Natural tropical rainforests in China’s Xishuangbanna region have undergone dramatic conversion to rubber plantations in recent decades, resulting in altering the region’s environment and ecological systems. Therefore, it is of great importance for local environmental and ecological protection agencies to research the distribution and expansion of rubber plantations. The objective of this paper is to monitor dynamic changes of rubber plantations in China’s Xishuangbanna region based on multitemporal Landsat images (acquired in 1989, 2000, and 2013) using a C5.0-based decision-tree method. A practical and semiautomatic data processing procedure for mapping rubber plantations was proposed. Especially, haze removal and deshadowing were proposed to perform atmospheric and topographic correction and reduce the effects of haze, shadow, and terrain. Our results showed that the atmospheric and topographic correction could improve the extraction accuracy of rubber plantations, especially in mountainous areas. The overall classification accuracies were 84.2%, 83.9%, and 86.5% for the Landsat images acquired in 1989, 2000, and 2013, respectively. This study also found that the Landsat-8 images could provide significant improvement in the ability to identify rubber plantations. The extracted maps showed the selected study area underwent rapid conversion of natural and seminatural forest to a rubber plantations from 1989 to 2013. The rubber plantation area increased from 2.8% in 1989 to 17.8% in 2013, while the forest/woodland area decreased from 75.6% in 1989 to 44.8% in 2013. The proposed data processing procedure is a promising approach to mapping the spatial distribution and temporal dynamics of rubber plantations on a regional scale.

By separating the surface and subsurface components of foliar hyperspectral signatures using polarization optics, it is possible to enhance the remote discrimination of different plant species and optimize the assessment of different factors associated with their health status. These initiatives, in turn, can lead to higher crop yield and lower environmental impact. It is important to consider, however, that the main varieties of crops, represented by C3 (e.g., soy) and C4 (e.g., maize) plants, have markedly distinct morphological characteristics. Accordingly, the influence of these characteristics on their interactions with impinging light may affect the selection of optimal probe wavelengths for specific applications making use of combined hyperspectral and polarization measurements. In this paper, we compare the sensitivity of the total (including surface and subsurface components) and subsurface reflectance responses of C3 and C4 plants to different spectral and geometrical light incidence conditions. This investigation is supported by measured biophysical data and predictive light transport simulations. The results of our comparisons indicate that the total and subsurface reflectance responses of C3 and C4 plants depict well-defined patterns of sensitivity for varying illumination conditions. We believe that these patterns should be considered in the design of high-fidelity crop discrimination and monitoring procedures.

Colorado potato beetle (CPB) adults and larvae devour leaves of potato and other solanaceous crops and weeds, and may quickly develop resistance to pesticides. With early detection of CPB damage, more options are available for precision integrated pest management, which reduces the amount of pesticides applied in a field. Remote sensing with small unmanned aircraft systems (sUAS) has potential for CPB detection because low flight altitudes allow image acquisition at very high spatial resolution. A five-band multispectral sensor and up-looking incident light sensor were mounted on a six-rotor sUAS, which was flown at altitudes of 60 and 30 m in June 2014. Plants went from visibly undamaged to having some damage in just 1 day. Whole-plot normalized difference vegetation index (NDVI) and the number of pixels classified as damaged (0.70≤NDVI≤0.80) were not correlated with visible CPB damage ranked from least to most. Area of CPB damage estimated using object-based image analysis was highly correlated to the visual ranking of damage. Furthermore, plant height calculated using structure-from-motion point clouds was related to CPB damage, but this method required extensive operator intervention for success. Object-based image analysis has potential for early detection based on high spatial resolution sUAS remote sensing.

An effective super-resolution (SR) algorithm is proposed for actual spectral remote sensing images based on sparse representation and wavelet preprocessing. The proposed SR algorithm mainly consists of dictionary training and image reconstruction. Wavelet preprocessing is used to establish four subbands, i.e., low frequency, horizontal, vertical, and diagonal high frequency, for an input image. As compared to the traditional approaches involving the direct training of image patches, the proposed approach focuses on the training of features derived from these four subbands. The proposed algorithm is verified using different spectral remote sensing images, e.g., moderate-resolution imaging spectroradiometer (MODIS) images with different bands, and the latest Chinese Jilin-1 satellite images with high spatial resolution. According to the visual experimental results obtained from the MODIS remote sensing data, the SR images using the proposed SR algorithm are superior to those using a conventional bicubic interpolation algorithm or traditional SR algorithms without preprocessing. Fusion algorithms, e.g., standard intensity-hue-saturation, principal component analysis, wavelet transform, and the proposed SR algorithms are utilized to merge the multispectral and panchromatic images acquired by the Jilin-1 satellite. The effectiveness of the proposed SR algorithm is assessed by parameters such as peak signal-to-noise ratio, structural similarity index, correlation coefficient, root-mean-square error, relative dimensionless global error in synthesis, relative average spectral error, spectral angle mapper, and the quality index Q4, and its performance is better than that of the standard image fusion algorithms.

