Remote Sensing

The development of analytical software for big Earth observation data faces several challenges. Designers need to balance between conflicting factors. Solutions that are efficient for specific hardware architectures can not be used in other environments. Packages that work on generic hardware and open standards will not have the same performance as dedicated solutions. Software that assumes that its users are computer programmers are flexible but may be difficult to learn for a wide audience. This paper describes sits, an open-source R package for satellite image time series analysis using machine learning. To allow experts to use satellite imagery to the fullest extent, sits adopts a time-first, space-later approach. It supports the complete cycle of data analysis for land classification. Its API provides a simple but powerful set of functions. The software works in different cloud computing environments. Satellite image time series are input to machine learning classifiers, and the results are post-processed using spatial smoothing. Since machine learning methods need accurate training data, sits includes methods for quality assessment of training samples. The software also provides methods for validation and accuracy measurement. The package thus comprises a production environment for big EO data analysis. We show that this approach produces high accuracy for land use and land cover maps through a case study in the Cerrado biome, one of the world’s fast moving agricultural frontiers for the year 2018. Full article

Land use/land cover (LULC) changes have been observed in the Gaborone dam catchment since the 1980s. A comprehensive analysis of future LULC changes is therefore necessary for the purposes of future land use and water resource planning and management. Recent advances in geospatial modelling techniques and the availability of remotely sensed data have become central to the monitoring and assessment of both past and future environmental changes. This study employed the cellular automata and Markov chain (CA-Markov) model combinations to simulate future LULC changes in the Gaborone dam catchment. Classified Landsat images from 1984, 1995, 2005 and 2015 were used to simulate the likely LULCs in 2015 and 2035. Model validation compared the simulated and observed LULCs of 2015 and showed a high level of agreement with Kappa variation estimates of K_no (0.82), K_loc (0.82) and K_standard (0.76). Simulation results indicated a projected increase of 26.09%, 65.65% and 55.78% in cropland, built-up and bare land categories between 2015 and 2035, respectively. Reductions of 16.03%, 28.76% and 21.89% in areal coverage are expected for shrubland, tree savanna and water body categories, respectively. An increase in built-up and cropland areas is anticipated in order to meet the population’s demand for residential, industry and food production, which should be taken into consideration in future plans for the sustainability of the catchment. In addition, this may lead to water quality and quantity (both surface and groundwater) deterioration in the catchment. Moreover, water body reductions may contribute to water shortages and exacerbate droughts in an already water-stressed catchment. The loss of vegetal cover and an increase in built-up areas may result in increased runoff incidents, leading to flash floods. The output of the study provides useful information for land use planners and water resource managers to make better decisions in improving future land use policies and formulating catchment management strategies within the framework of sustainable land use planning and water resource management. Full article

Although measurement data from the civil engineering sector are an important basis for scientific analyses in the field of non-destructive testing (NDT), there is still no uniform representation of these data. An analysis of data sets across different test objects or test types is therefore associated with a high manual effort. Ontologies and the semantic web are technologies already used in numerous intelligent systems such as material cyberinfrastructures or research databases. This contribution demonstrates the application of these technologies to the case of the ¹H nuclear magnetic resonance relaxometry, which is commonly used to characterize water content and porosity distribution in solids. The methodology implemented for this purpose was developed specifically to be applied to materials science (MS) tests. The aim of this paper is to analyze such a methodology from the perspective of data interoperability using ontologies. Three benefits are expected from this approach to the study of the implementation of interoperability in the NDT domain: First, expanding knowledge of how the intrinsic characteristics of the NDT domain determine the application of semantic technologies. Second, to determine which aspects of such an implementation can be improved and in what ways. Finally, the baselines of future research in the field of data integration for NDT are drawn. Full article

