and stored in an R-tree, and initial objects are segmented by eCognition and organized in the region adjacency graph; in the second stage, meaningful image objects are obtained by incorporating multiscale segmentation and analyzing the edge completeness curve. The evaluation results of edge completeness are obtained within the process of multiscale segmentation, and the assessment for the segmentation results shows its merit in coastal remote sensing. Images containing plenty of weak edges or distributing scene objects with various sizes and shapes can fully embody the strength of this method.

Description: Evapotranspiration (ET) is one of the key variables in water cycle and ecological systems, whereas it is difficult to quantify ET variations from traditional observations in large river basins, e.g., Mississippi River basin (MRB). In this paper, a new geodetic tool, i.e., Global Positioning System (GPS), is used for the first time to estimate monthly ET variations at a regional scale. Based on the water balance equation, the monthly ET variation is estimated using the GPS-derived terrestrial water storage (TWS) from January 2006 to July 2015 in MRB. The annual amplitude of GPS-inferred TWS in MRB agrees well with the results of Gravity Recovery and Climate Experiment. The ET variations from the water balance approach agree well with the land surface modeling and remote sensing data. The correlation of GPS-inferred ET with other ET products is higher than 0.8, which indicates that the GPS-estimated ET well characterizes the ET variations in MRB. The annual amplitude of GPS-inferred ET variations is 47.9 mm/month, which is close to that from land surface modeling of North American Land Data Assimilation System, and a little larger than MODerate Resolution Imaging Spectroradiometer. The mean monthly ET reaches its maximum in June–July and its minimum in December, which is consistent with the periodic pattern of radiative energy in a year. Furthermore, the ET variations are mainly dominated by the temperature change in MRB.

Description: Modeling 3-D trees from terrestrial laser scanning (TLS) point clouds remains a challenging task for several well-known reasons, including their complex structure and severe occlusions. In order to accurately reconstruct 3-D tree models from TLS point clouds that typically suffer from significant occlusions, in this paper, a novel local structure and direction-aware approach is presented to successfully complete missing structures of trees. In this method, we first extract the coarse tree skeleton from the input point cloud, and thus, the branch dominant direction and the point density of each branch are obtained. By a skeleton-based Laplacian algorithm, the point cloud is further shrunk into a skeleton point cloud to highlight the branch dominant direction of each branch. For obtaining even more accurate point densities, a dictionary-based algorithm is utilized to learn and reconstruct the local structure. Finally, the branch dominant direction and point density are integrated into an iterative optimization process to recover the missing data. Extensive experimental results have shown that the proposed method is very robust to incomplete data sets, and it is capable of accurately reconstructing 3-D trees, which are partially, or even to a large extent, missing from the input point cloud.

Description: The midinfrared (MIR) spectral region (3–5 $mutext{m}$ ), which penetrates most haze layers in the atmosphere and is less sensitive to variations in atmospheric water vapor, seems to be appropriate for retrieving land surface temperature (LST). However, there are currently few studies of LST retrieval with MIR data because it is difficult to eliminate solar irradiance from the total energy measured in the MIR during the daytime. This paper proposes a physics-based method to retrieve LST from MODIS daytime MIR data. The bidirectional reflectivity describing the reflected solar direct irradiance is determined using the method by Tang and Li. The directional emissivity, representing the surface emitted radiance, is determined by a kernel-driven bidirectional reflectance distribution function model, i.e., RossThick-LiSparse-R. Intercomparisons using the MODIS-derived LST product MYD11_L2, for the Baotou experimental site in Urad Qianqi, Inner Mongolia, China, have a maximum root-mean-square error (RMSE) of 1.69 K and a minimum RMSE of 1.31 K, for four scenes of MODIS images. Furthermore, in situ LSTs measured at the Hailar field site in northeastern Inner Mongolia, China, were also used to validate the proposed method. Comparisons of the LSTs retrieved from MODIS daytime MIR data and those calculated using in situ measurements have a bias and RMSE of −0.17 K and 1.42 K, respectively, which indicates that the proposed method can accurately retrieve LST from MODIS daytime MIR data.

Description: This paper describes a new search strategy within the scope of factorized geometrical autofocus (FGA) and synthetic-aperture-radar processing. The FGA algorithm is a fast factorized back-projection formulation with six adjustable geometry parameters. By tuning the flight track step by step and maximizing focus quality by means of an object function, a sharp image is formed. We propose an efficient two-stage approach for the geometrical variation. The first stage is a low-order (few parameters) parallel search procedure involving small image areas. The second stage then combines the local hypotheses into one global autofocus solution, without the use of images. This method has been applied successfully on ultrawideband CARABAS II data. Errors due to a constant acceleration are superposed on the measured track prior to processing, giving a 6-D autofocus problem. Image results, including resolution, peak-to-sidelobe ratio and magnitude values for point-like targets, finally confirm the validity of the strategy. The results also verify the prediction that there are several satisfying autofocus solutions for the same radar data.

Description: This paper presents a novel hybrid method to the estimation of bio/geophysical parameters, which models and corrects deviations from correct target values when theoretical electromagnetic models are used for the inversion process. The proposed hybrid method integrates theoretical models with empirical observations associated to a few field reference samples. This is achieved based on two steps. In the first step, deviations between estimations obtained by a theoretical model and empirical observations are initially computed. Then, deviations associated to unlabeled samples (for which reference measures are not existing) are characterized based on two different strategies: 1) the global deviation bias strategy (which assumes that the deviations of samples are constant within the input space); and 2) the local deviation bias strategy (which assumes that the deviations of samples are variable within different portions of the input space). In the second step, the theoretical model estimates of unlabeled samples are corrected based on the estimated deviations. The experimental analysis carried out in the context of soil moisture content retrieval from microwave remotely sensed data confirms the effectiveness of the proposed hybrid estimation method.