Hyperspectral remote sensing has been widely used in mineral identification using the particularly useful short-wave infrared (SWIR) wavelengths (1.0 to 2.5  μm). Current mineral mapping methods are easily limited by the sensor’s radiometric sensitivity and atmospheric effects. Therefore, a simple mineral mapping algorithm (SMMA) based on the combined application with multitype diagnostic SWIR absorption features for hyperspectral data is proposed. A total of nine absorption features are calculated, respectively, from the airborne visible/infrared imaging spectrometer data, the Hyperion hyperspectral data, and the ground reference spectra data collected from the United States Geological Survey (USGS) spectral library. Based on spectral analysis and statistics, a mineral mapping decision-tree model for the Cuprite mining district in Nevada, USA, is constructed. Then, the SMMA algorithm is used to perform mineral mapping experiments. The mineral map from the USGS (USGS map) in the Cuprite area is selected for validation purposes. Results showed that the SMMA algorithm is able to identify most minerals with high coincidence with USGS map results. Compared with Hyperion data (overall accuracy=74.54%), AVIRIS data showed overall better mineral mapping results (overall accuracy=94.82%) due to low signal-to-noise ratio and high spatial resolution.

Urban surface water is characterized by complex surface continents and small size of water bodies, and the mapping of urban surface water is currently a challenging task. The moderate-resolution remote sensing satellites provide effective ways of monitoring surface water. This study conducts an exploratory evaluation on the performance of the newly available Sentinel-2A multispectral instrument (MSI) imagery for detecting urban surface water. An automatic framework that integrates pixel-level threshold adjustment and object-oriented segmentation is proposed. Based on the automated workflow, different combinations of visible, near infrared, and short-wave infrared bands in Sentinel-2 image via different water indices are first compared. Results show that object-level modified normalized difference water index (MNDWI with band 11) and automated water extraction index are feasible in urban surface water mapping for Sentinel-2 MSI imagery. Moreover, comparative results are obtained utilizing optimal MNDWI from Sentinel-2 and Landsat 8 images, respectively. Consequently, Sentinel-2 MSI achieves the kappa coefficient of 0.92, compared with that of 0.83 from Landsat 8 operational land imager.

The development of satellite altimeters (SALTs) has brought huge benefits, among which is the ability to more adequately sense ocean-surface topography. The radar altimeter database system was used to capture and process ENVISAT, CRYOSAT-2, SARAL, JASON-1, and JASON-2 SALT data of 5 years between 2011 and 2015. The time series of monthly multimission SALT data showed an estimated sea level trend of 1.0, 2.4, 2.4, 3.6, and 12.0  mm/year at Gelang, Port Kelang, Kukup, Cendering, and Keling. The correlation analysis for the selected tide gauge stations produced satisfying results of R-squared with 0.86, 0.89, 0.91, and 0.97 for Cendering, Sedili, Gelang, and Geting, respectively. The ITG-Grace2010s geoid model was used to compute the mean dynamic topography (MDT) and plot to a grid of 0.25 deg for the Malacca Strait and South China Sea of Peninsular Malaysia, with Keling, Port Kelang, Geting, Sedili, and Johor Bahru tide gauge stations having values determined by interpolation to be 1.14, 1.19, 1.26, 1.88, and 2.91 m, respectively. MDT is computed from the SALT with respect to Port Kelang, the north–south sea slope ranges between −0.64 and 0.29  m/50  km and −0.01 and 0.52  m/50  km along the east and west coasts of Peninsular Malaysia, respectively.

