Search

For very-high-resolution (VHR) remote sensing images with complex objects and rich textural information, multi-difference image fusion has been proven as an effective method to improve the performance of change detection. However, errors are superimposed during this process and a single spectral feature cannot fully utilize the correlation between pixels, resulting in low robustness. To overcome these problems and optimize the performance of multi-difference image fusion in change detection, we propose a novel multi-difference image fusion change detection method based on a visual attention model (VA-MDCD). First, we construct difference images using change vector analysis (CVA) and spectral gradient difference (SGD). Second, we use the visual attention model to calculate multiple color, intensity and orientation features of the difference images to obtain the difference saliency images. Third, we use the wavelet transform fusion algorithm to fuse two saliency images. Finally, we execute the OTSU threshold segmentation algorithm (OTSU) to obtain the final change detection map. To validate the effectiveness of VA-MDCD on VHR images, two datasets of Jilin 1 and Beijing 2 are selected for experiments. Compared with classical methods, the proposed method has a better performance with fewer missed alarms (MA) and false alarms (FA), which proves that the method has a strong robustness and generalization ability. The F-measure of the two datasets is 0.6671 and 0.7313, respectively. In addition, the results of ablation experiments confirm that the three feature extraction modules of the model all play a positive role. [ABSTRACT FROM AUTHOR]

On 6 February 2023 (UTC), an earthquake doublet, consisting of the Mw 7.8 Pazarcik earthquake and the Mw 7.5 Elbistan earthquake, struck south-central Turkey and northwestern Syria, which was the largest earthquake that occurred in Turkey since the 1939 Erzincan earthquake. The faulting model of this earthquake was estimated based on the coseismic InSAR and GPS displacements. In addition, the best-fitting coseismic faulting model indicates that both the Pazarcik earthquake and the Elbistan earthquake were controlled by predominated left-lateral strike-slip motion, with slip peaks of 9.7 m and 10.8 m, respectively. The Coulomb failure stress (CFS) change suggests that the Pazarcik earthquake has a positive effect in triggering the rupture of the seismogenic fault of the Elbistan earthquake. Furthermore, these two main shocks promoted the occurrence of the Mw 6.3 strong aftershock. Additionally, it is found that the 2023 Turkey-Syria earthquake doublet increased the rupture risk of the Puturge segment of the EAF fault and the northern segment of the Dead Sea Fault (DSF). It is crucial to note that the northern segment of the DSF has not experienced a large earthquake in several centuries, highlighting the need for heightened attention to the potential seismic hazard of this segment. Finally, a deformation zone adjacent to the DSF was identified, potentially attributed to the motion of a blind submarine fault. [ABSTRACT FROM AUTHOR]

The census and management of hazard-bearing entities, along with the integrity of data quality, form crucial foundations for disaster risk assessment and zoning. By addressing the challenge of feature confusion, prevalent in single remotely sensed image recognition methods, this paper introduces a novel method, Spatially Constrained Deep Learning (SCDL), that combines deep learning with spatial constraint strategies for the extraction of disaster-bearing bodies, focusing on dams as a typical example. The methodology involves the creation of a dam dataset using a database of dams, followed by the training of YOLOv5, Varifocal Net, Faster R-CNN, and Cascade R-CNN models. These models are trained separately, and highly confidential dam location information is extracted through parameter thresholding. Furthermore, three spatial constraint strategies are employed to mitigate the impact of other factors, particularly confusing features, in the background region. To assess the method's applicability and efficiency, Qinghai Province serves as the experimental area, with dam images from the Google Earth Pro database used as validation samples. The experimental results demonstrate that the recognition accuracy of SCDL reaches 94.73%, effectively addressing interference from background factors. Notably, the proposed method identifies six dams not recorded in the GOODD database, while also detecting six dams in the database that were previously unrecorded. Additionally, four dams misdirected in the database are corrected, contributing to the enhancement and supplementation of the global dam geo-reference database and providing robust support for disaster risk assessment. In conclusion, leveraging open geographic data products, the comprehensive framework presented in this paper, encompassing deep learning target detection technology and spatial constraint strategies, enables more efficient and accurate intelligent retrieval of disaster-bearing bodies, specifically dams. The findings offer valuable insights and inspiration for future advancements in related fields. [ABSTRACT FROM AUTHOR]