Description: The uncertainty of differential code bias (DCB) is one of the main error sources in the low Earth orbit (LEO) based total electron content (TEC) retrieval, whereas the derivation of the LEO DCB is not systematically studied. In this paper, we propose an improved DCB estimation method (ZERO method) based on the assumption that the LEO-based TEC can reach zero and also optimize the parameter configuration in the commonly used least square method (LSQ method). In the improved ZERO method, the combination of the lower quartile minimum relative TEC during each orbital revolution with the daily minimum relative TEC gives a stable and reliable DCB estimation. For the LSQ method, the 3-TECU cutoff vertical TEC with 10° cutoff elevation is considered to offer a reasonable DCB estimation. Subsequently, Global Positioning System (GPS) observations from multiple LEO satellites at different altitudes are used to study the variability of the LEO DCBs. Our results revealed that the LEO DCBs underwent obvious long-term variation and periodic oscillations of months. Moreover, the CHAMP data illustrated that the long-term variation of LEO DCBs is partly associated with the GPS satellite replacement, and the periodic variation can be attributed to the variation of the hardware thermal status, represented by the receiver CPU temperature in this study.

Description: Since its launch in 2009, the European Space Agency mission Soil Moisture and Ocean Salinity (SMOS) has provided valuable information on soil moisture, sea surface salinity data, and other geophysical variables. Due to its innovative instrument, an L-band 2-D synthetic aperture interferometric radiometer, SMOS is able to provide high-resolution L2 data (as compared with other L-band missions). However, SMOS processing is complex, giving rise to the emergence of some unexpected biases. In this paper, we have analyzed the spatial structure of two-point correlations owing to the SMOS synthetic antenna, finding that they are not negligible. Those correlations can be characterized by means of effective point spread functions (PSFs). This paper indicates that the SMOS PSF matrix can be computed in a fast way from measured data without the need for any model or auxiliary data. Furthermore, this matrix can be described in terms of a convolution kernel. The knowledge of that convolution kernel can be used to improve the quality of the SMOS image and to assess the effect of changes of processing procedures, including calibration methods.

Description: In this paper, we propose a self-learning approach for remote sensing image classification. The main work of this paper aims at providing a new framework of semisupervised learning technique for multiple-source synergetic classification, thereby improving the classification accuracy under the condition of small samples. Considering the high spectral resolution of a hyperspectral (HS) image and the high spatial resolution of a panchromatic (PAN) image, the proposed approach combines image segmentation with an active learning algorithm and adopts a standard active learning method for a self-learning strategy, in which the learning algorithm automatically selects informative unlabeled samples by itself according to their collaborative spatial–spectral features and the predicted information of a spectral-based classifier. This way, no extra cost of human expertise is required for labeling the selected pixels when compared with conventional active learning methods. Experiments on three data sets, including HS and PAN images, indicate that our proposed approach has a great enhancement on overall classification accuracy compared with classical supervised algorithms and turns out to be a promising strategy in synergetic classification of HS and PAN images.

Description: 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 m 3 $cdot$ m −3 , well within the mission requirement of 0.04 m 3 $cdot$ 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.

Description: The dielectric properties of sea ice are important for both passive and active microwave remote sensing of sea ice. In this paper, we present a new technique for dielectric measurements of artificially grown sea ice in the frequency range between 0.3 and 12 GHz using an open-ended coaxial probe. To provide a solid contact between the probe and ice, we slightly submerge and then freeze the probe's flange in sea water in a cold laboratory with a preset temperature. Once the ice is formed, we conduct a measurement of the complex reflection coefficient in the cold room using a vector network analyzer. To calibrate the system, we propose a set of measurements from air, shorting block (short), and pure methanol to be conducted immediately after. Both the real and imaginary parts of the complex dielectric constant as functions of frequency are then derived using a coaxial probe inverse model fed by these data. X-ray microtomography analysis of our samples revealed that the ice formed under the described conditions has completely isotropic microstructure typical for the frazil layer of natural first-year sea ice. To evaluate the experimental system's accuracy, we conducted extensive test measurements of standard materials (saline water, methanol, butanol, and pure ice). We also demonstrate that our sea ice dielectric measurements are close to corresponding values previously reported in the literature. The proposed measurement technique is valuable for developing a sea ice dielectric mixture model at microwave frequencies for different temperatures and salinities.

Description: In hyperspectral unmixing applications, one typically assumes that a single spectrum exists for every endmember. In many scenarios, this is not the case, and one requires a set or a distribution of spectra to represent an endmember or class. This inherent spectral variability can pose severe difficulties in classical unmixing approaches. In this paper, we present a new algorithm for dealing with endmember variability in spectral unmixing, based on the geometrical interpretation of the resulting unmixing problem, and an alternating optimization approach. This alternating-angle-minimization algorithm uses sets of spectra to represent the variability present in each class and attempts to identify the subset of endmembers which produce the smallest reconstruction error. The algorithm is analogous to the popular multiple endmember spectral mixture analysis technique but has a much more favorable computational complexity while producing similar results. We illustrate the algorithm on several artificial and real data sets and compare with several other recent techniques for dealing with endmember variability.

Description: The National Aeronautics and Space Administration (NASA) Soil Moisture Active Passive (SMAP) satellite mission was launched on January 31, 2015. The observatory was developed to provide global mapping of high-resolution soil moisture and freeze-thaw state every two to three days using an L-band (active) radar and an L-band (passive) radiometer. After an irrecoverable hardware failure of the radar on July 7, 2015, the radiometer-only soil moisture product became the only operational soil moisture product for SMAP. The product provides soil moisture estimates posted on a 36 km Earth-fixed grid produced using brightness temperature observations from descending passes. Within months after the commissioning of the SMAP radiometer, the product was assessed to have attained preliminary (beta) science quality, and data were released to the public for evaluation in September 2015. The product is available from the NASA Distributed Active Archive Center at the National Snow and Ice Data Center. This paper provides a summary of the Level 2 Passive Soil Moisture Product (L2_SM_P) and its validation against in situ ground measurements collected from different data sources. Initial in situ comparisons conducted between March 31, 2015 and October 26, 2015, at a limited number of core validation sites (CVSs) and several hundred sparse network points, indicate that the V-pol Single Channel Algorithm (SCA-V) currently delivers the best performance among algorithms considered for L2_SM_P, based on several metrics. The accuracy of the soil moisture retrievals averaged over the CVSs was 0.038 m 3 /m 3 unbiased root-mean-square difference (ubRMSD), which approaches the SMAP mission requirement of 0.040 m 3 /m 3 .