A band selection algorithm based on information entropy is proposed for hyperspectral image classification. First, original spectral features are transformed into discrete features and represented by a discrete space model. Then, the band selection algorithm based on information entropy is adopted to reduce feature dimensionality. The bands with weak class separability are effectively abandoned by the band selection algorithm. Moreover, support vector machine classifiers with composite kernels are employed to incorporate spatial features into spectral features, reducing speckle errors in the classification maps. The proposed methods are applied to three benchmark hyperspectral data sets for classification. The performance of the proposed methods is compared with a band selection algorithm based on mutual information. The experimental results demonstrate that the band selection algorithm based on information entropy can effectively reduce feature dimensionality and improve classification accuracy.

Accurate and continuous estimation of evapotranspiration (ET) is crucial for effective water resource management. We used the moderate resolution imaging spectroradiometer (MODIS) standard ET algorithm forced by the MODIS land products and the three-hourly solar radiation datasets to estimate daily actual evapotranspiration of China (ET_MOD) for the years 2001 to 2015. From the point scale validations using seven eddy covariance tower sites, the results showed that the agreement of ET_MOD estimates and observations was higher for monthly and daily values than that of instantaneous values. Under the major river basin and subbasin levels’ comparisons with the variable infiltration capacity hydrological model estimates, the ET_MOD exhibited a slight overestimation in northern China and underestimation in southern China. The mean annual ET_MOD estimates agreed favorably with the hydrological model with coefficients of determination (R2) of 0.93 and 0.83 at major river basin and subbasin scale, respectively. At national scale, the spatiotemporal variations of ET_MOD estimates matched well with those ET estimates from various sources. However, ET_MOD estimates were generally lower than the other estimates in the Tibetan Plateau. This underestimation may be attributed to the plateau climate along with low air temperature and sparsely vegetated surface on the Tibetan Plateau.

Spartina alterniflora is an aggressive invasive plant species that replaces native species, changes the structure and function of the ecosystem across coastal wetlands in China, and is thus a major conservation concern. Mapping the spread of its invasion is a necessary first step for the implementation of effective ecological management strategies. The performance of a phenology-based approach for S. alterniflora mapping is explored in the coastal wetland of the Yangtze Estuary using a time series of GaoFen satellite no. 1 wide field of view camera (GF-1 WFV) imagery. First, a time series of the normalized difference vegetation index (NDVI) was constructed to evaluate the phenology of S. alterniflora. Two phenological stages (the senescence stage from November to mid-December and the green-up stage from late April to May) were determined as important for S. alterniflora detection in the study area based on NDVI temporal profiles, spectral reflectance curves of S. alterniflora and its coexistent species, and field surveys. Three phenology feature sets representing three major phenology-based detection strategies were then compared to map S. alterniflora: (1) the single-date imagery acquired within the optimal phenological window, (2) the multitemporal imagery, including four images from the two important phenological windows, and (3) the monthly NDVI time series imagery. Support vector machines and maximum likelihood classifiers were applied on each phenology feature set at different training sample sizes. For all phenology feature sets, the overall results were produced consistently with high mapping accuracies under sufficient training samples sizes, although significantly improved classification accuracies (10%) were obtained when the monthly NDVI time series imagery was employed. The optimal single-date imagery had the lowest accuracies of all detection strategies. The multitemporal analysis demonstrated little reduction in the overall accuracy compared with the use of monthly NDVI time series imagery. These results show the importance of considering the phenological stage for image selection for mapping S. alterniflora using GF-1 WFV imagery. Furthermore, in light of the better tradeoff between the number of images and classification accuracy when using multitemporal GF-1 WFV imagery, we suggest using multitemporal imagery acquired at appropriate phenological windows for S. alterniflora mapping at regional scales.

This current study explores satellite-based soil moisture downscaling approaches and applies them to common passive microwave retrievals. Three variations of a second-order polynomial regression were tested based on the surface temperature/greenness index concept and merged information from higher spatial resolution moderate-resolution imaging spectroradiometer with soil moisture active passive (SMAP) and soil moisture and ocean salinity (SMOS) products to obtain soil moisture estimates at higher resolutions (1 km). Downscaled products were evaluated at the Walnut Gulch Experimental Watershed (WGEW) in southeastern Arizona. Results show slight differences in performance among the three downscaling methods and little improvement between original low-resolution products and downscaled (1 km) products. Spatial analysis over WGEW demonstrates downscaled products were able to decipher small-scale heterogeneities in surface soil moisture, though spatial variability remains low compared to observations with a difference of only 0.06  m3/m3 in spatial standard deviation between observations and the mean between downscaling techniques. Results demonstrate the ability of both SMOS and SMAP to represent soil moisture accurately on the point scale without applying downscaling techniques in the region under study.