Sentinel-1 Terrain Observation by Progressive Scans (TOPS) data have been widely applied in earthquake studies due to their open-source policy, short revisit cycle and wide coverage. However, significant near-fault displacement gradients and the moderate azimuth resolution of TOPS data make achieving high-precision along-track measurements challenging, which prevents the generation of high-quality three-dimensional (3D) displacement maps. Here, we propose an integrated method to retrieve high-quality 3D displacements based on the differential interferometric SAR (DInSAR), burst-overlap interferometry (BOI), multiple-aperture InSAR (MAI) and pixel offset tracking (POT) techniques, which are achieved to use only two track Sentinel-1 TOPS data with different viewing geometries. The key step of this method is using a weighted fusion algorithm with the interpolated BOI-derived and MAI-derived 3D displacements. In a case study of the 2021 Maduo earthquake, the calculated root mean square errors (RMSEs) from global navigation satellite system (GNSS) data and the InSAR-derived 3D displacement fields were found to be 6.3, 5.8 and 1.7 cm in north–south, east–west and up–down components, respectively. Moreover, the slip model of the 2021 Maduo earthquake jointly estimated by DInSAR and BOI measurements indicates that this seismic event was dominated by sinistral strike-slip motion mixed with some dip-slip movements; the estimated seismic moment was 1.75 × 1020 Nm, corresponding to a Mw 7.44 event. [ABSTRACT FROM AUTHOR]

Urban building morphology has a significant impact on the urban thermal environment (UTE). The sky view factor (SVF) is an important structure index of buildings and combines height and density attributes. These factors have impact on the land surface temperature (LST). Thus, it is crucial to analyze the relationship between SVF and LST in different spatial-temporal scales. Therefore, we tried to use a building vector database to calculate the SVF, and we used remote sensing thermal infrared band to retrieve LST. Then, we analyzed the influence between SVF and LST in different spatial and temporal scales, and we analyzed the seasonal variation, day–night variation, and the impact of building height and density of the SVF–LST relationship. We selected the core built-up area of Beijing as the study area and analyzed the SVF–LST relationship in four periods in 2018. The temporal experimental results indicated that LST is higher in the obscured areas than in the open areas at nighttime. In winter, the maximum mean LST is in the open areas. The spatial experimental results indicate that the SVF and LST relationship is different in the low SVF region, with 30 m and 90 m pixel scale in the daytime. This may be the shadow cooling effect around the buildings. In addition, we discussed the effects of building height and shading on the SVF–LST relationship, and the experimental results show that the average shading ratio is the largest at 0.38 in the mid-rise building area in winter. [ABSTRACT FROM AUTHOR]

We here present an example of the 2016 Kumamoto earthquake with its coseismic surface deformation mapped by the ALOS-2 satellite both in the right- and left-looking observation modes. It provides the opportunity to reveal the coseismic surface deformation and to explore the performance of the unusual left-looking data in faulting model inversion. Firstly, three tracks (ascending and descending right-looking and descending left-looking) of ALOS PALSAR-2 images are used to extract the surface deformation fields. It suggests that the displacements measured by the descending left-looking InSAR coincide well with the ascending right-looking track observations. Then, the location and strike angle of the fault are determined from the SAR pixel offset-tracking technique. A complicated four-segment fault geometry is inferred for explaining the coseismic faulting of the Kumamoto earthquake due to the interpretation of derived deformation fields. Quantitative comparisons between models constrained by the right-looking only data and by joint right- and left-looking data suggest that left-looking InSAR could provide comparable constraints for geodetic modelling to right-looking InSAR. Furthermore, the slip model suggests that the series of events are dominated by the dextral strike-slip with some normal fault motions. The fault rupture initiates on the Hinagu fault segment and propagates from southwest to northeast along the Hinagu fault, then transforms to Futagawa fault with a slip maximum of 4.96 m, and finally ends up at ~7 km NW of the Aso caldera, with a rupture length of ~55 km. The talent of left-looking InSAR in surface deformation detection and coseismic faulting inversion indicates that left-looking InSAR can be effectively utilized in the investigation of the geologic hazards in the future, same as right-looking InSAR. [ABSTRACT FROM AUTHOR]

With rapid urbanization, the disposal and management of urban construction waste have become the main concerns of urban management. The distribution of urban construction waste is characterized by its wide range, irregularity, and ease of confusion with the surrounding ground objects, such as bare soil, buildings, and vegetation. Therefore, it is difficult to extract and identify information related to urban construction waste by using the traditional single spectral feature analysis method due to the problem of spectral confusion between construction waste and the surrounding ground objects, especially in the context of very-high-resolution (VHR) remote sensing images. Considering the multi-feature analysis method for VHR remote sensing images, we propose an optimal method that combines morphological indexing and hierarchical segmentation to extract the information on urban construction waste in VHR images. By comparing the differences between construction waste and the surrounding ground objects in terms of the spectrum, geometry, texture, and other features, we selected an optimal feature subset to improve the separability of the construction waste and other objects; then, we established a classification model of knowledge rules to achieve the rapid and accurate extraction of construction waste information. We also chose two experimental areas of Beijing to validate our algorithm. By using construction waste separability quality evaluation indexes, the identification accuracy of construction waste in the two study areas was determined to be 96.6% and 96.2%, the separability indexes of the construction waste and buildings reached 1.000, and the separability indexes of the construction waste and vegetation reached 1.000 and 0.818. The experimental results show that our method can accurately identify the exposed construction waste and construction waste covered with a dust screen, and it can effectively solve the problem of spectral confusion between the construction waste and the bare soil, buildings, and vegetation. [ABSTRACT FROM AUTHOR]