Description: This paper describes a general-purpose parallel scheme for efficiently focusing synthetic aperture radar (SAR) data on multicore-based shared-memory architectures. The rationale of the proposed tiling-based parallel focusing model is first discussed, and then, its implementation structure is illustrated. The adopted parallel solution, which is based on a canonical processing pattern, exploits a segmented-block-based approach and works successfully on data acquired by different spaceborne SAR platforms. Insofar as a significant portion of the focusing algorithm is amenable to tiling, our approach decomposes the problem into simpler subproblems of the same type, also providing a suitable mechanism to explicitly control the granularity of computation through the proper specification of the tiling at the different stages of the algorithm itself. Relevant implementation makes use of multithreading and high-performance libraries. Achievable performances are then experimentally investigated by quantifying the benefit of the parallelism incorporated into the prototype solution, thus demonstrating the validity of our approach. Accordingly, canonical performance metrics have been evaluated, and the pertinent scalability has been examined on different multicore architectures. Furthermore, in order to emphasize the practical ability of the proposed parallel model implementation to efficiently deal with data of different SAR sensors, a performance analysis has been carried out in different realistic scenarios including data acquired by the Envisat/ASAR, RADARSAT-1, and COSMO-SkyMed platforms.

Description: This paper investigates the problems of outliers and/or noise in surface segmentation and proposes a statistically robust segmentation algorithm for laser scanning 3-D point cloud data. Principal component analysis (PCA)-based local saliency features, e.g., normal and curvature, have been frequently used in many ways for point cloud segmentation. However, PCA is sensitive to outliers; saliency features from PCA are nonrobust and inaccurate in the presence of outliers; consequently, segmentation results can be erroneous and unreliable. As a remedy, robust techniques, e.g., RANdom SAmple Consensus (RANSAC), and/or robust versions of PCA (RPCA) have been proposed. However, RANSAC is influenced by the well-known swamping effect, and RPCA methods are computationally intensive for point cloud processing. We propose a region growing based robust segmentation algorithm that uses a recently introduced maximum consistency with minimum distance based robust diagnostic PCA (RDPCA) approach to get robust saliency features. Experiments using synthetic and laser scanning data sets show that the RDPCA-based method has an intrinsic ability to deal with outlier- and/or noise-contaminated data. Results for a synthetic data set show that RDPCA is 105 times faster than RPCA and gives more accurate and robust results when compared with other segmentation methods. Compared with RANSAC and RPCA based methods, RDPCA takes almost the same time as RANSAC, but RANSAC results are markedly worse than RPCA and RDPCA results. Coupled with a segment merging algorithm, the proposed method is efficient for huge volumes of point cloud data consisting of complex objects surfaces from mobile, terrestrial, and aerial laser scanning systems.

Description: We develop an approach for the detection of ruins of livestock enclosures (LEs) in alpine areas captured by high-resolution remotely sensed images. These structures are usually of approximately rectangular shape and appear in images as faint fragmented contours in complex background. We address this problem by introducing a rectangularity feature that quantifies the degree of alignment of an optimal subset of extracted linear segments with a contour of rectangular shape. The rectangularity feature has high values not only for perfectly regular enclosures but also for ruined ones with distorted angles, fragmented walls, or even a completely missing wall. Furthermore, it has a zero value for spurious structures with less than three sides of a perceivable rectangle. We show how the detection performance can be improved by learning a linear combination of the rectangularity and size features from just a few available representative examples and a large number of negatives. Our approach allowed detection of enclosures in the Silvretta Alps that were previously unknown. A comparative performance analysis is provided. Among other features, our comparison includes the state-of-the-art features that were generated by pretrained deep convolutional neural networks (CNNs). The deep CNN features, although learned from a very different type of images, provided the basic ability to capture the visual concept of the LEs. However, our handcrafted rectangularity-size features showed considerably higher performance.

Description: In this paper, a novel machine learning algorithm is presented for disaggregation of satellite soil moisture (SM) based on self-regularized regressive models (SRRMs) using high-resolution correlated information from auxiliary sources. It includes regularized clustering that assigns soft memberships to each pixel at a fine scale followed by a kernel regression that computes the value of the desired variable at all pixels. Coarse-scale remotely sensed SM was disaggregated from 10 to 1 km using land cover (LC), precipitation, land surface temperature, leaf area index, and in situ observations of SM. This algorithm was evaluated using multiscale synthetic observations in NC Florida for heterogeneous agricultural LCs. It was found that the rmse for 96% of the pixels was less than 0.02 m 3 /m 3 . The clusters generated represented the data well and reduced the rmse by up to 40% during periods of high heterogeneity in LC and meteorological conditions. The Kullback–Leibler divergence (KLD) between the true SM and the disaggregated estimates is close to zero, for both vegetated and bare-soil LCs. The disaggregated estimates were compared with those generated by the principle of relevant information (PRI) method. The rmse for the PRI disaggregated estimates is higher than the rmse for the SRRM on each day of the season. The KLD of the disaggregated estimates generated by the SRRM is at least four orders of magnitude lower than those for the PRI disaggregated estimates, whereas the computational time needed was reduced by three times. The results indicate that the SRRM can be used for disaggregating SM with complex nonlinear correlations on a grid with high accuracy.

Description: Two-step ways are often used for fusing both panchromatic (PAN) and multispectral (MS) images for classification, e.g., classifying MS images sharpened by PAN images or directly pouring fine spatial details of PAN images into a classification result of MS images. In this paper, we present a unified Bayesian framework to iteratively discovering semantic segments from PAN images and allocating cluster labels for the segments using MS images. Specifically, the probabilistic generative process of both PAN and MS images is explained with a generalized metaphor of the Chinese restaurant franchise (CRF) (gCRF), in which the two iterative random processes, i.e., table selection and dish selection , are adapted to discovering semantic segments in PAN images and inferring cluster labels for the discovered segments using MS images, respectively. Our major contributions are twofold: 1) The CRF is generalized into an image fusion framework by elegantly decomposing its two random processes, and 2) the random process of table selection in the CRF is transformed into stochastic image segmentation by enforcing spatial constraints over adjacent pixels. The qualitative analysis of experimental results shows that the gCRF can effectively utilize both the spatial details of the PAN images and the spectral information of the MS images. In terms of quantitative evaluation, the gCRF is comparable with support vector machine-based supervised classification methods.

