This PDF file contains the editorial “2016 List of Reviewers” for JARS Vol. 11 Issue 01

Sparsity preserving projection is a well-known dimensionality reduction method that preserves the sparse representation relationship among data in low-dimensional space, which is beneficial for classification. The idea of sparsity preserving is applied to band selection for hyperspectral classification. Considering the spatial distribution characteristic of hyperspectral image (HSI), a spatial–spectral regularized sparse graph (ssRSG), which could utilize the spatial–spectral information in HSI to promote the discriminability of extracted local structure, is proposed. For band selection, the L2,1 norm is applied to restrain the projection matrix and make a few bands with high importance scores, which are computed by the contribution of bands in a projection matrix. According to the importance score, more important bands are selected. Two real hyperspectral images are used to validate the performance of the proposed method.

This PDF file contains the editorial “Special Section Guest Editorial: Remote Sensing for Investigating the Coupled Biogeophysical and Biogeochemical Process of Harmful Algal Blooms” for JARS Vol. 11 Issue 01

Satellite ocean color remote sensing techniques, coupled with <italic<in situ</italic< data, were used to examine the spatial extent and evolution of four <italic<Karenia brevis</italic< blooms on the West Florida Shelf (WFS) in 2004, 2005, 2006, and 2011. Observations were obtained with the moderate resolution imaging spectroradiometer (MODIS-Aqua). These four blooms were delineated by combining remote-sensing reflectance at 555 nm and normalized fluorescence line height. In 2004 and 2005, the WFS was affected by several hurricanes, including the category 5 storm Hurricane Katrina. These hurricanes led to increased river discharge and vertical mixing which favored bloom intensification and dispersion. No hurricanes passed over the WSF in 2006; however, storms in south Florida may have aided bloom intensification via increased river discharge. In 2011, a bloom appeared off Venice, Florida, where several small creeks discharge. The bloom moved south toward Charlotte Harbor where it intensified and lingered for several months as it received nutrients from riverine discharge and upwelling events. While it is difficult to identify initiation stages of a <italic<K. brevis</italic< bloom (<inline-formula< <mml:math display="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML"< <mml:mrow< <mml:mo form="prefix"<<</mml:mo< <mml:mo<∼</mml:mo< <mml:mn<50,000</mml:mn< <mml:mtext<  </mml:mtext< <mml:mtext<cells</mml:mtext< <mml:mtext< </mml:mtext< <mml:msup< <mml:mrow< <mml:mi mathvariant="normal"<L</mml:mi< </mml:mrow< <mml:mrow< <mml:mo<−</mml:mo< <mml:mn<1</mml:mn< </mml:mrow< </mml:msup< </mml:mrow< </mml:math< </inline-formula<) using satellite imagery, the techniques used here provide information about bloom evolution (size, duration, and advection) and insight into factors affecting bloom dynamics.

Several satellite-based methods have been used to detect and trace Karenia brevis red tide blooms in the eastern Gulf of Mexico. Some require data statistics and multiple data products while others use a single data product. Of these, the MODIS normalized fluorescence line height (nFLH) has shown its advantage of detecting blooms in waters rich in colored dissolved organic matter, thus having been used routinely to assess bloom conditions by the Florida Fish and Wildlife Conservation Commission (FWC), the official state agency of Florida responsible for red tide monitoring and mitigation. However, elevated sediment concentrations in the water column due to wind storms can also result in high nFLH values, leading to false-positive bloom interpretation. Here, a modified nFLH data product is developed to minimize such impacts through empirical adjustments of the nFLH values using MODIS-derived remote sensing reflectance in the green band at 547 nm. The new product is termed as an algal bloom index (ABI), which has shown improved performance over the original nFLH in both retrospective evaluation statistics and near real-time applications. The ABI product has been made available in near real-time through a Web portal and has been used by the FWC on a routine basis to guide field sampling efforts and prepare for red tide bulletins distributed to many user groups.

Numerous empirical algorithms have been operationally used to retrieve global ocean chlorophyll-a (Chla) concentrations from ocean color satellite data, for example, the OC4V4 algorithm for sea-viewing wide field-of-view sensor and the OC3M algorithm for moderate-resolution imaging spectroradiometer. However, the algorithms have been established and validated based on in situ data mainly measured under low to moderate solar zenith angle (SZA) (<70  deg). Currently, with the development of geostationary satellite ocean color remote sensing, enabling observations from early morning to late afternoon, it has become necessary to know whether the empirical Chla algorithms can be applied to high SZA. The performances of seven widely used Chla algorithms (i.e., OC2, OC3M, OC3V, OC4V4, Clark, ocean-color index, and Yellow Sea Large Marine Ecosystem Ocean Color Work Group) under high SZAs were evaluated using the global in situ ocean color dataset (NASA bio-optical marine algorithm dataset). The results show that the performances of all seven algorithms decreased significantly under high SZAs compared with those under low to moderate SZAs. For instance, for the OC4V4 algorithm, the relative percent difference (RPD) and root-mean-square error (RMSE) were 13.78% and 1.66  μg/L for the whole dataset and 3.95% and 1.49  μg/L for SZAs ranging from 30 deg to 40 deg, respectively. However, RPD and RMSE values increased to 30.45% and 6.10  μg/L for SZAs larger than 70 deg. Adjusting the coefficients of the algorithms using the in situ dataset with high SZA can only slightly improve the performance. The bidirectional remote sensing reflectance can explain the underestimation of the retrieved Chla under high SZA, and the low sensitivity of the blue–green ratio is responsible for the relatively larger scatter between the retrieved and the in situ Chla under high SZA.

Monitoring algal blooms using traditional methods is expensive and labor intensive. The use of satellite technology can attenuate such limitations. A common problem associated with the application of such technology is the need to eliminate the effects of atmosphere, which can be, at least, a time-consuming task. Thus, a remote sensed algal bloom monitoring system needs a simple algorithm which is nonsensitive to atmospheric correction and that could be applied to small aquatic systems. A slope algorithm (SAred−NIR) was developed to detect and map the extension of algal blooms using the Landsat 8/Operational Land Imager. SAred−NIR was shown to have advantages over other commonly used indices to monitor algal blooms, such as normalized difference vegetation index (NDVI), normalized difference water index, and floating algae index. SAred−NIR was shown to be less sensitive to different atmospheric corrections, less sensitive to thin clouds, and less susceptible to confusion when classifying water and moderate bloom conditions. Based on ground truth data from Eagle Creek Reservoir, Indiana, SAred−NIR showed an accuracy of 88.46% while NDVI only showed a 46.15% accuracy. Finally, based on qualitative and quantitative results, SAred−NIR can be used as a tool to improve the governance of small size water resources.

To compensate for the limitations of optical remote sensing when restricted by cloud cover, it is worth exploring how to detect cyanobacterial blooms using synthetic aperture radar (SAR), which can penetrate clouds. A satellite–ground synchronous experiment was conducted in Lake Taihu, the third largest freshwater lake in China. A lipopeptide biosurfactant was detected in the algal scum layer, with an average content of 1.8  g/L. The viscosity (1.41 to 332 mPa.s) of the scum was significantly higher than that of scum-free water. The surface tension of the algal scum decreased by 12.5%, and the SAR microwave backscatter was reduced by 7.3 dB. This indicated that the cyanobacterial scum could effectively attenuate capillary waves and appear as dark patches in SAR images. SAR has the potential to be developed as a tool for the remote sensing of algal scum in lake waters.

Previous studies have shown that Terra moderate resolution imaging spectroradiometer (MODIS) has low detection and characterization efficiency when mapping a green tide (Ulva prolifera) in the Yellow Sea. To quantify the uncertainty in mapping of the green tide using MODIS data, comparisons were conducted between quasi synchronous MODIS images and in situ observation data, as well as an unmanned aerial vehicle (UAV) image. The results show that MODIS images could detect the location of large (<100  m) floating green algae patches with good positional accuracy but tended to ignore the existence of small patches less than 10 m in width. The floating macroalgae area extracted using MODIS was several times larger than the area mapped using the UAV image. The Sentinel-2 multispectral instrument, the Chinese high-resolution GF-1 wide field camera, and the Chinese HJ-1 charge-coupled device are recommended for early green tide detection, whereas MODIS is suitable for green tide monitoring. The UAV could also play an important role in regional green tide monitoring with the advantages of flexibility, smaller dimensions, high spatial resolution, and low cost.

The Doppler spectra of first-order sea echo are studied in a high frequency radar system for hybrid sky-surface wave propagation when the receiving antenna array is mounted on a shipborne platform. Here, the platform motion and ionosphere perturbed phases are assumed to be the key factors resulting in the sea clutter spectral fluctuation. Based on the received first-order electric field equation, a closed form of cross sections is derived to account for the ship motion influence. It is shown numerically that the resultant clutter spectra exhibit significant discrepancies with different system operating parameters and platform motion states. In addition, the phase path variations along the inhomogeneous ionosphere are considered by dividing the ionosphere into the quiescent and distorted components. Such manipulation makes it feasible to simulate the phase fluctuations with a method associated with deterministic and stochastic models. Eventually, simulation results for a composite modulation taking account of both effects are provided. Compared with the spread Doppler spectra due to the platform motion, the attached ionosphere perturbed phases will further split and leak the spectral energy, which may bring in difficulties for the applications in remote sensing and target detection.

A system for measuring the two-dimensional (2-D) spatial distribution of atmospheric CO2 over complex industrial sites and urban areas on the order of 1 to 30  km2 every few minutes with a spatial resolution as high as tens of meters has been developed and demonstrated over the past 3 years. The greenhouse gas (GHG) laser imaging tomography experiment (GreenLITE™) provides improved measurement capabilities for applications ranging from automated 24/7 monitoring of ground carbon storage/sequestration (GCS) sites to long-duration real-time analyses of GHG sources and sinks in urban environments. GreenLITE combines a set of sensors based on an intensity modulated continuous wave approach with 2-D sparse tomographic reconstruction mechanisms to compute a 2-D map of CO2 concentrations over the area of interest. GreenLITE systems have recently been deployed at a number of test facilities, including a 4000-h demonstration at a GCS site in Illinois and an urban deployment in Paris, France, from November 2015 to the present. This paper describes the GreenLITE concept and the associated measurement capabilities and provides proof of concept results and analyses of observations from both short-term tests as well as longer-term industrial and urban deployments.