Analyzing the surface and bedrock locations in radar imagery enables the computation of ice sheet thickness, which is important for the study of ice sheets, their volume and how they may contribute to global climate change. However, the traditional handcrafted methods cannot quickly provide quantitative, objective and reliable extraction of information from radargrams. Most traditional handcrafted methods, designed to detect ice-surface and ice-bed layers from ice sheet radargrams, require complex human involvement and are difficult to apply to large datasets, while deep learning methods can obtain better results in a generalized way. In this study, an end-to-end multi-scale attention network (MsANet) is proposed to realize the estimation and reconstruction of layers in sequences of ice sheet radar tomographic images. First, we use an improved 3D convolutional network, C3D-M, whose first full connection layer is replaced by a convolution unit to better maintain the spatial relativity of ice layer features, as the backbone. Then, an adjustable multi-scale module uses different scale filters to learn scale information to enhance the feature extraction capabilities of the network. Finally, an attention module extended to 3D space removes a redundant bottleneck unit to better fuse and refine ice layer features. Radar sequential images collected by the Center of Remote Sensing of Ice Sheets in 2014 are used as training and testing data. Compared with state-of-the-art deep learning methods, the MsANet shows a 10% reduction (2.14 pixels) on the measurement of average mean absolute column-wise error for detecting the ice-surface and ice-bottom layers, runs faster and uses approximately 12 million fewer parameters. Full article

The identification and management of ecological restoration areas play important roles in promoting sustainable urban development. However, current research lacks a scientific basis for the scope and scale of ecological restoration. Further, the absence of a framework to assess policy goals and public preferences that leads to identification of ecological restoration areas across the science-policy interface is difficult, and the existing frameworks’ performance has little applicability. We proposed a transdisciplinary framework to combine ecological quality, ecological health, and ecosystem services as an assessment endpoint to identify priority restoration areas. Further, we classified the ecological restoration areas on a township scale by K-means. Based upon policy goals and public preferences of the Pearl River Delta urban agglomeration, we chose air quality, biodiversity, soil fragility, recreation quality, ecosystem vigor, landscape metrics, and the water supply ecosystem service as elements of the evaluation system. This study showed that priority restoration areas accounted for 10.8% of the urban agglomeration area and classified township, largely in the difference between natural and semi-natural ecosystems and the human environment. Policymakers can use this framework comprehensively and flexibly to identify and classify ecological restoration areas to achieve policy goals and fulfil public preferences. Full article

Deep learning has been widely used in various computer vision tasks. As a result, researchers have begun to explore the application of deep learning for pansharpening and have achieved remarkable results. However, most current pansharpening methods focus only on the mapping relationship between images and the lack overall structure enhancement, and do not fully and completely research optimization goals and fusion rules. Therefore, for these problems, we propose a pansharpening generative adversarial network with multilevel structure enhancement and a multistream fusion architecture. This method first uses multilevel gradient operators to obtain the structural information of the high-resolution panchromatic image. Then, it combines the spectral features with multilevel gradient information and inputs them into two subnetworks of the generator for fusion training. We design a comprehensive optimization goal for the generator, which not only minimizes the gap between the fused image and the real image but also considers the adversarial loss between the generator and the discriminator and the multilevel structure loss between the fused image and the panchromatic image. It is worth mentioning that we comprehensively consider the spectral information and the multilevel structure as the input of the discriminator, which makes it easier for the discriminator to distinguish real and fake images. Experiments show that our proposed method is superior to state-of-the-art methods in both the subjective visual and objective assessments of fused images, especially in road and building areas. Full article

A fast voxel traversal algorithm for ray tracing was applied to build a 4 × 4 × 20 tomography model using the observation data of 11 ground-based Global Navigation Satellite System (GNSS) meteorology (GNSS/MET) stations in Hebei Province, China. The precipitation water vapor (PWV) observed at 05 a.m. (Universal Time Coordinated: UTC) on 10 December 2019, was used to reconstruct three-dimensional (3D) water vapor density fields over the test area. The tomographic results (GNSS_T) show that the water vapor density above this area is mainly below 25 g/m³ and is concentrated between the first to the fourth layers. The vertical distribution conforms to the exponential characteristics, while the horizontal distribution shows a decreasing trend from southwest to northeast. In addition, the results of the 0.25° grid dataset generated by the Global Forecast System (GFS) of the National Center for Environmental Forecasting (NCEP) (GFS_L) were interpolated to the height of the tomographic grid, which is in good agreement with the tomographic results. GFS_L is larger than GNSS_T on the first floor at the surface, with an average deviation of 0.19 g/m³. In contrast, GFS_L from the second floor to the top of the model is smaller than GNSS_T, with the average deviations distributed between −0.08 and −0.15 g/m³. Full article