Description: Ice core drillings have been performed in various zones in Antarctica and Greenland to obtain climatological information, study ice properties, or analyze air and dust encapsulated in the ice during the quaternary period. During these procedures, a set of measurements to characterize the ice and to evaluate its physical and chemical properties are usually performed in situ . In particular, using known temperature and dielectric profiles (DEP measurements), it is possible to evaluate the ice electromagnetic power absorption profile, valid at the drilling site. In the last decades, bedrock characterization through radio echo sounding surveys has been improved by the analysis of the power of radar echoes. This way, analysis of the electromagnetic properties of bedrock interfaces makes it possible to assess the physical characteristics and to distinguish between wet and dry conditions. Power variation of the received echoes also depends on ice absorption and on bedrock reflectivity due to specific physical conditions of the ice. In this paper, the propagation of electromagnetic waves through the ice sheet is examined, and in particular, a new method for establishing the electromagnetic absorption profile for ice from core drilling measurements is proposed and discussed. Variation in the ice absorption is deduced, starting from the analysis of ice core data from the European Project for Ice Coring in Antarctica (EPICA) at the Concordia station (Antarctica) and from the Greenland Ice Core Project (GRIP) site (Greenland). This direct method of measurement is proposed with the aim of defining common characteristics of the ice absorption rate that are valid both in Antarctica and in Greenland.

Description: The algorithm of synthetic aperture radar automatic target recognition (SAR-ATR) is generally composed of the extraction of a set of features that transform the raw input into a representation, followed by a trainable classifier. The feature extractor is often hand designed with domain knowledge and can significantly impact the classification accuracy. By automatically learning hierarchies of features from massive training data, deep convolutional networks (ConvNets) recently have obtained state-of-the-art results in many computer vision and speech recognition tasks. However, when ConvNets was directly applied to SAR-ATR, it yielded severe overfitting due to limited training images. To reduce the number of free parameters, we present a new all-convolutional networks (A-ConvNets), which only consists of sparsely connected layers, without fully connected layers being used. Experimental results on the Moving and Stationary Target Acquisition and Recognition (MSTAR) benchmark data set illustrate that A-ConvNets can achieve an average accuracy of 99% on classification of ten-class targets and is significantly superior to the traditional ConvNets on the classification of target configuration and version variants.

Description: The demand for renewable power production has fostered an exponential increase in the size and number of wind turbines. This expanding source of power generation is sometimes at odds with maintaining the effective operation of radar systems in air traffic control, defense, weather prediction, and severe-storm-tracking applications. With the recent upgrade of the NEXRAD weather radar system to enable dual-polarization observations, a dual-polarization characterization effort of wind turbines is warranted. Focusing on weather radar applications, a characterization of ground clutter, precipitation, and wind turbines is presented here using a consistent unified treatment. This characterization effort directly compares the dual-polarization radar signatures of these three classes of scatterers. The physical characteristics of wind turbines (particularly their cyclostationary behavior) are exploited to identify unique dual-polarization radar signatures.

Description: The panchromatic and multispectral (PMS) sensor is a high spatial resolution sensor aboard the GF-1 satellite launched on April 26, 2013. This paper focuses on the cross-calibration of the PMS sensor using Terra/Moderate-Resolution Imaging Spectroradiometer (MODIS) and Landsat 8/Operational Land Imager (OLI). Two matched-image adjustment factors (MIAFs) are used in the cross-calibration which are the radiance MIAF and reflectance MIAF. Two test sites are chosen as the regions of interest. One is the Dunhuang test site, which has been used for the vicarious calibration of Chinese satellites since later 1990s. The other is the Golmud test site, which is a new site with no ground measured data available. The results show that both the Dunhuang and Golmud test sites can be used for cross-calibration. This paper reveals that the cross-calibration of the PMS sensor using OLI is better than using MODIS, as the calibration coefficient difference between the two test sites with OLI is smaller than that with MODIS. The uncertainty analysis results show that the uncertainty of cross-calibration using OLI is 5%–7% when the ground data are not available.

Description: A novel approach for a wireless transmission of MWD and LWD signals in lossy formations has been proposed in this paper. First, the quasi-TEM mode of the extremely low frequency (ELF) near field is excited by a voltage gap on the pipe, guided by and propagated along the metal pipe string in LWD or MWD (logging/measurement while drilling) environment. Once the signal received on the ground is weaker than a certain threshold as the drilling depth increases, a relay segment (transceiver) is installed into the pipe string to receive, amplify, and retransmit the signals to extend the transmission distance and ensure the required propagation range of the ELF wave. In this paper, the details of the transceiver antennas and the relay segment suitable for efficient coupling to quasi-TEM mode have been given first, and then, the numerical simulations of the ELF wave propagation in LWD or MWD environment and the coupling between the guided wave and the relay segment have been made to show the coupling performance of the relay segment to quasi-TEM mode quantitatively. It has been shown that the propagation attenuation of the ELF near field with the guidance of the metal pipe string is much less than that of the space wave in the same lossy media without the guidance of the pipe string, and the scheme of the wireless relay provides an efficient way to extend the propagation range of the EM telemetry evidently and open a new freedom of frequency optimization for wireless transmission systems.