In general, considerable human and material resources are required for performing a forest inventory survey. Using remote sensing technologies to save forest inventory costs has thus become an important topic in forest inventory-related studies. Leica ADS-40 digital aerial photographs feature advantages such as high spatial resolution, high radiometric resolution, and a wealth of spectral information. As a result, they have been widely used to perform forest inventories. We classified ADS-40 digital aerial photographs according to the complex forest land cover types listed in the Fourth Forest Resource Survey in an effort to establish a classification method for categorizing ADS-40 digital aerial photographs. Subsequently, we classified the images using the knowledge-based classification method in combination with object-based analysis techniques, decision tree classification techniques, classification parameters such as object texture, shape, and spectral characteristics, a class-based classification method, and geographic information system mapping information. Finally, the results were compared with manually interpreted aerial photographs. Images were classified using a hierarchical classification method comprised of four classification levels (levels 1 to 4). The classification overall accuracy (OA) of levels 1 to 4 is within a range of 64.29% to 98.50%. The final result comparisons showed that the proposed classification method achieved an OA of 78.20% and a kappa coefficient of 0.7597. On the basis of the image classification results, classification errors occurred mostly in images of sunlit crowns because the image values for individual trees varied. Such a variance was caused by the crown structure and the incident angle of the sun. These errors lowered image classification accuracy and warrant further studies. This study corroborates the high feasibility for mapping complex forest land cover types using ADS-40 digital aerial photographs.

A hierarchical method is proposed for synthetic aperture radar (SAR) and optical image coregistration. We use local invariant salient points extracted by the binary robust invariant scalable keypoints (BRISK) algorithm for SAR and optical image coregistration. However, the matched points are highly erroneous because of the speckle noise in SAR images and the different structures of SAR and optical images. Therefore, an adaptive and elliptical bilateral filter is used to remove the speckle noise. Additionally, a hierarchical approach is used for coregistration using the local and global geometrical relationship of BRISK features. For each salient point in the optical image, three closest matched points are found in the SAR image. The geometrical relationship of the matched points is determined in the local areas around the salient and matched points, and matched pairs with fewer geometrical matching scores are removed. At the final stage of the algorithm, the projective global model between optical and SAR images is obtained using a robust statistic and the remaining false matches are refined. Experimental results and the comparison of the results of the proposed algorithm with those of the existing approaches show that the proposed algorithm is more efficient.

A quantized block compressive sensing (QBCS) framework, which incorporates the universal measurement, quantization/inverse quantization, entropy coder/decoder, and iterative projected Landweber reconstruction, is summarized. Under the QBCS framework, this paper presents an improved reconstruction algorithm for aerial imagery, QBCS, with entropy-aware projected Landweber (QBCS-EPL), which leverages the full-image sparse transform without Wiener filter and an entropy-aware thresholding model for wavelet-domain image denoising. Through analyzing the functional relation between the soft-thresholding factors and entropy-based bitrates for different quantization methods, the proposed model can effectively remove wavelet-domain noise of bivariate shrinkage and achieve better image reconstruction quality. For the overall performance of QBCS reconstruction, experimental results demonstrate that the proposed QBCS-EPL algorithm significantly outperforms several existing algorithms. With the experiment-driven methodology, the QBCS-EPL algorithm can obtain better reconstruction quality at a relatively moderate computational cost, which makes it more desirable for aerial imagery applications.

We investigate optimal waveform design using fractional Fourier transform in signal-dependent interference, as well as additive channel noise for stochastic extended target. Within constraints on waveform energy and bandwidth, optimal waveform design in fractional Fourier domain based on the signal-to-interference-plus-noise ratio criterion, probability of detection criterion, and mutual information criterion are modeled, respectively. In addition, the relationship between the optimal waveforms that are designed based on the three criteria is discussed. Simulations are conducted to illustrate that for all of the three criteria, the energy of optimal waveform can be distributed in some narrow bands where the target power is large and the interference power is small in fractional Fourier domain. Finally, the fractional Fourier domain waveform design method is proved more flexible and effective than traditional Fourier domain waveform design method, especially when the spectral density of target response and interference are relatively dispersed and flat.

Automatic cloud masking of Earth observation images is one of the first required steps in optical remote sensing data processing since the operational use and product generation from satellite image time series might be hampered by undetected clouds. The high temporal revisit of current and forthcoming missions and the scarcity of labeled data force us to cast cloud screening as an unsupervised change detection problem in the temporal domain. We introduce a cloud screening method based on detecting abrupt changes along the time dimension. The main assumption is that image time series follow smooth variations over land (background) and abrupt changes will be mainly due to the presence of clouds. The method estimates the background surface changes using the information in the time series. In particular, we propose linear and nonlinear least squares regression algorithms that minimize both the prediction and the estimation error simultaneously. Then, significant differences in the image of interest with respect to the estimated background are identified as clouds. The use of kernel methods allows the generalization of the algorithm to account for higher-order (nonlinear) feature relations. After the proposed cloud masking and cloud removal, cloud-free time series at high spatial resolution can be used to obtain a better monitoring of land cover dynamics and to generate more elaborated products. The method is tested in a dataset with 5-day revisit time series from SPOT-4 at high resolution and with Landsat-8 time series. Experimental results show that the proposed method yields more accurate cloud masks when confronted with state-of-the-art approaches typically used in operational settings. In addition, the algorithm has been implemented in the Google Earth Engine platform, which allows us to access the full Landsat-8 catalog and work in a parallel distributed platform to extend its applicability to a global planetary scale.

In recent years, oil spill surveillance with space-borne synthetic aperture radar (SAR) has received unprecedented attention and has been gradually developed into a common technique for maritime environment protection. A typical SAR-based oil spill detection process consists of three steps: (1) dark-spot segmentation, (2) feature extraction, and (3) oil spill and look-alike discrimination. As a preliminary task in the oil spill detection process chain, dark-spot segmentation is a critical and fundamental step prior to feature extraction and classification, since its output has a direct impact on the two subsequent stages. The balance between the detection probability and false alarm probability has a vital impact on the performance of the entire detection system. Unfortunately, this problem has not drawn as much attention as the other two stages. A specific effort has been placed on dark-spot segmentation in single-pol SAR imagery. A combination of fine designed features, including gray features, geometric features, and textural features, is proposed to characterize the oil spill and seawater for improving the performance of dark-spot segmentation. In the proposed process chain, a histogram stretching transform is incorporated before the gray feature extraction to enhance the contrast between possible oil spills and water. A simple but effective multiple-level thresholding algorithm is developed to conduct a binary classification before the geometric feature extraction to obtain more accurate area features. A local binary pattern code is computed and assigned as the textural feature for a pixel to characterize the physical difference between oil spills and water. The experimental result confirms that the proposed fine designed feature combination outperforms existing approaches in both aspects of overall segmentation accuracy and the capability to balance detection probability and false alarm probability. It is a promising alternative that can be incorporated into existing oil spill detection systems to further improve system performance.

In this paper, we proposed an azimuth-range decouple-based L₁ regularization method for wide ScanSAR imaging via extended chirp scaling (ECS) and applied it to the TerraSAR-X data to achieve large-scale sparse reconstruction. Compared with ECS, the conventional ScanSAR imaging algorithm based on matched filtering, the proposed method can improve the synthetic aperture radar image performance with full-sampling raw data for not only sparse but also nonsparse surveillance regions. It can also achieve high-resolution imaging for sparse considered scenes efficiently using down-sampling raw data. Compared with a typical L₁ regularization imaging approach, which requires transfer of the two-dimensional (2-D) echo data into a vector and reconstruction of the scene via 2-D matrix operation, our proposed method has less computational cost and hence makes the large-scale regularization reconstruction of considered area become possible. The experimental results via real data validate the effectiveness of the proposed method.

Regularization technology is an efficient and robust method to implement forward-looking imaging of scanning radar, and the selection of regularization parameter is quite important for the imaging performance. When the imaging scene contains sparse strong targets as well as a weak background, different regularization parameters should be adopted. A dual-channel fast iterative shrinkage-thresholding (D-FIST) regularization algorithm is proposed. First, by setting a threshold with mean segmentation method, the strong targets are extracted from the echo to formulate the data of channel I, and the entire data are included in channel II. The L-curve method is adopted to determine the regularization parameter for two channels data, respectively. Then, the data are processed by the FIST algorithm with corresponding parameters. At last, the imaging result is obtained by combining the results of two channels. Simulation experiments show that the imaging results of D-FIST are better than that of direct method, especially in a low SNR situation.

For synthetic aperture radar (SAR) imaging, a stepped-frequency chirp signal is widely used to obtain ultrahigh range resolution. There are mainly two challenges in the stepped-frequency SAR imaging. One is the subbands synthesis, and the other is the estimation and compensation of amplitude and phase errors of the subbands caused by the radar system and the propagation. An imaging approach for the airborne squinted sliding-spotlight SAR with stepped-frequency chirps is proposed to achieve ultrahigh two-dimensional (2-D) resolution. Due to the squint angle, a Doppler centroid shift is introduced and then a modified space-variant compensation method for the range difference between the subbands is realized via multiplying a phase ramp in the 2-D frequency domain. Moreover, a range-dependent phase gradient autofocus algorithm based on raw data is proposed to estimate and compensate the residual phase error after the internal calibration. Afterward, the frequency-domain synthetic bandwidth method is used to complete the bandwidth synthesis. Finally, the extended Omega-K algorithm is applied to focus the synthesized data. The simulation results and the real data processing results prove the validity of the proposed approach.