Airborne synthetic aperture radar (SAR) data have been successfully used for forest height inversion; however, there is limited applicability in spaceborne scenarios due to high temporal decorrelation. This study investigates the potential of a high-resolution fully polarimetric interferometric pair of TerraSAR-X/TanDEM-X SAR data with no temporal decorrelation to analyze the backscatter and coherence response and to implement polarimetric SAR interferometry-based height inversion algorithms. The data were acquired over Barkot forest region of Uttarakhand state in India. Yamaguchi decomposition was implemented onto the dataset to express total backscatter as a sum of different scattering components from a single SAR resolution cell. Coherency matrix was used to compute complex coherence for different polarization channels. Forest areas suffered from low coherence due to volume decorrelation, whereas a dry river bed had shown high coherence. The coherence amplitude inversion approach overestimated the forest height and also resulted in false heights for this dry river bed. These limitations were overcome by implementing three-stage inversion modeling, which assumes polarization-independent volume coherence. The results were validated using ground truth data available for 49 plots, and the latter was found to be more accurate with an overall accuracy of 90.15% and root-mean-square error of 2.42 m.

Separating sea surface and land areas in synthetic aperture radar (SAR) images is challenging yet of great importance to coastline extraction and subsequent coastal classification. Results of the previous state-of-art methods often suffer from a number of limitations that arise from the presence of the speckle effect and the inadequate returned signal around the boundaries. We propose a graph cut (GC)-based approach to tackle these limitations and achieve accurate sea–land segmentation results. To be more specific, as the first step, three powerful multipolarization features are extracted from the polarimetric SAR data as descriptors to fully characterize the sea area and land area. Starting from that, seeds of the sea and land are selected automatically to build the prior model for GC. Based on the prior model, we construct the undirected graph in GC using the multipolarization descriptors. Finally, we incorporate the ratio of average operator to eliminate the speckle effect and get finer results for some finer structures. Experiments on Radarsat-2 quad-polarization images demonstrate significantly improved results of our proposed algorithms compared with several state-of-the-art methods in terms of both quantitative and visual performance.

It has been suggested that Tianjin, China, has significant land subsidence due to excessive extraction of water. Although it is presently under control, the land subsidence around Tianjin suburbs in recent years should not be ignored. However, existing research work on land subsidence is based on traditional synthetic aperture radar satellite images in which the research time spans are mainly before 2012. An advanced time-series method, namely small baselines subset (SBAS) technique, is applied to a total of 27 Sentinel-1A images over Tianjin acquired between May 31, 2015, and May 13, 2016, to derive the subsidence magnitude and distribution of Tianjin. Furthermore, the overall and quantitative validations of SBAS-derived results are implemented. First, the overall subsidence distribution derived by SBAS is compared with the annual report of land subsidence in Tianjin 2015, which shows the same subsidence trend and distribution. Then, 44 benchmarks and 2 continuously operating reference station datasets, i.e., CH01 and XQYY, are processed to provide a specific validation of SBAS-derived results of Sentinel-1A. Finally, through investigation, an interpretation from two aspects of groundwater extraction and geological structures of the surrounding Wangqingtuo settlement funnel area is given.

Hyperspectral absorption features are important indicators of characterizing plant biophysical variables for the automatic diagnosis of crop diseases. Continuous wavelet analysis has proven to be an advanced hyperspectral analysis technique for extracting absorption features; however, specific wavelet features (WFs) and their relationship with pathological characteristics induced by different infestations have rarely been summarized. The aim of this research is to determine the most sensitive WFs for identifying specific pathological lesions from yellow rust and powdery mildew in winter wheat, based on 314 hyperspectral samples measured in field experiments in China in 2002, 2003, 2005, and 2012. The resultant WFs could be used as proxies to capture the major spectral absorption features caused by infestation of yellow rust or powdery mildew. Multivariate regression analysis based on these WFs outperformed conventional spectral features in disease detection; meanwhile, a Fisher discrimination model exhibited considerable potential for generating separable clusters for each infestation. Optimal classification returned an overall accuracy of 91.9% with a Kappa of 0.89. This paper also emphasizes the WFs and their relationship with pathological characteristics in order to provide a foundation for the further application of this approach in monitoring winter wheat diseases at the regional scale.