Description: The seasonal and interannual dynamics of tropical rainforests play a critical role in the global carbon cycle and climate change. This paper retrieved and compared land surface phenology from observations acquired by the Spinning Enhanced Visible and Infrared Imager (SEVIRI) onboard geostationary satellites and the Moderate Resolution Imaging Spectroradiometer (MODIS) on polar-orbiting satellites over the Congo Basin. To achieve this, we first retrieved canopy greenness cycles (CGCs) and their transition timing from two-band enhanced vegetation index (EVI2) derived from SEVIRI and MODIS data between 2006 and 2013. We then assessed the influences of SEVIRI and MODIS data quality on the reconstruction of the EVI2 temporal trajectory, the detection of the CGC onset and end timing, and the total number of successful CGC retrievals. The significance of influences was determined using the one-tailed two-sample Kolmogorov–Smirnov test. The results indicate that diurnal SEVIRI observations greatly increased the probability of capturing cloud-free daily EVI2 in the rainforest-dominated region of the Congo Basin, where the proportion of good quality (PGQ) observations during a CGC was up to 80% higher than that from MODIS. As a result, the double annual CGCs of the Congo Basin rainforests were well identified from SEVIRI but sparsely detected from MODIS, whereas the single annual CGC in the savanna-dominated northern and southern Congo Basin was successfully retrieved from both SEVIRI and MODIS. Moreover, the decreases of PGQ in an EVI2 time series were found to significantly increase the uncertainties of retrieved phenological timings and increase the probabilities of CGC retrieval failures.

Description: Active learning is an area of significant ongoing research interest for the classification of remotely sensed data, where obtaining efficient training data is both time consuming and expensive. The goal of active learning is to achieve high classification performance by querying as few samples as possible from a large unlabeled data pool. Traditional active learning frameworks all assume the existence of labeled samples for all classes of interest. However, in real-world applications, the unlabeled data pool may contain data from unknown classes that we are not aware of in advance, and a quick detection of them is useful for enriching our training set. In this scenario, traditional active learning methods may not effectively and rapidly detect the unknown classes. We proposed an active learning framework which provides robust classification performance with minimum manual labeling effort while simultaneously discovering unknown (missing) classes. The discovery of unknown classes is particularly suited to an active learning framework where an annotator is in the loop. A Dirichlet process mixture model is utilized in our proposed method to cluster the labeled and unlabeled samples as a whole. If unknown classes exist, they will emerge as new clusters which are different from other existing clusters occupied by known classes, and then, the proposed query strategy will give priority to querying samples in the new clusters. We present experimental results with hyperspectral data to show that our method provides better classification performance compared to existing active learning methods with or without unknown classes.

Description: A novel oriented volume over ground (OVoG) inversion scheme is developed and tested on a data set of simulated agricultural scenarios and real SAR acquisitions. The algorithm makes use of multibaseline measurements to estimate the whole set of the OVoG structural parameters (e.g., crop height, differential extinction between eigenpolarizations, and ground-to-volume ratios) and is significantly robust against nonvolumetric decorrelation contributions. The theoretical assessment points out that, in the dual-baseline case, the vegetation height $h_{V}$ can be estimated with a relative root-mean-square deviation (%RMSD) of 7.8% if the selected baselines fulfill the condition $1.2 <kappa_{z}h_{V} <2.8 text{rad}$ ( $kappa_{z}$ is the vertical wavenumber). Furthermore, the variance of the estimates is inversely related to the number of baselines $Nb$ . Compared with the dual-baseline case, the RMSD of the differential extinction is reduced by 45% (from 1.1 to 0.6 dB/m) when $Nb=5$ baselines are employed, whereas its mean bias is independent of $Nb$ . The proposed scheme has been assessed using a set of repeat-pass F-SAR acquisitions at L-, C-, and X-band of an agricultural area in Germany. Using two baselines, the height of maize and rape fields is estimated with an average 10% %RMSD if the inversion is carried out over L-band acquisitions. On the other hand, when X-band data are employed, one can obtain reliable estimates of wheat and barley height, with a %RMSD better than 24%. The - tudy also indicates the existence of differential wave propagation effects through maize ( $Deltasigma=sigma_mathrm{VV}-sigma_mathrm{HH}$ between 0.7 and 1 dB/m) and rape $(Deltasigma=-0.8 text{dB/m})$ canopies at L-band.

Description: To obtain accurate classification results of hyperspectral images, both spectral and spatial information should be fully exploited in the classification process. In this paper, we propose a novel method using independent component analysis (ICA) and edge-preserving filtering (EPF) via an ensemble strategy for the classification of hyperspectral data. First, several subsets are randomly selected from the original feature space. Second, ICA is used to extract spectrally independent components followed by an effective EPF method, to produce spatial features. Two strategies (i.e., parallel and concatenated) are presented to include the spatial features in the analysis. The spectral–spatial features are then classified with a random forest or a rotation forest classifier. Experimental results on two real hyperspectral data sets demonstrate the effectiveness of the proposed methods. A sensitivity analysis of the new classifiers is also performed.

Description: In a plethora of applications dealing with inverse problems, e.g., image processing, social networks, compressive sensing, and biological data processing, the signal of interest is known to be structured in several ways at the same time. This premise has recently guided research into the innovative and meaningful idea of imposing multiple constraints on the unknown parameters involved in the problem under study. For instance, when dealing with problems whose unknown parameters form sparse and low-rank matrices, the adoption of suitably combined constraints imposing sparsity and low rankness is expected to yield substantially enhanced estimation results. In this paper, we address the spectral unmixing problem in hyperspectral images. Specifically, two novel unmixing algorithms are introduced in an attempt to exploit both spatial correlation and sparse representation of pixels lying in the homogeneous regions of hyperspectral images. To this end, a novel mixed penalty term is first defined consisting of the sum of the weighted $ell_{1}$ and the weighted nuclear norm of the abundance matrix corresponding to a small area of the image determined by a sliding square window. This penalty term is then used to regularize a conventional quadratic cost function and impose simultaneous sparsity and low rankness on the abundance matrix. The resulting regularized cost function is minimized by: 1) an incremental proximal sparse and low-rank unmixing algorithm; and 2) an algorithm based on the alternating direction method of multipliers . The effectiveness of the proposed algorithms is illustrated in experiments conducted both on simulated and real data.