We propose a hyperspectral image (HSI) classification method combining low-dimensional spectral–spatial features with joint sparsity model (JSM). First, for the high-dimensional data sets, we introduced image fusion for feature reduction. Then fast bilateral filtering is adopted to exploit spatial features, which will be combined with the original spectral features for classification. Based on the low-dimensional spectral–spatial features, we utilize JSM to serve as a classifier. Considering the strong relationship between the neighboring pixels in HSI, this model can achieve a promising performance by exploiting regional spectral–spatial information. Overall accuracies (with 10% and 2% training samples) of the proposed method are 97.84% and 97.52% for the Indian Pines image and University of Pavia image. Experimental results on different HSI data sets show that the proposed method shows outstanding performance in terms of classification accuracy and computational efficiency.

A fusion algorithm based on target extraction for infrared image (IIR) and visible image fusion in the nonsubsampled contourlet transform (NSCT) domain is proposed. Commonly, the target information in IIR is important; in order to fully retain the target information in a final fused image, first, use maximum between-class variance method to segment IIR, such that the target regions with salient objects are extracted to produce the background and target images. Next, the visible and background images are decomposed to a series of low-pass and band-pass images by NSCT, respectively. Then, fuse the obtained images to produce the fused background image by different strategies in each band, in which Gaussian fuzzy logic is used to produce the low-pass coefficient; the spatial frequency of each band-pass image is used to determine the linking strength β value of pulse coupled neural network structure, and the result is used to fuse the band-pass images. Eventually, the fused image is produced combining the target image and the fused background image. The experiments show that this algorithm can retain more background details of the two images and highlight the target in the infrared image more effectively, as well as obviously improve the visual effect of the fusion image.

This paper presents an investigation into the underlying geometry and performance of various statistical target detection algorithms for hyperspectral imagery, presents results from algorithm testing, and investigates general trends and observable principles for understanding performance. Over the variety of detection algorithms, there is no universally best performing algorithm. In our test, often top performing algorithms on one class of targets obtain mediocre results on another class of targets. However, there are two clear trends: quadratic detectors such as ACE generally performed better than linear ones especially for subpixel targets (our top 15 scoring algorithms were quadratic detectors), and using anomaly detection to prescreen image spectra improved the performance of the quadratic detectors (8 of our top 9 scoring algorithms using anomaly prescreening). We also demonstrate that simple combinations of detection algorithms can outperform single algorithms in practice. In our derivation of detection algorithms, we provide exposition on the underlying mathematical geometry of the algorithms. That geometry is then used to investigate differences in algorithm performance. Tests are conducted using imagery and targets freely available online. The imagery was acquired over Cooke City, Montana, a small town near Yellowstone National Park, using the HyMap V/NIR/SWIR sensor with 126 spectral bands. There are three vehicle and four fabric targets located in the town and surrounding area.

We have developed a method for using hyperspectral (HS) data to identify and locate chemical materials on arbitrary surfaces using the materials’ reflection or emission spectra that makes no prior modeling assumptions about the presence of pure pixels or the statistics of the background clutter and sensor noise. To our knowledge, this is the first time that surface detection without dependence on background information has been achieved. There are three main components to the method: (1) an HS unmixing algorithm based on the alternating direction method of multipliers that is applied over local subsets of the imaging to resolve the HS data into a set of linearly independent spectral and spatial components; (2) the fitting of those unmixing spectra to a set of candidate template spectra; and (3) a support vector machine classifier for chemical detection, identification, and location. The algorithm is illustrated on HS data collected by a Telops Hyper-Cam infrared camera on data resulting from the deposition of chemical agent simulants on various surfaces.

The near-range clutter of airborne nonside-looking array greatly depends on range. Conventional phased-array space-time adaptive processing (STAP) radar suffers severe performance degradation in the presence of a near-range clutter scenario. To efficiently suppress no-stationary clutter with only one snapshot, an STAP algorithm for airborne multiple-input multiple-output (MIMO) radar with nonside-looking array based on sparse representation is first presented, which is referred to as MIMOSR-STAP in this paper. By exploiting the waveform diversity of MIMO radar, each snapshot of a tested range-cell is transformed into the multisnapshots of phased array radar, which are used to estimate the high-resolution space-time spectrum with multiple measurement vectors technique. The proposed approach is effective in estimating the spectrum by utilizing temporally correlated multiple sparse Bayesian learning. In the sequel, the clutter covariance matrix and corresponding adaptive weight vector are efficiently obtained. MIMOSR-STAP enjoys high accuracy and robustness so that it achieves better performance of output signal-to-clutter-plus-noise-ratio and minimum detectable velocity than the single measurement vector sparse representation methods in the literature. Thus, MIMOSR-STAP performs well in a serious nonstationary clutter scenario and is suitable for an insufficient independent and identically distributed samples environment.

This paper proposes an absolute attitude measurement approach by utilizing a monostatic wideband radar. In this approach, the three-dimensional electromagnetic-model (3-D em-model) and the parametric motion model of a target are combined to estimate absolute attitude. The 3-D em-model is established offline based on the target’s geometric structure. Scattering characteristics such as radar cross section and radar images from one-dimension to 3-D can be conveniently predicted by this model. By matching the high-resolution range profiles (HRRPs) of measurements with the HRRPs predicted by the 3-D em-model, the directions of the lines of sight relative to the target at different measuring times are first obtained. Then, based on the obtained directions and the parametric motion model of the target, the target absolute attitude at each measuring time can be acquired. Experiments using both data predicted by a high-frequency em-code and data measured in an anechoic chamber verify the validity of the proposed method.

Remote sensing technology plays a vital role in the initial stages of ore deposits exploration, with special significance in arid and semiarid regions. Advanced spaceborne thermal emission and reflection radiometer (ASTER) creates opportunities for geologists and many researchers to study the extraction of hydrothermal alterations of the porphyry copper and epithermal gold deposits. ASTER data were used to map distribution of hydrothermal alteration of deposits in the Baogutu region which is located in western Junggar, Xinjiang, China, including Baogutu II and Baogutu V deposits. Band ratio logical operator (BRLO), principal component analysis (PCA), mixture-tuned matched-filtering (MTMF), and constrained energy minimization (CEM) techniques were used for detailed hydrothermal alteration mapping. The alteration zones were identified using BRLO and PCA. The minerals subject to alteration were extracted by MTMF and CEM. The results were validated through field observation, spectral measurements, and petrographic studies. By comparing the extraction results and accuracy rates of four methods, CEM boasts the highest accuracy and identifies the altered minerals corresponding to three types of alterations at accuracy of over 80%. The identifying result of CEM method indicates that the prospect areas that were located in the southeast part of the Baogutu II rocks and in the southwest part of the Baogutu V rocks are valuable for further exploration.

The goal of ratio enhancement for hyperspectral (HS) image pansharpening is to obtain an enhancement ratio between a simulated low-resolution panchromatic (Pan) image and an original high-resolution Pan image. However, the simulated low-resolution Pan image often suffers from gray-level distortion. To solve these problems, the original HS bands are synthesized to a smaller number of reduced HS bands, then the pixels of Pan and HS images are divided into different groups according to the linearity between the Pan band and the reduced bands. For each pixel group, a nonnegative least-squares algorithm is utilized to calculate the weights of reduced HS bands, so that the simulated Pan image is obtained by weighted summation of reduced HS bands. Finally, the HS image is sharpened by a ratio enhancement. The experiments demonstrated that the proposed method had a good performance on fusion quality.

The smooth variable structure filter (SVSF) has seen significant development and research activity in recent years. It is based on sliding mode concepts, which utilize a switching gain that brings an inherent amount of stability to the estimation process. In an effort to improve upon the numerical stability of the SVSF, a square-root formulation is derived. The square-root SVSF is based on Potter’s algorithm. The proposed formulation is computationally more efficient and reduces the risks of failure due to numerical instability. The new strategy is applied on target tracking scenarios for the purposes of state estimation, and the results are compared with the popular Kalman filter. In addition, the SVSF is reformulated to present a two-pass smoother based on the SVSF gain. The proposed method is applied on an aerospace flight surface actuator, and the results are compared with the Kalman-based two-pass smoother.

Edge extraction from high spatial resolution (HSR) remotely sensed images is one of the essential tasks for image segmentation and object identification. We present an optimal Gabor-based edge detection method which mainly focuses on selecting optimal parameters, including central frequency and spectrum scale, for Gabor filter. The central frequency is automatically optimized by phase randomization and the human visual system-based structure similarity index. Next, the optimal spectrum scale is determined based on two-dimensional power spectrum density. The edge detection method is comprehensively discussed in the analysis of parameter sensitivity, overall performance, and comparative tests with several widely used methods. Qualitative and quantitative experimental studies, performed on six test images with various spatial resolution, show that the proposed method provides a promising solution to edge detection from HSR remotely sensed images.

A graph-based segmentation technique has been tailored to segment airborne LiDAR points which, unlike images, are irregularly distributed. In our method, every LiDAR point is labeled as a node and interconnected as a graph extended to its neighborhood, defined in a 4-D feature space: the spatial coordinates (x,y,z) and the reflection intensity. The interconnections between pairs of neighboring nodes are weighted based on the distance in the feature space. The segmentation consists of an iterative process of classification of nodes into homogeneous groups based on their similarity. This approach is intended to be part of a complete system for the classification of structures from LiDAR point clouds in applications needing fast response times. In this sense, a study of the performance/accuracy trade-off has been performed, extracting some conclusions about the benefits of the proposed solution. In addition, an interlaced graph-based approach is proposed to increase the reliability in general purpose segmentations.

Cloud removal is a very important preprocessing step when using aerial and spaceborne optical sensors for land surface and cover applications. Methods that have been proposed for identifying cloud-affected pixels range from classification and segmentation type approaches applied to individual images to outlier detection type methods applied to time-series of images. The choice of method is influenced by considerations including the requirements of the application, the image characteristics, and how frequently images over a given area are acquired. When many images are acquired in a period where land surface cover exhibits negligible change, an image formed by compositing from a series of images taken in a relatively short period of time will suffice for further analysis. It is highly desirable to fully automate this compositing process. To this end, we propose the multiple temporal mosaicing (MTM) algorithm. It uses, in the first instance, a cloud score for each pixel in the images to separate/partially separate cloud-affected pixels from noncloud pixels. These cloud scores are then combined with the output from existing cloud identification methods and date preference to determine the likelihood of given pixels being considered as good candidates to be included in the final image. Moreover, the spatial smoothness is incorporated to ensure that the pixels of a small neighborhood are from the same image so that the final image looks smoother. We apply MTM to two Landsat scenes. The resulting images show the effectiveness of this method. The methodology can be applied to images acquired from other sensors.