Riparian vegetation restoration efforts require cost-effective, accurate, and replicable impact assessments. We present a method to use an unmanned aerial vehicle (UAV) equipped with a GoPro digital camera to collect photogrammetric data of a 0.8-ha riparian restoration. A three-dimensional point cloud was created from the photos using “structure from motion” techniques. The point cloud was analyzed and compared to traditional, ground-based monitoring techniques. Ground-truth data were collected on 6.3% of the study site and averaged across the entire site to report stem heights in stems/ha in three height classes. The project site was divided into four analysis sections, one for derivation of parameters used in the UAV data analysis and the remaining three sections reserved for method validation. Comparing the ground-truth data to the UAV generated data produced an overall error of 21.6% and indicated an R2 value of 0.98. A Bland–Altman analysis indicated a 95% probability that the UAV stems/section result will be within 61  stems/section of the ground-truth data. The ground-truth data are reported with an 80% confidence interval of ±1032  stems/ha; thus, the UAV was able to estimate stems well within this confidence interval.

Persistent scatterer interferometry (PSI) analysis of a large area is always a challenging task regarding the removal of the atmospheric phase component. This work presents an investigation of ground movement measurements based on a combination of differential SAR interferometry time-series (DTS) and PSI techniques, applied on a large area of extent with open pit iron mines located in Carajás (Brazilian Amazon Region), aiming at detecting linear and nonlinear ground movement. These mines have presented a history of instability, and surface monitoring measurements over sectors of the mines (pit walls) have been carried out based on ground-based radar and total station (prisms). Using a priori information regarding the topographic phase error and a phase displacement model derived from DTS, temporal phase unwrapping in the PSI processing and the removal of the atmospheric phases can be performed more efficiently. A set of 33 TerraSAR-X (TSX-1) images, acquired during the period from March 2012 to April 2013, was used to perform this investigation. The DTS analysis was carried out on a stack of multilook unwrapped interferograms using an extension of SVD to obtain the least-square solution. The height errors and deformation rates provided by the DTS approach were subtracted from the stack of interferograms to perform the PSI analysis. This procedure improved the capability of the PSI analysis for detecting high rates of deformation, as well as increased the numbers of point density of the final results. The proposed methodology showed good results for monitoring surface displacement in a large mining area, which is located in a rain forest environment, providing very useful information about the ground movement for planning and risk control.

Vehicle detection in high-resolution aerial images has received widespread interests when it comes to providing the required information for traffic management and urban planning. It is challenging due to the relatively small size of the vehicles and the complex background. Furthermore, it is particularly challenging if the higher detection efficiency is required. Therefore, an urban vehicle detection algorithm is proposed via improved entropy rate clustering (IERC) and correlation-based sequential dictionary learning (CSDL). First, to enhance the detection accuracy, IERC is designed to generate more regular superpixels. It aims to avoid the situation that one superpixel sometimes straddles multiple vehicles. The generated superpixels are then treated as the seeds for the training sample selection. Then, CSDL is constructed to achieve a fast sequential training and updating of the dictionary. In CSDL, only the atoms correlated with the sparse representation of the new training data are inferred. Finally, comprehensive analyses and comparisons on two data sets demonstrate that the proposed method generates satisfactory and competitive results.

The North China Plain contains 65% of the country’s agricultural land and 24% of its fresh water resources. Monthly (January 2003 to December 2012) Gravity Recovery and Climate Experiment (GRACE) data were used to quantify anthropogenic impacts on groundwater depletion rates in the northern China region; areas include the North China Plain and surroundings. Nongroundwater components of GRACE-derived terrestrial water storage (TWS) were removed using the outputs of the Community Land Model version 4.5 (CLM4.5) spanning the same period. Results indicate that the northern China region witnessed a TWS depletion of 14.09 ± 1.74 × 10 9    m 3 / year ( 9.39 ± 1.16    mm / year ). The GRACE-derived groundwater depletion rate was estimated at 12.78 ± 1.56 × 10 9    m 3 / year ( 8.52 ± 1.04    mm / year ), which represents 91% of the GRACE-derived TWS trend. Variations in combined soil moisture, snow, vegetation canopy, and river channel storages are minimal (9%) across northern China region. Anthropogenic variations in groundwater depletion rates were estimated at 19.50 ± 1.71 × 10 9    m 3 / year