Description: Land surface component temperatures are important inputs in longwave radiation and evapotranspiration estimation models. Most component temperature inversion approaches focus only on two components, namely, soil and leaves, because space-based multiangle observations are lacking. This approach is inconsistent with ground-based measurements, which suggest that the temperatures of sunlit and shaded soil may significantly differ. This paper explores a three-component temperature inversion scheme that uses airborne multiangle thermal infrared observations to decrease the difference between the retrieved data and the actual subpixel temperature distribution. The FR97 model, which is an analytical directional brightness temperature model that was modified by dividing the soil component into sunlit and shaded portions, is adopted to calculate the matrix of component effective emissivity, which links multiangular observations and component temperatures. The new forward model and the inversion scheme are assessed using simulated data sets from the Scattering by Arbitrarily Inclined Leaves (4SAIL) model. The results indicate that the modified FR97 model provides good precision and that the inversion scheme based on the modified FR97 model is appropriate because of the model's simplicity and accuracy and the inversion's low sensitivity to noise. The inversion scheme is validated using airborne data collected by the wide-angle infrared dual-mode line/area array scanner over an area planted with maize and ground measurements collected during the Heihe Watershed Allied Telemetry Experimental Research campaign. The results indicate that the root mean square errors of the component temperatures of the leaves, sunlit soil, and shaded soil were 0.72 °C, 1.55 °C, and 2.73 °C, respectively. Because of the modified FR97's straightforward form and acceptable precision, we recommend this new retrieval scheme as an option for retrieving the compon- nt temperatures of leaves, sunlit soil, and shaded soil.

Description: This paper presents a novel method for estimating the model parameters of the probability integral method (PIM) based on the line-of-sight deformation derived from the interferometric synthetic aperture radar. Then, it applies the settled PIM to forward predict the horizontal and vertical displacements induced by the extraction of a new working panel. The proposed method first constructed the functional relationship between the InSAR-derived LOS deformation and the model parameters of PIM. Subsequently, an improved genetic algorithm (GA), in which gross error elimination was imposed, was proposed, and used to estimate the model parameters of PIM with a large number of LOS deformation measurements. The estimated model parameters and PIM were then employed to forward predict the horizontal and vertical displacements induced by the extraction of a working panel. Simulated experiments show that the rmses of the predicted displacements along the up–down, west–east, and north–south directions are 1.5, 0.9, and 2.5 mm, respectively. Real data experiments over the Qianyingzi coal mining area of China indicate that the predicted displacements are highly consistent with those by field surveys, with rmses of 4.1 and 3 cm for the vertical and horizontal directions, respectively. These imply that the proposed approach can be a very promising tool for predicting the mining-induced displacements and will potentially contribute to the assessing and forecasting of possible geological hazards in the mining area.

Description: This paper reports the first use of a multifrequency range imaging (RIM) technique for observing E-region field-aligned irregularities (FAIs) in the midlatitude ionosphere. The Middle and Upper atmosphere Radar (MUR; 34.85°N, 136.10°E) was used to conduct experiments with five equally spaced frequencies between 46.25 and 46.75 MHz. Three types of RIM data were examined: data with 13-element binary Barker codes, with 16-element binary complementary codes, and without phase codes. Moreover, two calibration approaches were adopted to validate the applicability of the RIM technique, which functioned as intended. Excellent RIM performance such as the ability to resolve several striations in an echo region of FAIs was demonstrated. However, sidelobe echoes caused by pulse coding mechanisms were occasionally observed at altitudes above and below the source regions in the coded data. Therefore, a procedure was developed according to one of the calibration approaches to identify and remove such kind of sidelobe echoes, which was shown to be applicable for the complementary-coded data. In addition to FAIs, a thin plasma layer with a thickness of approximately 1 km was identified as being structured with some tilted finer structures, which could not be observed in the original intensity images with a range resolution of 600 m. Preliminary estimates of the Doppler velocities indicated that a wind shear effect could be the cause of such tilted finer structures.

Description: Characterization of the microphysical and thermodynamical properties of convective events over the tropical coastal station Thiruvananthapuram (TVM) has been carried out based on multiyear microwave radiometer profiler observations. The analyses have been extended to develop a methodology to identify convective events, which is based on the radiometric brightness temperature $(T_b)$ difference threshold, at 30 and 22.23 GHz channels, and the results are compared with reflectivity and rainfall intensity deduced from concurrent and collocated disdrometer measurements. Eighty-four of such convections were identified using the aforementioned methodology over the station during 2010–2013, i.e., both for pre- and post-Indian summer monsoon months, and further evaluated by computing their stability indexes. The occurrence of convective systems peaks in the afternoon and early-morning hours with genesis, respectively, over the land and the sea.

Description: Singularity analysis has proven to be a complementary tool to the Advanced Scatterometer (ASCAT) inversion residual (or maximum likelihood estimator) in terms of wind quality control (QC). In this paper, a new implementation scheme of singularity exponent (SE) is developed for ASCAT data analysis. It combines the wavelet projections of the gradient measurements of multiple parameters into the analysis, ensuring that the analyzed parameters contribute equally to the final singularity map. Therefore, the underlying geophysical phenomena in the different ASCAT-derived parameters can be effectively revealed simultaneously on a unique map of SEs. The validation using both buoy winds and European Centre for Medium-Range Weather Forecasting forecast wind output shows that the newly derived SE significantly improves the current ASCAT wind QC. In particular, poor-quality ASCAT measurements at low-wind and high-variability conditions $(w < 4 text{m/s})$ can be effectively screened using the new SE.

Description: We report an initial investigation of the new location method of a very high frequency (VHF) radiation source, using signals recorded at the International Space Station. A VHF interferometer (VITF) has two VHF sensors. Locating lightning with VHF bands is useful to locate the position of the charge distribution in the thunderstorm. The location method of a radio source proposed used two direction-of-arrival estimation techniques. One is the interferometric technique, and another is based on the ionospheric propagation delay measurement of received signals. The combination of the two techniques provides two angular positions of the radiation source. When an altitude of a radiation source is assumed, we can determine two possible positions. One of the two positions was associated with the radiation source, while the other was not. In this paper, we compared the position of lightning and sprite imager (LSI) data, which are simultaneously captured during a lightning emission, with the locating position near the emission. The data set of the VITF within 100 ms of the optical lightning emission captured with the LSI was used. The temporally simultaneous event seems to be associated with the same lightning event. The estimated radiation positions were spatially in close agreement with the optical lightning positions captured with LSI, under nighttime ionosphere conditions. From statistical analysis, the spatial difference of the standard deviation changed from 15.3 to 30.8 km depending on the installation direction of the VHF sensors. The usefulness and limitations of the method are also discussed.