Hyperspectral imagers collect hundreds of images corresponding to different wavelength channels for the same area on the surface of the earth. Since a lot of information is redundant in neighboring bands and pixels, reducing dimensionality is necessary. Due to the high resolution in the spatial–spectral domain and complex computation, this procedure is time-consuming. Advances in leveraging special hardware, such as GPUs and MICs, show new ways of accelerating dimensionality reduction. We propose a parallel framework of fast independent component analysis (FastICA) for hyperspectral image (HSI) dimensionality reduction on heterogeneous platforms, which offers six parallel implementations on different parallel platforms. We also use specialized optimizations for each hotspot of FastICA and implement them in our framework running on a local GPU server and the Tianhe-2 supercomputer. The experimental results show that all these parallel implementations in the framework can obtain excellent performances and good scalabilities: the speed-up is up to 169 times on the GPU server and 410 times using 64 nodes on the Tianhe-2 supercomputer compared to the sequential implementation. We conclude that by using our framework, HSI dimensionality reduction can be implemented in real-time on the heterogeneous platform.

In inverse synthetic aperture radar (ISAR) imaging, the migration through resolution cells (MTRCs) will occur when the rotation angle of the moving target is large, thereby degrading image resolution. To solve this problem, an ISAR imaging method based on segmented preprocessing is proposed. In this method, the echoes of large rotating target are divided into several small segments, and every segment can generate a low-resolution image without MTRCs. Then, each low-resolution image is rotated back to the original position. After image registration and phase compensation, a high-resolution image can be obtained. Simulation and real experiments show that the proposed algorithm can deal with the radar system with different range and cross-range resolutions and significantly compensate the MTRCs.

The problem of adaptive waveform design for target detection in cognitive radar (CR) is investigated. This problem is analyzed in signal-dependent interference, as well as additive channel noise for extended target with unknown target impulse response (TIR). In order to estimate the TIR accurately, the Kalman filter is used in target tracking. In each Kalman filtering iteration, a flexible online waveform spectrum optimization design taking both detection and range resolution into account is modeled in Fourier domain. Unlike existing CR waveform, the proposed waveform can be simultaneously updated according to the environment information fed back by receiver and radar performance demands. Moreover, the influence of waveform spectral phase to radar performance is analyzed. Simulation results demonstrate that CR with the proposed waveform performs better than a traditional radar system with a fixed waveform and offers more flexibility and suitability. In addition, waveform spectral phase will not influence tracking, detection, and range resolution performance but will greatly influence waveform forming speed and peak-to-average power ratio.

Detecting surface coal mining areas (SCMAs) using remote sensing data in a timely and an accurate manner is necessary for coal industry management and environmental assessment. We developed an approach to effectively extract SCMAs from remote sensing imagery based on object-oriented decision trees (OODT). This OODT approach involves three main steps: object-oriented segmentation, calculation of spectral characteristics, and extraction of SCMAs. The advantage of this approach lies in its effective integration of the spectral and spatial characteristics of SCMAs so as to distinguish the mining areas (i.e., the extracting areas, stripped areas, and dumping areas) from other areas that exhibit similar spectral features (e.g., bare soils and built-up areas). We implemented this method to extract SCMAs in the eastern part of Ordos City in Inner Mongolia, China. Our results had an overall accuracy of 97.07% and a kappa coefficient of 0.80. As compared with three other spectral information-based methods, our OODT approach is more accurate in quantifying the amount and spatial pattern of SCMAs in dryland regions.

Accurate and frequent monitoring of land surface changes arising from oil and gas exploration and extraction is a key requirement for the responsible and sustainable development of these resources. Petroleum deposits typically extend over large geographic regions but much of the infrastructure required for oil and gas recovery takes the form of numerous small-scale features (e.g., well sites, access roads, etc.) scattered over the landscape. Increasing exploitation of oil and gas deposits will increase the presence of these disturbances in heavily populated regions. An object-based approach is proposed to utilize RapidEye satellite imagery to delineate well sites and related access roads in diverse complex landscapes, where land surface changes also arise from other human activities, such as forest logging and agriculture. A simplified object-based change vector approach, adaptable to operational use, is introduced to identify the disturbances on land based on red–green spectral response and spatial attributes of candidate object size and proximity to roads. Testing of the techniques has been undertaken with RapidEye multitemporal imagery in two test sites located at Alberta, Canada: one was a predominant natural forest landscape and the other landscape dominated by intensive agricultural activities. Accuracies of 84% and 73%, respectively, have been achieved for the identification of well site and access road infrastructure of the two sites based on fully automated processing. Limited manual relabeling of selected image segments can improve these accuracies to 95%.

Multichannel synthetic aperture radar systems are usually employed to suppress azimuth ambiguity and realize high-resolution wide-swath imaging. However, unavoidable array errors will significantly degrade the performance of ambiguity suppression and imaging quality. This paper presents an array error estimation method based on cross correlation. First, unambiguous Doppler spectra are obtained by selecting a short length of range profiles from strong targets. Then, array errors are estimated by a proposed cross-correlation method. Finally, a preprocessing method to improve the estimation accuracy is proposed. The proposed method takes full advantage of the training samples from strong targets and estimates array errors by the coherent integration technique, which improves the estimation accuracy and robustness. Theoretical analysis and experiments based on simulations and measurements showed the validity of the proposed method, especially in low signal-to-noise ratios.

This study addresses the problems associated with high dimensionality of hyperspectral images in reference to clustering. A new local band selection approach that takes both relevancy and redundancy among the bands into account while obtaining the multiple relevant set of bands is developed. The local band selection approach is then incorporated within a multistage clustering framework that includes three stages: segmentation, region merging, and projected clustering. At first, k-means is used to produce initial segments/regions. Then, in the region merging stage, the modified local mutually best region merging strategy is applied on the initial segments to produce the refined segmentation map. Finally, an improved projected clustering technique is used to group these segments into a fixed number of clusters. Further, the main principle of projected clustering, that different sets of point may cluster better for different subsets of dimensions, is extended to region merging by incorporating the suggested local band selection approach. The framework requires input for two major parameters, which are number of clusters (k) and number of relevant bands (l). The framework is tested over two hyperspectral images and compared with other clustering frameworks. The experimental results confirm the effectiveness of the proposed framework.

Forest height plays a crucial role in investigating the biophysical parameters of forests and the terrestrial carbon. PolInSAR-based inversion modeling has been successfully implemented on airborne and spaceborne synthetic aperture radar (SAR) data. SAR tomography is a recent approach to separate scatterers in the cross-range direction and generate its vertical profile. This study highlights the potential of tomographic processing of multibaseline fully polarimetric Radarsat-2 C-band SAR data to estimate radar reflectivity at different forest height levels. A teak patch of Haldwani forest in Uttarakhand state of India was chosen as the test site to perform tomography. Since SAR tomography is a spectral estimation problem, Fourier transform (FT), beamforming (BF), and Capon-based spectral estimators were applied on the dataset to obtain the backscattering power contributions at different forest height levels. Fourier showed high backscatter power retrieval at different forest heights. The radar reflectivities at different heights were significantly reduced by BF followed by Capon. Tomographic profile of FT severely suffered from high sidelobes, which was drastically reduced by implementing BF. Capon further reduced the sidelobes and achieved a substantially improved tomographic profile. The height maps were generated for these algorithms and validated with ground truth data.

A multitemporal algorithm, originally conceived for the C-band radar aboard the Sentinel-1 satellite, has been updated to retrieve soil moisture from L-band radar data, such as those provided by the National Aeronautics and Space Administration Soil Moisture Active/Passive (SMAP) mission. This type of algorithm may deliver more accurate soil moisture maps that mitigate the effect of roughness and vegetation changes. Within the multitemporal inversion scheme based on the Bayesian maximum a posteriori probability (MAP) criterion, a dense time series of radar measurements is integrated to invert a forward backscattering model. The model calibration and validation tasks have been accomplished using the data collected during the SMAP validation experiment 12 spanning several soil conditions (pasture, wheat, corn, and soybean). The data have been used to update the forward model for bare soil scattering at L-band and to tune a simple vegetation scattering model considering two different classes of vegetation: those producing mainly single scattering effects and those characterized by a significant multiple scattering involving terrain surface and vegetation elements interaction. The algorithm retrievals showed a root mean square difference (RMSD) around 5% over bare soil, soybean, and cornfields. As for wheat, a bias was observed; when removed, the RMSD went down from 7.7% to 5%.

Asphalt road reflectance spectra change as pavement ages. This provides the possibility for remote sensing to be used to monitor a change in asphalt pavement conditions. However, the relatively narrow geometry of roads and the relatively coarse spatial resolution of remotely sensed imagery result in mixtures between pavement and adjacent landcovers (e.g., vegetation, buildings, and soil), increasing uncertainties in spectral analysis. To overcome this problem, multiple endmember spectral mixture analysis (MESMA) was used to map the asphalt pavement condition using Worldview-2 satellite imagery in this study. Based on extensive field investigation and in situ measurements, aged asphalt pavements were categorized into four stages—preliminarily aged, moderately aged, heavily aged, and distressed. The spectral characteristics in the first three stages were further analyzed, and a MESMA unmixing analysis was conducted to map these three kinds of pavement conditions from the Worldview-2 image. The results showed that the road pavement conditions could be detected well and mapped with an overall accuracy of 81.71% and Kappa coefficient of 0.77. Finally, a quantitative assessment of the pavement conditions for each road segment in this study area was conducted to inform road maintenance management.