An airgun source in a water reservoir has been developed in the past decade as a green active source that had been proven effective to derive short-term subsurface structural changes. However, seasonal water level fluctuation in the reservoir affects the airgun signal, and thus whether the airgun signals can be used to derive robust seasonal variation in subsurface structure remains unclear. We use the airgun data observed in the Binchuan basin to estimate the seasonal variation of seismic travel time and compare the results with those derived from ambient noise data in the same frequency band. Our main observation is that seasonal change δt/t from airgun is negatively correlated to the variation of dominant frequency and water table fluctuation in the reservoir. One possible explanation is that water table fluctuation in the reservoir affects the dominant frequency of the airgun signal and causes significant phase shift. We also compute the travel time changes in P-wave from the empirical Green’s function after deconvolving the waveforms from a reference station that is 50 m from the airgun source. The dominant frequency after deconvolution still shows seasonal variation and correlates inversely to the travel time changes, suggesting that deconvolution cannot completely eliminate the source effect on travel time changes. We also use ambient noise cross-correlation to retrieve coda waves and then derive travel time changes in monthly stacked cross-correlations relative to a yearly average cross-correlation. We observe that seismic travel time increases to its local maximum in the end of August. The travel time changes lag behind the precipitation for about one month. We apply a poroelastic physical model to explain seismic travel time changes and find that a combined effect from precipitation and evaporation might induce the seasonal changes as shown in the ambient noise data. However, the pattern of travel time changes from the airgun differs from that from ambient noise, reflecting the strong effects of airgun source property changes. Therefore, we should be cautious to derive long-term subsurface structural variation from the airgun source and put more attention on stabilizing the dominant frequency of each excitation in the future experiments. Full article

The DLR Earth Sensing Imaging Spectrometer (DESIS) is the first hyperspectral imaging spectrometer installed in the Multi-User System for Earth Sensing (MUSES) on the International Space Station (ISS) for acquiring routine science grade images from orbit. It was launched on 29 June 2018 and integrated into MUSES. DESIS measures energy in the spectral range of 400 to 1000 nm with high spatial and spectral resolution: 30 m and 2.55 nm, respectively. DESIS data should be sufficiently quantitative and accurate to use it for different applications and research. This work performs a radiometric evaluation of DESIS Level 1 product (Top of Atmosphere (TOA) reflectance) by comparing it with coincident Radiometric Calibration Network (RadCalNet) measurements at Railroad Valley Playa (RVUS), Gobabeb (GONA), and La Crau (LCFR). RVUS, GONA, and LCFR offer 4, 15, and 5 coincident datasets between DESIS and RadCalNet measurements, respectively. The results show an agreement between DESIS and RadCalNet TOA reflectance within ~5% for most spectral regions. However, there is an additional ~5% disagreement across the wavelengths affected by water vapor absorption and atmospheric scattering. Among the three RadCalNet sites, RVUS and GONA show a similar measurement disagreement with DESIS of ~5%, while LCFR differs by ~10%. Agreement between DESIS and RadCalNet measurements is variable across all three sites, likely due to surface type differences. DESIS and RadCalNet agreement show a precision of ~2.5%, 4%, and 7% at RVUS, GONA, and LCFR, respectively. RVUS and GONA, which have a similar surface type, sand, have a similar level of radiometric accuracy and precision, whereas LCFR, which consists of sparse vegetation, has lower accuracy and precision. The observed precision of DESIS Level 1 products from all the sites, especially LCFR, can be improved with a better Bidirectional Reflection Distribution Function (BRDF) characterization of the RadCalNet sites. Full article

In this paper, we present a method for retrieving sea surface wind field (SSWF) from HaiYang-2B (HY-2B) scatterometer data. In contrast to the conventional algorithm, i.e., using a point-to-point (P2P) method based on geophysical model functions (GMF) to retrieve SSWF by spaceborne scatterometer, we introduce a more accurate field-to-field (F2F) retrieval method based on convolutional neural network (CNN). We fully consider the spatial correlation and continuity between adjacent observation points, and input the observation data of continuous wind field within a certain range into the neural network to construct the neural network model, and then synchronously obtain the wind field within the range. The wind field obtained by our retrieval method maintains its continuity and solves the problem of ambiguity removal in traditional wind direction retrieval methods. Comparing the retrieval results with the buoy data, the results show that the root mean square errors (RMSE) of wind direction and wind speed are less than 0.18 rad (10.31°) and 0.75 m/s, respectively. The retrieval accuracy is better than the L2B product of HY-2B. Full article