Description: Modeling of terrain topography is crucial for vegetated areas given that even small slopes impact and alter the radar wave interactions between the ground and the overlying vegetation. Current missions either exclude pixels with large topographic slopes or disregard the terrain topography entirely, potentially accumulating substantial modeling errors and therefore impacting the retrieval performance over such sloped pixels. The underlying terrain topography needs to be considered and modeled to obtain a truly general and accurate radar scattering model. In this paper, a flexible and modular model is developed: the vegetation is considered by a multilayered multispecies vegetation model capable of representing a wide range of vegetation cover types ranging from bare soil to dense forests. The ground is incorporated with the stabilized extended boundary condition method, allowing the representation of an $N$ -layered soil structure with rough interfaces. Terrain topography is characterized by a 2-D slope with two tilt angles $(alpha, beta)$ . Simulation results for an evergreen forest show the impact of a 2-D slope for a range of tilt angles: a 10 $^{circ}$ tilt in the plane of incidence translates to a change of up to 15 dB in HH, 10 dB in VV, and 1.5 dB in HV for the total radar backscatter. Terrain topography is shown to be crucial for accurate forward modeling, especially over forested areas.

Description: This paper presents an automated algorithm for rapidly and effectively detecting cars directly from large-volume 3-D point clouds. Rather than using low-order descriptors, a multilayer feature generation model is created to obtain high-order feature representations for 3-D local patches through deep learning techniques. To handle cars with different levels of incompleteness caused by data acquisition ways and occlusions, a hierarchical visibility estimation model is developed to augment Hough voting. Considering scale and orientation variations in the azimuth direction, a set of multiscale Hough forests is constructed to rotationally cast votes to estimate cars' centroids. Quantitative assessments show that the proposed algorithm achieves average completeness, correctness, quality, and $F_{1}$ -measure of 0.94, 0.96, 0.90, and 0.95, respectively, in detecting 3-D cars. Comparative studies also demonstrate that the proposed algorithm outperforms the other four existing algorithms in accurately and completely detecting 3-D cars from large-scale 3-D point clouds.

Description: We propose an approach to voxelize bathymetric full-waveform LiDAR (Light Detection and Ranging) to generate orthowaveforms and use them to estimate shallow water bathymetry and turbidity with a nonparametric support vector regression (SVR) method. Two distinct shallow rivers were investigated ranging from clear to turbid water; hyperspectral imagery and traditional full-waveform LiDAR processing were also investigated as a baseline for comparison with the proposed orthowaveform strategy. The orthowaveform showed significant correlation to water depth in both scenarios and outperformed hyperspectral imagery for water depth estimation in more turbid water. The orthowaveforms showed similar performance to full-waveform LiDAR point observations for bathymetry estimation in clear water and outperformed the bathymetry performance of full-waveform processing in turbid water. The orthowaveforms also showed similar performance to hyperspectral imagery for predicting water turbidity in turbid water, with a root mean square error (RMSE) of 1.32 NTU. The fusion of both hyperspectral imagery and orthowaveforms was also investigated and gave superior performance to using either data set alone. The fused data set was able to estimate depth in clear and turbid water with an RMSE of 10 and 21 cm, respectively, and turbidity with an RMSE of 1.16 NTU.

Description: Linearly derived features have been widely used in hyperspectral image classification to find linear separability of certain classes in recent years. Moreover, nonlinearly transformed features are more effective for class discrimination in real analysis scenarios. However, few efforts have attempted to combine both linear and nonlinear features in the same framework even if they can demonstrate some complementary properties. Moreover, conventional multiple-feature learning-based approaches deal with different features equally, which is not reasonable. This paper proposes an efficient multiple-feature learning-based model with adaptive weights for effectively classifying complex hyperspectral images with limited training samples. A new diversity kernel function is proposed first to simulate the vision perception and analysis procedure of human beings. It could simultaneously evaluate the contrast differences of global features and spatial coherence. Since existing multiple-kernel feature models are always time-consuming, we then design a new adaptive weighted multiple kernel learning method. It employs kernel projection, which could lower the dimensionalities and also learn kernel weights to further discriminate the classification boundaries. For combining both linear and nonlinear features, this paper also proposes a novel decision fusion strategy. The method combines linear and multiple kernel features to balance the classification results of different classifiers. The proposed scheme is tested on several hyperspectral data sets and extended to multisource feature classification environment. The experimental results show that the proposed classification method outperforms most of the existing ones and significantly reduces the computational complexity.

Description: Against an intricate infrared cloudy-sky background, jamming objects such as the edges of clouds in the scene have a similar thermal intensity measure with respect to the background as small targets. This may cause high false alarm rates and low probabilities of detection according to conventional small target detection methods. In this paper, we propose a weighted local difference measure (WLDM)-based scheme for the detection of small targets against various complex cloudy-sky backgrounds. Initially, a WLDM map is achieved to simultaneously enhance targets and suppress background clutters and noise. In this way, the true targets can be easily separated from jamming objects. After that, a simple adaptive threshold is used to segment the targets. More than 460 infrared small target images against diverse intricate cloudy-sky backgrounds were utilized to validate the detection capability of the WLDM-based method. Experimental results demonstrate that the proposed algorithm not only works more robustly for different cloudy-sky backgrounds, target movements, and signal-to-clutter ratio (SCR) values but also has a better performance with regard to the detection accuracy, in comparison to traditional baseline methods. In particular, the proposed method is able to significantly improve SCR values of the images.

Description: The FLuorescence EXplorer (FLEX) mission, selected as the European Space Agency's eighth Earth Explorer, aims to globally measure the sun-induced-chlorophyll-fluorescence spectral emission from terrestrial vegetation. In the frame of the FLEX mission, several industrial and scientific studies have analyzed the instrument design, image processing algorithms, or modeling aspects. At the same time, a common tool is needed to address the overall FLEX mission performance by combining all these features. For this reason, an end-to-end mission performance simulator has been developed for the FLEX mission (FLEX-E). This paper describes the FLEX-E software design, which combines the generation of complex synthetic scenes with an advanced modeling of the instrument behavior and the full processing scheme up to the final fluorescence product. The results derived from FLEX-E simulations indicate that the instrument and developed image processing algorithms are able to retrieve the sun-induced fluorescence with an accuracy below the 0.2 $text{mW}cdottext{m}^{-2}cdottext{sr}^{-1}cdot text{nm}^{-1}$ mission requirement. It is expected that FLEX-E will not only optimize the FLEX retrieval algorithms and technical requirements, but also serve as the baseline for the ground processing implementation and testing of calibration/validation procedures.