The applicability of the normalized difference water index (NDWI) to the delineation of dam failure-induced floods is demonstrated for the case of the Sparmos dam (Larissa, Central Greece). The approach followed was based on the differentiation of NDWI maps to accurately define the extent of the inundated area over different time spans using multimission Earth observation optical data. Besides using Landsat data, for which the index was initially designed, higher spatial resolution data from Sentinel-2 mission were also successfully exploited. A geospatial analysis approach was then introduced to rapidly identify potentially affected segments of the road network. This allowed for further correlation to actual damages in the following damage assessment and remediation activities. The proposed combination of geographic information systems and remote sensing techniques can be easily implemented by local authorities and civil protection agencies for mapping and monitoring flood events.

Linear mixtures of materials in a scene often occur because the resolution of a sensor is relatively coarse, resulting in pixels containing patches of different materials within them. This phenomenon causes nonoverlapping areal mixing and can be modeled by a linear mixture model. More complex phenomena, such as the multiple scattering in mixtures of vegetation, soils, granular, and microscopic materials within pixels can result in intimate mixing with varying degrees of nonlinear behavior. In such cases, a linear model is not sufficient. This study considers two approaches for unmixing pixels in a scene that may contain linear or intimate (nonlinear) mixtures. The first method is based on earlier studies that indicate nonlinear mixtures in reflectance space are approximately linear in albedo space. The method converts reflectance to single-scattering albedo according to Hapke theory and uses a constrained linear model on the computed albedo values. The second method is motivated by the same idea, but uses a kernel that seeks to capture the linear behavior of albedo in nonlinear mixtures of materials. This study compares the two approaches and pays particular attention to these dependencies. Both laboratory and airborne collections of hyperspectral imagery are used to validate the methods.

In agricultural remote sensing, unmanned aerial vehicle (UAV) platforms offer many advantages over conventional satellite and full-scale airborne platforms. One of the most important advantages is their ability to capture high spatial resolution images (1–10 cm) on-demand and at different viewing angles. However, UAV platforms typically rely on the use of multiple cameras, which can be costly and difficult to operate. We present the development of a simple low-cost imaging system for remote sensing of crop health and demonstrate it on lettuce (Lactuca sativa) grown in Hong Kong. To identify the optimal vegetation index, we recorded images of both healthy and unhealthy lettuce, and used them as input in an expectation maximization cluster analysis with a Gaussian mixture model. Results from unsupervised and supervised clustering show that, among four widely used vegetation indices, the blue wide-dynamic range vegetation index is the most accurate. This study shows that it is readily possible to design and build a remote sensing system capable of determining the health status of lettuce at a reasonably low cost (<US$100). When combined with recent advances in UAV technology, this system could lead to yield increases for lettuce growers.

Interpretation of synthetic aperture radar (SAR) data processing is difficult because the geometry and spectral range of SAR are different from optical imagery. Consequently, SAR imaging can be a complementary data to multispectral (MS) optical remote sensing techniques because it does not depend on solar illumination and weather conditions. This study presents a multisensor fusion of SAR and MS data based on the use of classification and regression tree (CART) and support vector machine (SVM) through a decision fusion system. First, different feature extraction strategies were applied on SAR and MS data to produce more spectral and textural information. To overcome the redundancy and correlation between features, an intrinsic dimension estimation method based on noise-whitened Harsanyi, Farrand, and Chang determines the proper dimension of the features. Then, principal component analysis and independent component analysis were utilized on stacked feature space of two data. Afterward, SVM and CART classified each reduced feature space. Finally, a fusion strategy was utilized to fuse the classification results. To show the effectiveness of the proposed methodology, single classification on each data was compared to the obtained results. A coregistered Radarsat-2 and WorldView-2 data set from San Francisco, USA, was available to examine the effectiveness of the proposed method. The results show that combinations of SAR data with optical sensor based on the proposed methodology improve the classification results for most of the classes. The proposed fusion method provided approximately 93.24% and 95.44% for two different areas of the data.

Early identification of cotton fields is important for advancing boll weevil eradication progress and reducing the risk of reinfestation. Remote sensing has long been used for crop identification, but limited work has been reported on early identification of cotton fields. The objective of this study is to evaluate aerial imagery for identifying cotton fields before cotton plants start to bloom. A two-camera imaging system was used to acquire red-green-blue and near-infrared images with 1-m pixel resolution along two flight lines over an 8  km×12  km cropping area. The images were mosaicked using two approaches: manual georeferencing followed by position-based mosaicking in Erdas Imagine and content-based automatic mosaicking in Pix4DMapper. The mosaicked images were then classified into different crops and cover types using supervised classification techniques. Results showed that both types of mosaics were effective for cotton identification and that maximum likelihood classification produced the best overall accuracy of 90% for the position-based approach and 91% for the content-based approach. The methodologies presented in this study will be useful for boll weevil eradication program managers to quickly and efficiently identify cotton fields at relatively early growth stages using mosaicked aerial imagery.

This study investigated the associations between land cover changes and evapotranspiration (ET) in the Yellow River Delta during the last 30 years using Landsat imagery. The result showed that the Delta region experienced a distinct increase in area due to sea–land interaction and sediment deposition, accompanied by substantial change in land cover fractions. From 1986 to 2015, 35.48% of land cover changed, mainly due to a transformation into salterns and culture ponds from other land cover types. In general, land cover was converted from less developed into highly developed types. The monowindow algorithm for retrieval of land surface temperature (LST) and the SEBAL model were used to explore the effects of land cover changes on regional ET. The results indicated that the average relative error of daily ET was 9.46%, and there was a significant linear correlation (R2≥0.959, p<0.001) between ET and LST. Relationships existed between LST, ET, fractional vegetation cover, and other relevant vegetation indices, and there were positive and negative correlations between different threshold ranges. During the study period, the transformation of large areas of different land cover types into salterns and culture ponds led to an average increase of 1.43 mm in daily ET.

Autonomous off-road navigation and obstacle avoidance are central to applications of unmanned ground vehicles (UGV). This requires the sensors to have the ability to sense obstacles outdoors or in unstructured environments. Radar sensors, which have the ability to operate in all weather, all lighting, and foliage-covered conditions, have been widely used. We focus on the perception ability of forward-imaging interferometric synthetic aperture radar (InSAR), which can provide three-dimensional scene information of unstructured environments for UGV. An InSAR signal processing flow, similar to the techniques of stereo pair airborne InSAR, is tailored to acquiring the height information of the scene. The InSAR can output three different kinds of images: the scattering image, the coherence image, and the elevation image. Simulation and real data experiments have been carried out on a terrain scene where there are positive and negative obstacles to validate the proposed method and the perception performance of the InSAR, and the special feature differences of obstacles are prominent in the InSAR images.

With the rapid advance of remote sensing technology, the amount of three-dimensional point-cloud data has increased extraordinarily, requiring faster processing in the construction of digital elevation models. There have been several attempts to accelerate the computation using parallel methods; however, little attention has been given to investigating different approaches for selecting the most suited parallel programming model for a given computing environment. We present our findings and insights identified by implementing three popular high-performance parallel approaches (message passing interface, MapReduce, and GPGPU) on time demanding but accurate kriging interpolation. The performances of the approaches are compared by varying the size of the grid and input data. In our empirical experiment, we demonstrate the significant acceleration by all three approaches compared to a C-implemented sequential-processing method. In addition, we also discuss the pros and cons of each method in terms of usability, complexity infrastructure, and platform limitation to give readers a better understanding of utilizing those parallel approaches for gridding purposes.

Water level was estimated, using AltiKa radar altimeter onboard the SARAL satellite, over the Ukai reservoir using modified algorithms specifically for inland water bodies. The methodology was based on waveform classification, waveform retracking, and dedicated inland range corrections algorithms. The 40-Hz waveforms were classified based on linear discriminant analysis and Bayesian classifier. Waveforms were retracked using Brown, Ice-2, threshold, and offset center of gravity methods. Retracking algorithms were implemented on full waveform and subwaveforms (only one leading edge) for estimating the improvement in the retrieved range. European Centre for Medium-Range Weather Forecasts (ECMWF) operational, ECMWF re-analysis pressure fields, and global ionosphere maps were used to exactly estimate the range corrections. The microwave and optical images were used for estimating the extent of the water body and altimeter track location. Four global positioning system (GPS) field trips were conducted on same day as the SARAL pass using two dual frequency GPS. One GPS was mounted close to the dam in static mode and the other was used on a moving vehicle within the reservoir in Kinematic mode. In situ gauge dataset was provided by the Ukai dam authority for the time period January 1972 to March 2015. The altimeter retrieved water level results were then validated with the GPS survey and in situ gauge dataset. With good selection of virtual station (waveform classification, back scattering coefficient), Ice-2 retracker and subwaveform retracker both work better with an overall root-mean-square error <15  cm. The results support that the AltiKa dataset, due to a smaller foot-print and sharp trailing edge of the Ka-band waveform, can be utilized for more accurate water level information over inland water bodies.

Change detection is defined as the process that helps in determining the changes associated with land use and land cover properties with reference to georegistered multitemporal remote sensing data. A unique idea of integrating sparse fusion with constrained clustering has been attempted to detect changes in multitemporal and multispectral remote sensing images in a framework called ASCC (to be pronounced as “ask”). Unlike the other traditional methods of change detection, this approach focuses on the fusion of difference images followed by a semisupervised clustering to construct a change detection map in binary form. An enhanced sparse representation is adopted to improve the performance of the fusion process by creating a locally adaptive dictionary that extracts patches from the two difference images. The reconstruction of the fused image is performed using maximum absolute coefficients of the learned dictionary. The resultant fused image is subjected to the change detection process through constrained clustering to discern the changed pixels from the unchanged pixels. To evaluate the designed solution, the method is quantified through percentage correct classification and the process is compared with the existing methods of clustering such as k-means, adaptive k-means, fuzzy C-means, and enhanced constrained k-means. The simulation results demonstrated that the designed approach performed better when compared with state-of-art change detection methods for detecting changes in multitemporal images.