Land reclamation has been increasingly employed in many coastal cities to resolve issues associated with land scarcity and natural hazards. Especially, land subsidence is a non-negligible environmental geological problem in reclamation areas, which is essentially caused by soil consolidation. However, spatial-scale evaluation on the average degree of consolidation (ADC) of soil layers and the effects of soil consolidation on land subsidence have rarely been reported. This study aims to carry out the integrated analysis on soil consolidation and subsidence mechanism in Chongming East Shoal (CES) reclamation area, Shanghai, at spatial-, macro-, and micro-scale so that appropriate guides can be provided to resist the potential environmental hazards. The interferometric synthetic aperture radar (InSAR) technique was utilized to retrieve the settlement curves of the selected onshore (Ra) and offshore (Rb) areas. Then, the hyperbolic (HP) model and three-point modified exponential (TME) model were combined applied to predict the ultimate settlement and to determine the range of ADC rather than a single pattern. With two boreholes Ba and Bb set within Ra and Rb, conventional tests, MIP test, and SEM test were conducted on the collected undisturbed soil to clarify the geological features of exposed soil layers and the micro-scale pore and structure characteristics of representative compression layer. The preliminary results showed that the ADC in Rb (93.1–94.1%) was considerably higher than that in Ra (60.8–78.7%); the clay layer was distinguished as the representative compression layer; on micro-scale, the poor permeability conditions contributed to the low consolidation efficiency and slight subsidence in Rb, although there was more compression space. During urbanization, the offshore area may suffer from potential subsidence when it is subjected to an increasing ground load, which requires special attention. Full article

Estimating the height of buildings and vegetation in single aerial images is a challenging problem. A task-focused Deep Learning (DL) model that combines architectural features from successful DL models (U-NET and Residual Networks) and learns the mapping from a single aerial imagery to a normalized Digital Surface Model (nDSM) was proposed. The model was trained on aerial images whose corresponding DSM and Digital Terrain Models (DTM) were available and was then used to infer the nDSM of images with no elevation information. The model was evaluated with a dataset covering a large area of Manchester, UK, as well as the 2018 IEEE GRSS Data Fusion Contest LiDAR dataset. The results suggest that the proposed DL architecture is suitable for the task and surpasses other state-of-the-art DL approaches by a large margin. Full article

Accurately identifying potential land use conflicts (LUCs) is critical for alleviating the ever-intensifying contradictions between humans and nature. The previous studies using the method of suitability analysis did not take full advantage of the current land use and multi-function characteristics of land resources. In this study, an improved model of suitability analysis was realized. In order to explore the LUCs status, including the types, intensity and distribution, a multi-objective suitability evaluation model was constructed from the perspective of production-living-ecological functions. And it was applied to Hengkou District, a typical region of the Qin-Ba mountainous area in the central part of China. The results show that the suitability distribution of living- production-ecological functions vary widely from the center to the periphery with altitude in Hengkou District; 22.03% of the land is at a risk of land use conflict. Among them, the high potential conflict areas account for 55.32%, and the conflicts between production and ecological lands (L2P1E1, L3P1E1) are the largest, which are located at the fringe of the central urban and ecologically dominant area. Therefore, it is necessary to adopt effective strategies to achieve a balance between the differential demands of land use. This research could better reflect the true situation of land use in ecologically sensitive mountainous areas and would provide theoretical and methodological support for the identification and prevention of potential LUCs. Full article