Description: Nonnegative matrix factorization (NMF), which can lead to nonsubtractive parts-based representation, has been demonstrated to be effective for dimensionality reduction of hyperspectral imagery (HSI). However, existing NMF methods applied to HSI use only a single spectral feature and do not take into consideration spatial information, such as texture or morphological features, while it has been widely acknowledged that exploiting multiple features can improve performance. Consequently, a variant of orthogonal NMF, which can not only achieve a nonnegative factorization but also exploit the complementary information that arises among heterogeneous features, is proposed for hyperspectral dimensionality reduction. The proposed method, which couples orthogonal NMF with a previous multiple-features-combining algorithm, yields a discriminative low-dimensional feature representation that matches the intuition that parts should sum to produce a whole. An efficient multiplicative updating procedure is derived, and its local convergence is guaranteed theoretically. Experimental results on two hyperspectral data sets demonstrate the effectiveness of the proposed method.

Description: Polarimetric incoherent target decomposition aims at accessing physical parameters of illuminated scatters through the analysis of the target coherence or covariance matrix. In this framework, independent component analysis (ICA) was recently proposed as an alternative method to eigenvector decomposition to better interpret non-Gaussian heterogeneous clutter (inherent to high-resolution synthetic aperture radar systems). Until now, the two main drawbacks reported of the aforementioned method are the greater number of samples required for an unbiased estimation, when compared to the classical eigenvector decomposition, and the inability to be employed in scenarios under the Gaussian clutter assumption. In this paper, both drawbacks are analyzed. First, a Monte Carlo approach is performed in order to investigate the bias in estimating Touzi's target-scattering-vector-model parameters when ICA is employed. Simulated data and a RAMSES X-band image acquired over Brétigny, France, are taken into consideration to investigate the bias estimation under different scenarios. Finally, the performance of the algorithm is also evaluated under the Gaussian clutter assumption and when spatial correlation is introduced in the model.

Description: Satellite-derived land surface phenology (LSP) serves as a valuable input source for many environmental applications such as land cover classifications and global change studies. Commonly, LSP is derived from coarse-resolution (CR) sensors due to their well-suited temporal resolution. However, LSP is increasingly demanded at medium resolution (MR), but inferring LSP directly from MR imagery remains a challenging task (e.g., due to acquisition frequency). As such, we present a methodology that directly predicts MR LSP on the basis of the respective CR LSP and MR reflectance imagery. The approach considers information from the local pixel neighborhood at both resolutions by utilizing several prediction proxies, including spectral distance and multiscale heterogeneity metrics. The prediction performs well with simulated data $(R^{2} = 0.84)$ , and the approach substantially reduces noise. The size of the smallest reliably predicted object coincides with the effective CR pixel size (i.e., field-of-view). Nevertheless, even subpixel objects can be reliably predicted provided that pure CR pixels are located within the search radius. The application to real MODIS LSP and Landsat reflectance well preserves the phenological landscape composition, and the spatial refinement is especially striking in heterogeneous agricultural areas, where, for example, the circular shape of center pivot irrigation schemes is successfully restored at MR.

Description: Floods are among the most destructive natural disasters worldwide. In flood disaster management programs, flood mapping is an initial step. This research proposes an efficient methodology to recognize and map flooded areas by using TerraSAR-X imagery. First, a TerraSAR-X satellite image was captured during a flood event in Kuala Terengganu, Malaysia, to map the inundated areas. Multispectral Landsat imagery was then used to detect water bodies prior to the flooding. In synthetic aperture radar (SAR) imagery, the water bodies and flood locations appear in black; thus, both objects were classified as one. To overcome this drawback, the class of the water bodies was extracted from the Landsat image and then subtracted from that extracted from the TerraSAR-X image. The remaining water bodies represented the flooded locations. Object-oriented classification and Taguchi method were implemented for both images. The Landsat images were categorized into three classes, namely, urban, vegetation, and water bodies. By contrast, only water bodies were extracted from the TerraSAR-X image. The classification results were then evaluated using a confusion matrix. To examine the efficiency of the proposed method, iterative self-organizing data analysis technique (ISODATA) classification method was applied on TerraSAR-X after employing the segmentation process during object-oriented–rule-based method, and the results were compared. The overall accuracy values of the classified maps derived from TerraSAR-X using the rule-based method and Landsat imagery were 86.18 and 93.04, respectively. Consequently, the flooded locations were recognized and mapped by subtracting the two classes of water bodies from these images. The acquired overall accuracy for TerraSAR-X using ISODATA was considerably low at only 57.98. The current research combined the methods and the optimization technique used as an innovative flood detection application. The successful production of a reliable and accu- ate flood inventory map confirmed the efficiency of the methodology. Therefore, the proposed method can assist researchers and planners in implementing and expediting flood inventory mapping.

Description: Processing scanning synthetic aperture radar (ScanSAR) data using a stripmap processor, which is called full-aperture processing, has been the choice of many researchers. ScanSAR data are known to require very high azimuth coregistration precision which is usually achieved by a geometrical coregistration followed by a spectral diversity coregistration on the ScanSAR burst. However, for full-aperture processing, since individual bursts are no longer available for spectral diversity coregistration, the cross-correlation method in practice is still used to coregister ScanSAR data as stripmap data. We analyze the azimuth coregistration precision requirement of full-aperture processing and find that its requirement can be significantly relaxed. This is confirmed by a number of experiments, including simulations and real data experiments whose results are in good agreement with each other. An additional experiment on the cross-correlation method supports its use in full-aperture processing. Concluding from the experimental results, we further propose a simple method to evaluate the azimuth coregistration precision requirement for practical use. Finally, we present examples with ALOS-2 ScanSAR data.