MODIS consists of a cross-track, two-sided scan mirror, whose reflectance is not uniform but is a function of angle of incidence (AOI). This feature, known as response versus scan-angle (RVS), was characterized for all reflective solar bands of both MODIS instruments prior to launch. The RVS characteristic has changed on orbit, which must be tracked precisely over time to ensure the quality of MODIS products. The MODIS characterization support team utilizes the onboard calibrators and the earth view responses from multiple pseudoinvariant desert sites to track the RVS changes at different AOIs. The drawback of using deserts is the assumption that these sites are radiometrically stable during the monitoring period. In addition, the 16-day orbit repeat cycle of MODIS allows for only a limited set of AOIs over a given desert. We propose a novel and robust approach of characterizing the MODIS RVS using tropical deep convective clouds (DCC). The method tracks the monthly DCC response at specified sets of AOIs to compute the temporal RVS changes. Initial results have shown that the Aqua-MODIS collection 6 band 1 level 1B radiances show considerable residual RVS dependencies, with long-term drifts up to 2.3% at certain AOIs.

A kHz optical remote sensing system is implemented to determine insect melanization features. This is done by measuring the backscatter signal in the visible and near-infrared (VIS-NIR) and short-wave infrared (SWIR) in situ. It is shown that backscatter cross section in the SWIR is insensitive to melanization and absolute melanization can be derived from the ratio of backscatter cross section of different bands (SWIR/VIS-NIR). We have shown that reflectance from insect is stronger in the SWIR as compared to NIR and VIS. This reveals that melanization plays a big role to determine backscatter cross section. One can use this feature as a tool to improve insect species and age classification. To support the findings, we illustrated melanization feature using three different insects [dead, dried specimens of snow white moth (Spilosoma genus), fox moth (Macrothylacia), and leather beetle (Odontotaenius genus)]. It is shown that reflectance from the leather beetle in the VIS and NIR is more affected by melanization as compared with snow white moth.

In the performance of change detection analysis, the change/unchanged pixel categorization is usually made through empirical methods or trial-and-error manual procedures whose reliability may affect the results. To detect land-cover changes, an unsupervised change detection technique was applied to Landsat images from an area in the south of México. At first, normalized surface reflectance, principal components, and tasselled cap images were used to develop the chi-square transformation (CST) applied to each kind of image organized in absolute and relative values and thus, obtain the continuous image of change. Later, the histogram secant technique was applied to change images to automatically define the thresholds and categorize as change/unchanged the pixels. Finally, to assess the change detection accuracy, 86 polygons (14,512 pixels) were sampled, classified as real change/unchanged sites, and defined as ground-truth, from the interpretation of color aerial photo slides aided by the land-cover maps to obtain omission/commission errors and kappa coefficient of agreement. The results show that the CST and automatic histogram secant thresholding are suitable techniques that can be applied for unsupervised analysis change detection.

We propose a local feature representation based on two types of linear filtering, feature pooling, and nonlinear divisive normalization for remote sensing image classification. First, images are decomposed using a bank of log-Gabor and Gaussian derivative filters to obtain filtering responses that are robust to changes in various lighting conditions. Second, the filtering responses computed using the same filter at nearby locations are pooled together to enhance position invariance and compact representation. Third, divisive normalization with channel-wise strategy, in which each pooled feature is divided by a common factor plus the sum of the neighboring features to reduce dependencies among nearby locations, is introduced to extract divisive normalization features (DNFs). Power-law transformation and principal component analysis are applied to make DNF significantly distinguishable, followed by feature fusion to enhance local description. Finally, feature encoding is used to aggregate DNFs into a global representation. Experiments on 21-class land use and 19-class satellite scene datasets demonstrate the effectiveness of the channel-wise divisive normalization compared with standard normalization across channels and the fusion of the two types of linear filtering in improving classification accuracy. The experiments also illustrate that the proposed method is competitive with state-of-the-art approaches.

The strong backscattering homogeneous region, i.e., a uniform region with a high and constant backscattering coefficient, is important for synthetic aperture radar (SAR) image quality assessment and SAR radiometric calibration, which, however, is difficult to realize in practice with a known backscattering coefficient. We realize a strong backscattering homogeneous region with a known backscattering coefficient in SAR images by utilizing designed metal grids. First, we propose a manmade grid-structure target and realize it with aluminum in practice, which is named the metal grid. Then, the backscattering coefficient of the designed metal grid is simulated in the computer simulation technology (CST) microwave studio and measured by a radar cross-section (RCS) measurement instrument in a microwave anechoic chamber. Both CST simulation results and RCS measurement results confirm the strong backscattering property of the designed target. In addition, by utilizing the designed target, we realize a test field consisting of several strong backscattering homogeneous regions with different sizes at Shanghai Jiao Tong University, Shanghai, China. The spaceborne experiments have been carried out by the TerraSAR-X sensor over the test field in two flight campaigns in X-band with VV polarization. Experimental results demonstrate the strong backscattering property and homogeneity of the realized regions.

The phytoplankton pigment indices were used to characterize the spatial succession of the community composition in the frontal regions of the subtropical front (STF), sub-Antarctic front (SAF), and polar front (PF) in the Indian Ocean sector of the Southern Ocean during austral summer 2013. Diagnostic indices revealed that the flagellates were dominant in STF (51%) and progressively declined toward SAF (39%) and PF (11%). Similarly, the prokaryotes were highest in STF (43%) and decreased to SAF (32%) and PF (28%). In contrast, the diatoms were gradually increased from STF (6%) to SAF (29%) and PF (61%). The variability of flagellates and diatoms from the STF to PF is attributed to the variability of photosynthetically available radiation, sea surface temperature, and sea surface wind speed. The in-situ pigment indices were then compared to the NASA Ocean Biogeochemical model that shows the similar patterns of frontal community distribution except their magnitude. Similarly, the satellite retrieved phytoplankton biomass (chlorophyll a) was checked for its consistency after comparing with the in-situ observations and the result shows underestimation of satellite measured values. The result suggests that the conjunctive analysis of in-situ, satellite, and model archive is suitable to study the impact of climate variability on the structure of marine ecosystems.

We propose a versatile model to describe the time-dependent spatial distribution of sun glint areas in satellite-based wave water imaging. This model can be used to identify whether the imaging is affected by sun glint and how strong the glint is. The observing geometry is calculated using an accurate orbit prediction method. The Cox-Munk model is used to analyze the bidirectional reflectance of wave water surface under various conditions. The effects of whitecaps and the reflectance emerging from the sea water have been considered. Using the moderate resolution atmospheric transmission radiative transfer model, we are able to effectively calculate the sun glint distribution at the top of the atmosphere. By comparing the modeled data with the medium resolution imaging spectrometer image and Feng Yun 2E (FY-2E) image, we have proven that the time-dependent spatial distribution of sun glint areas can be effectively predicted. In addition, the main factors in determining sun glint distribution and the temporal variation rules of sun glint have been discussed. Our model can be used to design satellite orbits and should also be valuable in either eliminating sun glint or making use of it.

Waste quantities, which were left over after smelting, have a wide spread distribution and threaten soils, vegetation, and above all human health. The analysis of the spectral behavior of mine wastes using reflectance spectroscopy is an alternative approach to traditional methods. It provides information about targets between visible and near-infrared and shortwave infrared (SWIR) (400 to 2500 nm) wavelengths. The spectral behavior of soils in the SWIR region using hyperspectral reflectance spectroscopy supported by ground-truth-based x-ray diffraction is analyzed, and the performance of the continuum removal (CR) method for carbonate mineral (XCO3) content estimation around mine wastes is examined. The method has been used to correlate spectral absorption band centered at 2338 nm with calcium carbonate calcite (CaCO3) concentrations in the presence of different carbonate minerals, dolomite (CaMgCO3), cerusite (PbCO3), within soil samples. The method was carried out on hyperspectral reflectance measurements collected around the mine site of Jebel Ressas in the North of Tunisia. The results show that the performance of the CR method is dependent on the spectral feature of carbonate minerals. Particularly, the CR2338   nm was found to be interesting for carbonate content prediction in presence of calcium carbonate.

Land use changes are a reflection of the effects of human activity on Earth’s ecosystems. Land use in Nagasaki has changed significantly in recent years with built-up land area increasing from 34 to 62  km2 during the study period (1976 to 2009). Landscape metrics are quantitative indexes that can streamline landscape pattern information and reflect its characteristic structural composition and spatial configuration. Using patches as landscape elements, 12 landscape metrics at two levels (class and landscape levels) were selected to analyze the dynamic changes in land use. At the class level, the following observations were made: (1) merger development of most land use types; (2) a trend toward merging and extension of built-up land and forest; and (3) land use types dominated by nature showed higher shape complexity and lower uniformity in patch area distribution than land use types dominated by human activities. At the landscape level, the results showed an increasing degree of fragmentation and decreasing degree of patch shape complexity in the study period. The study is the first analysis to be done on this scale in Nagasaki that showed the change characteristics in land use by landscape metrics.

Although unmanned aerial vehicles (UAV) can provide images with high resolution in a portable and easy way, the matching algorithms such as scale-invariant feature transform and speeded-up robust features (SURF) are often time-consuming. To reduce the time of image matching processes, a fast and low-cost method is proposed for constructing the UAV image matching framework using a satellite image. In this context, the satellite image is used as the base map of UAV images. To find the matching points between UAV and satellite images, a simplified version of SURF is designed to detect interest points. The simplified version of the SURF method uses only one octave of scale spaces to build filter response maps, and each octave is subdivided into four levels of scale spaces. Meanwhile, template matching is used to remove incorrectly matched points. The experimental results show that the method of this paper is robust and can deal with images acquired by small-sized UAVs without a position and orientation system. The method can calculate the rough overlap regions, which are then employed to narrow down the searching space. This will improve the speed of matching greatly, especially for an unordered database of images.

In order to fully consider the local spatial constraints between neighboring objects in object-based change detection (OBCD), an OBCD approach is presented by introducing a refined Markov random field (MRF). First, two periods of images are stacked and segmented to produce image objects. Second, object spectral and textual histogram features are extracted and G-statistic is implemented to measure the distance among different histogram distributions. Meanwhile, object heterogeneity is calculated by combining spectral and textual histogram distance using adaptive weight. Third, an expectation-maximization algorithm is applied for determining the change category of each object and the initial change map is then generated. Finally, a refined change map is produced by employing the proposed refined object-based MRF method. Three experiments were conducted and compared with some state-of-the-art unsupervised OBCD methods to evaluate the effectiveness of the proposed method. Experimental results demonstrate that the proposed method obtains the highest accuracy among the methods used in this paper, which confirms its validness and effectiveness in OBCD.