The Human Development Index (HDI) is a prevailing indicator to present the status and trend of sustainability of nations, hereby offers a valuable measurement on the Sustainable Development Goals (SDGs). Revealing the dynamics of the HDI of the Eastern Hemisphere countries is vital for measurement and evaluation of the human development process and revealing the spatial disparities and evolutionary characteristics of human development. However, the statistical data-based HDI, which is currently widely applied, has defects in terms of data availability and inconsistent statistical caliber. To tackle such an existing gap, we applied nighttime lights (NTL) data to reconstruct new HDI indicators named HDI_NTL and quantify the HDI_NTL at multispatial scales of Eastern Hemisphere countries during 1992–2013. Results showed that South Central Asia countries had the smallest discrepancies in HDI_NTL, while the largest was found in North Africa. The national-level HDI_NTL values in the Eastern Hemisphere ranged between 0.138 and 0.947 during 1992–2013. At the subnational scale, the distribution pattern of HDI_NTL was spatially clustered based on the results of spatial autocorrelation analysis. The evolutionary trajectory of subnational level HDI_NTL exhibited a decreasing and then increasing trend along the northwest to the southeast direction of Eastern Hemisphere. At the pixel scale, 93.52% of the grids showed an increasing trend in HDI_NTL, especially in the urban agglomerations of China and India. These results are essential for the ever-improvement of policy making to reduce HDI’s regional disparity and promote the continuous development of humankind’s living qualities. This study offers an improved HDI accounting method. It expects to extend the channel of HDI application, e.g., potential integration with environmental, physical, and socioeconomic data where the NTL data could present as well. Full article

Evapotranspiration (ET) is a vital part of the hydrological cycle and the water–energy balance. To explore the characteristics of five typical remote sensing evapotranspiration datasets and provide guidance for algorithm development, we used reconstructed evapotranspiration (Recon) data based on ground and GRACE satellite observations as a benchmark and evaluated five remote sensing datasets for 592 watersheds across the continental United States. The Global Land Evaporation Amsterdam Model (GLEAM) dataset (with bias and RMSE values of 23.18 mm/year and 106.10 mm/year, respectively), process-based land surface evapotranspiration/heat flux (P-LSH) dataset (bias = 22.94 mm/year and RMSE = 114.44 mm/year) and the Penman–Monteith–Leuning (PML) algorithm generated ET dataset (bias = −17.73 mm/year and RMSE = 108.97 mm/year) showed the better performance on a yearly scale, followed by the model tree ensemble (MTE) dataset (bias = 99.45 mm/year and RMSE = 141.32 mm/year) and the moderate-resolution imaging spectroradiometer (MODIS) dataset (bias = −106.71 mm/year and RMSE = 158.90 mm/year). The P-LSH dataset outperformed the other four ET datasets on a seasonal scale, especially from March to August. Both PML and MTE showed better overall accuracy and could accurately capture the spatial variability of evapotranspiration in arid regions. The P-LSH and GLEAM products were consistent with the Recon data in middle-value section. MODIS and MTE had larger bias and RMSE values on a yearly scale, whereby the MODIS and MTE datasets tended to underestimate and overestimate ET values in all the sections, respectively. In the future, the aim should be to reduce bias in the MODIS and MTE algorithms and further improve seasonality of the ET estimation in the GLEAM algorithm, while the estimation accuracy of the P-LSH and MODIS algorithms should be improved in arid regions. Our analysis suggests that combining artificial intelligence algorithms or data-driven algorithms and physical process algorithms will further improve the accuracy of ET estimation algorithms and the quality of ET datasets, as well as enhancing their capacity to be applied in different climate regions. Full article

The interferometric imaging radar altimeter (InIRA) aboard the Chinese Tiangong-2 space laboratory is the first spaceborne imaging radar working at low incidence angles. This study focuses on the retrieval of significant wave heights (SWHs) from InIRA data. The retrieved SWHs can be used for correcting the sea state bias of InIRA-derived sea surface heights and can supplement SWH products from other spaceborne sensors. First, we analyzed tilt, range bunching and velocity bunching wave modulations at low incidence angles, and we found clear dependencies between the SWH and two defined factors, range and azimuth integration, for ocean waves in the range and azimuth directions, respectively. These dependencies were further confirmed using InIRA measurements and collocated WaveWatch III (WW3) data. Then, an empirical orthogonal SWH model using the range and azimuth integration factors as model inputs was proposed. The model was segmented by the incidence angle, and the model coefficients were estimated by fitting the collocation at each incidence angle bin. Finally, the SWHs were retrieved from InIRA data using the proposed model. The retrievals were validated using both WW3 and altimeter (JASON2, JASON3, SARAL, and HY2A) SWHs. The validation with WW3 data shows a root mean square error (RMSE) of 0.43 m, while the average RMSE with all traditional altimeter data is 0.48 m. This indicates that the InIRA can be used to measure SWHs. Full article