In the Wadi Biskra arid and semiarid areas, sustainable development is restricted by land degradation processes such as secondary salinization of soils. Being an important high-quality date production region of Algeria, this area needs continuous monitoring of desertification indicators, hence highly exposed to climate-related risks. Given the limited access to field data, appropriate methods were assessed for the identification and change detection of salt-affected areas, involving image interpretation and automated classifications employing Landsat imagery, ancillary and multisource ground truth data. First, a visual photointerpretation study of the land cover and land use classes was undergone according to acknowledged methodologies. Second, two automated classification approaches were developed: a customized decision tree classification (DTC) and an unsupervised one applied to the principal components of Knepper ratios composite. Five indices were employed in the DTC construction, among which also is a salinity index. The diachronic analysis was undergone for the 1984 to 2015 images (including seasonal approach), being supported by the interpreted land cover/land use map for error estimation. Considering also biophysical and socioeconomic data, comprehensive results are discussed. One of the most important aspects that emerged was that the accelerated expansion of agricultural land in the last three decades has led and continues to contribute to a secondary salinization of soils.

Due to weathering and external forces, solar panels are subject to fouling and defects after a certain amount of time in service. These fouling and defects have direct adverse consequences such as low-power efficiency. Because solar power plants usually have large-scale photovoltaic (PV) panels, fast detection and location of fouling and defects across large PV areas are imperative. A drone-mounted infrared thermography system was designed and developed, and its ability to detect rapid fouling on large-scale PV panel systems was investigated. The infrared images were preprocessed using the K neighbor mean filter, and the single PV module on each image was recognized and extracted. Combining the local and global detection method, suspicious sites were located precisely. The results showed the flexible drone-mounted infrared thermography system to have a strong ability to detect the presence and determine the position of PV fouling. Drone-mounted infrared thermography also has good technical feasibility and practical value in the detection of PV fouling detection.

Satellite images are widely used to map geological and environmental features at different map scales. The ability of visible to near-infrared (VNIR) scanner systems to map gossans, rich in iron and associated with weathered sulfide occurrences, as well as to characterize regoliths, is perhaps one of the most important current applications of this technology. Initial results of this study show that advanced space-borne thermal emission and reflection (ASTER), VNIR, and short-wave infrared radiometer scanner systems can be used successfully to map iron ores. By applying internal average relative reflectance, false color composite, minimum noise fraction transform, and mathematical evaluation method (MEM) techniques, iron contaminations were successfully detected in the Chadormalu iron mine area of central Iran. An attempt was also made to discriminate between the geogenic and anthropogenic iron contaminations in the vicinity of the Chadormalu iron deposit. This research compares ASTER and Landsat 8 data images and the MEM with the band ratio method in a full scope view scale and demonstrates ASTER image data capability in detecting iron contaminations in the Chadormalu area. This indicates that ASTER bands 3, 2, and 1 have a higher spatial (15 m) resolution compared with sensors used in previous works. In addition, the capability of the MEM in detecting Fe-contaminants, unlike the color judgments of the band ratio method, can discriminate between iron pollution in an alluvial plain and the Fe-contents of the host and country rocks in the study area. This study proved that Landsat 8 data illustrate exaggeration both in the MEM and band ratio final results (outputs) and cannot display iron contamination in detail.

The regional surface soil heat flux ( G 0 ) estimation is very important for the large-scale land surface process modeling. However, most of the regional G 0 estimation methods are based on the empirical relationship between G 0 and the net radiation flux. A physical model based on harmonic analysis was improved (referred to as “HM model”) and applied over the Heihe River Basin northwest China with multiple remote sensing data, e.g., FY-2C, AMSR-E, and MODIS, and soil map data. The sensitivity analysis of the model was studied as well. The results show that the improved model describes the variation of G 0 well. Land surface temperature (LST) and thermal inertia ( Γ ) are the two key input variables to the HM model. Compared with in situ G 0 , there are some differences, mainly due to the differences between remote-sensed LST and the in situ LST. The sensitivity analysis shows that the errors from − 7 to − 0.5    K in LST amplitude and from − 300 to 300    J   m − 2   K − 1  

Loktak is the largest natural wetland of Northeast India, the last home of endangered brow-antlered deer, and a site of global significance recognized under Ramsar convention. Ecological and human-meditated spatial patterns of Loktak wetland were characterized and classified using a Radarsat-2 C band synthetic aperture radar (SAR) satellite data. Radarsat-2 quad-pol scene of dry season was preprocessed and classified using PolSARpro software. Eigen vector–eigen value decomposition of coherency matrix (T3) was performed to characterize the scattering properties of wetland targets based on entropy (H)/anisotropy (A)/alpha angle (α) segmentation. Results illustrate that RGB color display of H/A/α images is a useful indicator of wetland structure and composition, and provide clear visual discrimination of open water, floating phumdi, permanent phumdi cover, and associated man-made features. Six classes, namely, floating phumdi, permanent phumdi, scrub/shrub, fallow land, built-up, and open water were mapped using Wishart classification of H/A/α images. Scattering mechanisms of natural and man-made targets synthesized from PolSAR data, and their classification using Wishart algorithm have been validated through a visually classified map and field reference points. The land cover generated would be useful for conservation and management of Loktak wetland and brow-antlered deer population.

Aiming to subsidize the rational water resources management, four Landsat 8 (L8) images along different conditions of the year 2014 were used for modeling the radiation and energy balances in the mixed agroecosystems inside a Brazilian reference semiarid area. The SAFER algorithm was applied to calculate the latent heat flux (λE); net radiation (Rn) was acquired by the Slob equation; ground heat flux (G) was considered a fraction of Rn; and the sensible heat flux (H) was retrieved by residue in the energy balance equation. For classifying the vegetation, the surface resistance algorithm (SUREAL) was used to estimate the surface resistance to the water fluxes (rs) with threshold values for rs. Clearly, one could see higher λE values from irrigated crops (ICs) than those for natural vegetation (NV) with some situations of heat horizontal advection. The respective λE, H, and G average ratios to Rn for the ICs ecosystem were 64% to 79%, 18% to 28%, and 3%, respectively. For the NV ecosystem, the corresponding fractions were 4% to 37%, 60% to 94%, and 4%, respectively. The algorithms proved to have strong sensibility to quantifying the large-scale energy and mass exchanges by applying L8 images in mixed agroecosystems of semiarid environments.

An important area of research for super-resolution reconstruction is the issue of motion estimation or mutual registration. Currently, the majority of domestic and international research focuses on registration or estimation methods. However, the reverse approach has not been considered, i.e., investigating the impact of the accuracy of motion estimation or registration on the quality of the reconstructed images, the relationship between the above-mentioned impact and the accuracy of motion estimation, and whether this relationship is quantifiable. There remains a lack of theoretical and systematic studies on these issues. By discussing super-resolution reconstruction of infrared scanning images while focusing on the impact of the registration error caused by the control system of the scanning device, the above mentioned issues are addressed. Nonlinear error transfer methods and impact analysis of blur parameters were conducted using both low-frequency and high-frequency information. After obtaining the theoretical results, experiments were performed to verify the accuracy of those results. The findings of this study can be used not only as guidelines for the theoretical evaluation of super-resolution reconstructed images but also to determine the design parameters of sensors and provide theoretical support for the design indicators of sensor control systems.

Filtering of airborne laser scanning (ALS) point clouds into ground and nonground points is a core postprocessing step for ALS data. A hierarchical filtering method, which has high operating efficiency and accuracy because of the combination of multiscale morphology and progressive triangulated irregular network (TIN) densification (PTD), is proposed. In the proposed method, the grid is first constructed for the ALS point clouds, and virtual seed points are set by analyzing the shape and elevation distribution of points within the grid. Then, the virtual seed points are classified as ground or nonground using the multiscale morphological method. Finally, the virtual ground seed points are utilized to generate the initial TIN, and the filter is completed by iteratively densifying the initial TIN. We used various ALS data to test the performance of the proposed method. The experimental results show that the proposed filtering method has strong applicability for a variety of landscapes and, in particular, has lower commission error than the classical PTD filtering method in urban areas.

A spectral–spatial classification method using a trilateral filter (TF) and stacked sparse autoencoder (SSA) for improving the classification accuracy of hyperspectral image (HSI) is proposed. The operation is carried out in two main stages: edge-preserved smoothing and high-level feature learning. First, a reference image obtained from dual tree complex wavelet transform is adopted in a TF for smoothing the HSI. As expected, the filter not only can effectively attenuate the mixed noise (e.g., Gaussian noise and impulse noise) where the bilateral filter shows poor performance but also can produce useful spectral–spatial features from HSI by considering geometric closeness and photometric similarity between pixels simultaneously. Second, an artificial fish swarm algorithm (AFSA) is first introduced into a SSA, and the proposed deep learning architecture is used to adaptively exploit more abstract and differentiable high-level feature representations from the smoothed HSI, based on the factor that AFSA provides better trade-off among concurrency, search efficiency, and convergence rate compared with gradient descent and back-propagation algorithms in a traditional SSA. Finally, a random forest classifier is utilized to perform supervised fine-tuning and classification. Experimental results on two real HSI data sets demonstrate that the proposed method generates competitive performance compared with those of conventional methods.

Evapotranspiration (ET) plays an important role in hydrological cycle by linking land surface and atmosphere through water and energy transfers. Based on the data from the Landsat-8 satellite for typical days with clear sky condition from 2013 to 2016, a two-layer daily ET remote sensing framework was built, which includes four compartments: surface feature parameter estimation, evaporative fraction estimation, daily net radiation estimation, and daily ET extension. Based on the model, evaporation, transpiration, and daily ET in Shahe River Basin were estimated. The estimated daily ET showed a mean absolute percentage error of 8.7% in the plain areas, and 12.1% in the mountainous areas, compared to observations using large aperture scintillometer and eddy covariance system. The method gave higher accuracy than other remote sensing models applied in the same area previously, including the surface energy balance system and the ETWatch. By analyzing the relationship between land use types and surface water/heat fluxes, it was found that the surface energy balance components in the basin have prominent spatial-temporal features, and the soil component’s features are more obvious. It indicated that the proposed two-layer approach is superior to others in terms of simulation accuracy, and applicable to daily scale ET estimations on complex terrains.

Information on distribution of forest types and land cover classes is essential for decision making and significant in climate regulation, biodiversity conservation, and societal issues. An approach for the combination of advanced polarimetric decompositions and textures of Advanced Land Observing Satellite Phased Array L-band Synthetic Aperture Radar full polarimetric data for the purpose of forest type classification is proposed. Using a support vector machine (SVM) classifier, we classified forest types over a selected Indian region. Further, we tested the classification performance of the Wishart method for the same forest types. The classified results were assessed with confusion matrix-based statistics. The results suggest that incorporation of various polarimetric decompositions features into gray-level co-occurrence matrix textures refines the SVM classification overall accuracy (OA) from 73.82% (k=0.69) to 76.34% (k=0.72). The Wishart supervised classification algorithm has the OA of 73.38% (kappa=0.68). We observed that integration of polarimetric information with textures can give complimentary information in forest type discrimination and produce high accuracy maps. Further, this approach overcomes the limitations of optical remote sensing data in continuous cloud coverage areas.

Long-term monitoring of shoreline changes is of significant importance for coastal erosion prediction and coastal planning. This paper presents the use of the Landsat archival dataset over 29 years to monitor shoreline changes at Narrabeen–Collaroy Beach, Australia. For each scene of Landsat data, a vector-based shoreline at subpixel level was automatically extracted by employing the super-resolution border segmentation method. Monthly cross-shore beach profile surveys at five locations along the beach are used as ground truth data to assess the results. The experimental results show that after the correction of tidal effects the RMSEs of the time-series shorelines at the five locations are all less than 10 m; RMSEs of annual mean shorelines are within 5.7 m. These results were achieved within the context that the Narrabeen–Collaroy Beach has been observed to vary in width by up to 80 m during this same period. In addition, the observed variabilities in shoreline change at different locations along the 3.6-km long embayment are also clearly observed from Landsat-derived results. The results indicate the feasibility of Landsat data for long-term shoreline change monitoring.

Tree species information is essential for forest research and management purposes, which in turn require approaches for accurate and precise classification of tree species. One such remote sensing technology, terrestrial laser scanning (TLS), has proved to be capable of characterizing detailed tree structures, such as tree stem geometry. Can TLS further differentiate between broad- and needle-leaves? If the answer is positive, TLS data can be used for classification of taxonomic tree groups by directly examining their differences in leaf morphology. An analysis was proposed to assess TLS-represented broad- and needle-leaf structures, followed by a Bayes classifier to perform the classification. Tests indicated that the proposed method can basically implement the task, with an overall accuracy of 77.78%. This study indicates a way of implementing the classification of the two major broad- and needle-leaf taxonomies measured by TLS in accordance to their literal definitions, and manifests the potential of extending TLS applications in forestry.

A triangle method has been frequently used for a regional evapotranspiration (ET) or evaporative fraction (EF) estimation using the relationship between a vegetation index and surface temperature. A simple histogram based on an on-orbit EF estimation algorithm is developed to determine hot and cold edges in the triangle distribution using only remotely sensed data. A new histogram based on an EF estimation algorithm is developed and compared with different options of existing hot and cold edge algorithms. The selected algorithms are applied to 5-year Terra MODIS data products, and the on-orbit EF estimations are validated using ground truth data from southern Great Plains. Using numerous iterations among the tested EF algorithms with the triangle method, the “histogram”-based hot edge and newly proposed cold edge algorithm combinations provide an excellent performance between the ground truth and on-orbit EF estimations. It is shown that the histogram-based EF algorithm has enhanced correlation and mean bias values of 0.74 and 0.087, respectively, providing excellent accuracy for the regional on-orbit EF estimations.

Lufft’s CHM 15k “Nimbus” ceilometer and a collocated Cimel sunphotometer were used, in tandem with satellite data, to observe the transport of atmospheric aerosols over Raciborz, Poland, during an exceptionally warm month of August 2015. Two distinct periods are identified: increased aerosol optical thickness (AOT) values, exceeding 0.65, during August 6 to 13 concurrent with the planetary boundary layer elevated up to ∼2.5  km and thin aerosol layers up to 7 km above ground visible from August 25 to 27. A newly developed scheme for backward air mass trajectory analysis is employed. The scheme utilizes satellite data on thermal anomalies as well as multiangle imaging of aerosol clouds. The obtained retrievals provide evidence that aerosols of biomass burning type were present during the first period, originating from a strong episode of wildfires in Ukraine. Moreover, satellite AOT data from the MODIS instrument show that a significant part of the observed aerosol was accumulated during transport between the region of intense biomass burning and the receptor, Raciborz. The same analysis scheme suggests that a long-range transport of biomass burning products from the United States was a source of the layers observed during the second period.

Mapping spatial distributions of tropical peatlands is important for properly estimating carbon emissions and for providing information that aids in the sustainable management of tropical peatlands, particularly in Indonesia. This study evaluated the performance of phased array type L-band synthetic aperture radar (SAR) (PALSAR) dual-polarization and fully polarimetric data for tropical peatlands classification. The study area was in Siak River Transect, Riau Province, Indonesia, a rapidly developing region, where the peatland has been intensively converted mostly into oil palm plantations over the last two decades. Thus, polarimetric features derived after polarimetric decompositions, backscatter coefficients measurements, and the radar vegetation index were evaluated to classify tropical peatlands using the decision tree classifier. Overall, polarimetric features generated by the combination of dual-polarization and fully polarimetric data yielded an overall accuracy (OA) of 69% and a kappa coefficient (K) of 0.57. The integration of an additional feature, “distance to river,” to the algorithm increased the OA to 76% and K to 0.66. These results indicated that the methodology in this study might serve as an efficient tool in tropical peatlands classification, especially when involving the use of L-band SAR dual-polarization and fully polarimetric data.

Based on a previous study that used the Defense Meteorological Satellite Program/Operational Linescan System (DMSP/OLS) nighttime light images to partition different types of night-lit areas within individual cities, 456 urban core zones in China, representing highly developed areas in 2012, were extracted. Then, several morphologic indices were selected for characterizing for each urban core zone, and the spatial differentiation and morphologic characteristics of urban core zones located within different geomorphologic regions in China were quantitatively analyzed. The results showed that urban core zones were most widely distributed in the eastern region comprising hilly plains, with a decreasing distribution trend from northeast to southwest, and the least distribution was in the Tibetan Plateau. The contours of most of these zones appeared to be relatively simple and compact, and evidenced seven shapes. Regions at lower altitudes with flat terrains were more likely to demonstrate a wide range of urban core zones, especially those with complex shapes. This study represents a preliminary effort toward the construction of an interactive coupling mechanism for urban and geomorphologic environments (e.g., altitude, relief of land surface, geomorphologic types, geomorphologic region). Its findings contribute to enhancing understanding of the spatial morphologic characteristics of highly developed areas in China.

This work proposes methodologies aimed at evaluating the sensitivity of optical and synthetic aperture radar (SAR) change features obtained from satellite images with respect to the damage grade due to an earthquake. The test case is the Mw 7.0 earthquake that hit Haiti on January 12, 2010, located 25 km west–south–west of the city of Port-au-Prince. The disastrous shock caused the collapse of a huge number of buildings and widespread damage. The objective is to investigate possible parameters that can affect the robustness and sensitivity of the proposed methods derived from the literature. It is worth noting how the proposed analysis concerns the estimation of derived features at object scale. For this purpose, a segmentation of the study area into several regions has been done by considering a set of polygons, over the city of Port-au-Prince, extracted from the open source open street map geo-database. The analysis of change detection indicators is based on ground truth information collected during a postearthquake survey and is available from a Joint Research Centre database. The resulting damage map is expressed in terms of collapse ratio, thus indicating the areas with a greater number of collapsed buildings. The available satellite dataset is composed of optical and SAR images, collected before and after the seismic event. In particular, we used two GeoEye-1 optical images (one preseismic and one postseismic) and three TerraSAR-X SAR images (two preseismic and one postseismic). Previous studies allowed us to identify some features having a good sensitivity with damage at the object scale. Regarding the optical data, we selected the normalized difference index and two quantities coming from the information theory, namely the Kullback–Libler divergence (KLD) and the mutual information (MI). In addition, for the SAR data, we picked out the intensity correlation difference and the KLD parameter. In order to analyze the capability of these parameters to correctly detect damaged areas, two different classifiers were used: the Naive Bayes and the support vector machine classifiers. The classification results demonstrate that the simultaneous use of several change features from Earth observations can improve the damage estimation at object scale.

HJ-1A hyperspectral data were used to distinguish topsoil salt components and estimate soil salinity, and the relationship between soil salt chemical components and sensitive bands of soil reflectance spectra was analyzed. The correlation between the soil salt content and the soil spectra obtained from the hyperspectral data was analyzed, proving that topsoil salinity has a very significant correlation with soil reflectance spectra. The relationship between soil reflectance spectra and salt chemical ions was investigated. The soil spectral reflectance at wavelength 510.975 nm and a difference vegetation index were selected to estimate soil salinity and the dominant salt chemical ion concentrations at a depth of 0 to 10 cm using a partial least squares regression model. It was found that the bands sensitive to various levels of chemical components of soil salt were shown to differ, controlled by the dominant component of the soil salt. The sensitive bands in the soil salinity estimation will change with differences in salt components. Estimating the dominant salt in the soil using soil reflectance spectra will lead to greater prediction accuracy. This study provided a possible method for the estimation of salinity and chemical component levels in topsoil, using the hyperspectral data to estimate topsoil salt components.