The Global Precipitation Measurement mission is a major U.S.–Japan joint mission to understand the physics of the Earth’s global precipitation as a key component of its weather, climate, and hydrological systems. The core satellite carries a dual-precipitation radar and an advanced microwave imager which provide measurements to retrieve the drop size distribution (DSD) and rain rates using a Combined Radar-Radiometer Algorithm (CORRA). Our objective is to validate key assumptions and parameterizations in CORRA and enable improved estimation of precipitation products, especially in the middle-to-higher latitudes in both hemispheres. The DSD parameters and statistical relationships between DSD parameters and radar measurements are a central part of the rainfall retrieval algorithm, which is complicated by regimes where DSD measurements are abysmally sparse (over the open ocean). In view of this, we have assembled optical disdrometer datasets gathered by research vessels, ground stations, and aircrafts to simulate radar observables and validate the scattering lookup tables used in CORRA. The joint use of all DSD datasets spans a large range of drop concentrations and characteristic drop diameters. The scaling normalization of DSDs defines an intercept parameter N_W, which normalizes the concentrations, and a scaling diameter D_m, which compresses or stretches the diameter coordinate axis. A major finding of this study is that a single relationship between N_W and D_m, on average, unifies all datasets included, from stratocumulus to heavier rainfall regimes. A comparison with the N_W–D_m relation used as a constraint in versions 6 and 7 of CORRA highlights the scope for improvement of rainfall retrievals for small drops (D_m < 1 mm) and large drops (D_m > 2 mm). The normalized specific attenuation–reflectivity relationships used in the combined algorithm are also found to match well the equivalent relationships derived using DSDs from the three datasets, suggesting that the currently assumed lookup tables are not a major source of uncertainty in the combined algorithm rainfall estimates. Full article

When soils are used for a long period of time as arable land, their properties change. This can lead to soil degradation and loss of fertility, as well as other important soil biosphere functions. Obtaining data on the trends in arable soil conditions over large areas using traditional field survey methods is expensive and time-consuming. Currently, there are large archives of satellite data that can be used to monitor the status of arable soils. The analysis of changes in the color of the surface of arable chernozem soils of the Belgorod region, for the period from 1985 to the present, has been carried out based on the analysis of Landsat TM5 satellite data and information about the spectral reflectance of the soils of the region. It is found that, on most parts of arable lands of the region, the color of the soil surface has not changed significantly since 1985. Color changes were revealed on 11% of the analyzed area. The greatest changes are connected with the humus content and moisture content of soils. The three most probable reasons for the change of humus content in an arable horizon of soils are as follows: the dehumidification of soils during plowing; the reduction of the humus content due to water erosion; and the increase in humus content due to changes in the land-use system of the region in recent years. The change in soil moisture regime has mainly been found in arable lands in river valleys, most likely conditioned by the natural evolution of soils. Trends of increasing soil moisture are prevalent. The revealed regularities testify to the high stability of arable soils in the region during the last few decades. Full article

Spatial patterns in long-term average evapotranspiration (ET) represent a unique source of information for evaluating the spatial pattern performance of distributed hydrological models on a river basin to continental scale. This kind of model evaluation is getting increased attention, acknowledging the shortcomings of traditional aggregated or timeseries-based evaluations. A variety of satellite remote sensing (RS)-based ET estimates exist, covering a range of methods and resolutions. There is, therefore, a need to evaluate these estimates, not only in terms of temporal performance and similarity, but also in terms of long-term spatial patterns. The current study evaluates four RS-ET estimates at moderate resolution with respect to spatial patterns in comparison to two alternative continental-scale gridded ET estimates (water-balance ET and Budyko). To increase comparability, an empirical correction factor between clear sky and all-weather ET, based on eddy covariance data, is derived, which could be suitable for simple corrections of clear sky estimates. Three RS-ET estimates (MODIS16, TSEB and PT-JPL) and the Budyko method generally display similar spatial patterns both across the European domain (mean SPAEF = 0.41, range 0.25–0.61) and within river basins (mean SPAEF range 0.19–0.38), although the pattern similarity within river basins varies significantly across basins. In contrast, the WB-ET and PML_V2 produced very different spatial patterns. The similarity between different methods ranging over different combinations of water, energy, vegetation and land surface temperature constraints suggests that robust spatial patterns of ET can be achieved by combining several methods. Full article