Over the 15 years of the Gravity Recovery and Climate Experiment (GRACE) mission, various data processing approaches were developed to derive time-series of global gravity fields based on sensor observations acquired from the two spacecrafts. In this paper, we compare GRACE-based mass anomalies provided by various processing groups against Global Navigation Satellite System (GNSS) station coordinate time-series and in-situ observations of ocean bottom pressure. In addition to the conventional GRACE-based global geopotential models from the main processing centers, we focus particularly on combined gravity field solutions generated within the Horizon2020 project European Gravity Service for Improved Emergency Management (EGSIEM). Although two validation techniques are fully independent from each other, it is demonstrated that they confirm each other to a large extent. Through the validation, we show that the EGSIEM combined long-term monthly solutions are comparable to CSR RL05 and ITSG2016, and better than the other three considered GRACE monthly solutions AIUB RL02, GFZ RL05a, and JPL RL05.1. Depending on the GNSS products, up to 25.6% mean Weighted Root-Mean-Square (WRMS) reduction is obtained when comparing GRACE to the ITRF2014 residuals over 236 GNSS stations. In addition, we also observe remarkable agreement at the annual period between GNSS and GRACE with up to 73% median WRMS reduction when comparing GRACE to the 312 EGSIEM-reprocessed GNSS time series. While the correspondence between GRACE and ocean bottom pressure data is overall much smaller due to lower signal to noise ratio over the oceans than over the continents, up to 50% agreement is found between them in some regions. The results fully confirm the conclusions found using GNSS. [ABSTRACT FROM AUTHOR]

Multi-source datasets, including the UAV-Lidar/Photogrammetry, InSAR, and field investigation data, have been used for revealing the complex surface displacement pattern and focal mechanism of the 2021 Maduo Mw 7.3 earthquake. First, a co-seismic surface deformation field was extracted from the Synthetic Aperture Radar (SAR) images captured by ALOS-2 and Sentinel-1 satellites. Second, the SAR pixel offset tracking results were adopted to detect the initial location of the seismogenic fault. Then, the Lidar digital elevation model with high spatial resolution and field investigation were employed to refine and verify the location of the seismogenic fault. It was found that bifurcated strike-slip rupture should account for the 2021 Maduo earthquake. As indicated by the estimated faulting model based on the InSAR data, the maximal fault slip was ~6.2 m, occurring in the southeast of the main seismogenic fault, and five remarkable slip concentrations controlled the surface displacement of the 2021 Maduo earthquake. Furthermore, the co- and post-seismic InSAR deformation, dilatation, shear strain, Coulomb failure stress, and aftershock sequence suggest that the co-seismic rupture of the two main seismogenic faults have triggered the aseismic slip along the Changmahe fault. Lastly, according to the Coulomb failure stress change due to the historical earthquakes and the 2021 Maduo earthquake, the 1937 M 7.8 earthquake predominantly controlled the Coulomb failure stress change along the Kunlun fault, and the Xidatan-Alake Lake and Maqin segments had a higher risk of future earthquake than the other segments. [ABSTRACT FROM AUTHOR]

Building change detection is a prominent topic in remote sensing applications. Scholars have proposed a variety of fully-convolutional-network-based change detection methods for high-resolution remote sensing images, achieving impressive results on several building datasets. However, existing methods cannot solve the problem of pseudo-changes caused by factors such as "same object with different spectrums" and "different objects with same spectrums" in high-resolution remote sensing images because their networks are constructed using simple similarity measures. To increase the ability of the model to resist pseudo-changes and improve detection accuracy, we propose an improved method based on fully convolutional network, called multitask difference-enhanced Siamese network (MDESNet) for building change detection in high-resolution remote sensing images. We improved its feature extraction ability by adding semantic constraints and effectively utilized features while improving its recognition performance. Furthermore, we proposed a similarity measure combining concatenation and difference, called the feature difference enhancement (FDE) module, and designed comparative experiments to demonstrate its effectiveness in resisting pseudo-changes. Using the building change detection dataset (BCDD), we demonstrate that our method outperforms other state-of-the-art change detection methods, achieving the highest F1-score (0.9124) and OA (0.9874), indicating its advantages for high-resolution remote sensing image building change detection tasks. [ABSTRACT FROM AUTHOR]

Mass redistribution within the Earth system deforms the surface elastically. Loading theory allows us to predict loading induced displacement anywhere on the Earth's surface using environmental loading models, e.g., Global Land Data Assimilation System. In addition, different publicly available loading products are available. However, there are differences among those products and the differences among the combinations of loading models cannot be ignored when precisions of better than 1 cm are required. Many scholars have applied these loading corrections to Global Navigation Satellite System (GNSS) time series from mainland China without considering or discussing the differences between the available models. Evaluating the effects of different loading products over this region is of paramount importance for accurately removing the loading signal. In this study, we investigate the performance of these different publicly available loading products on the scatter of GNSS time series from the Crustal Movement Observation Network of China. We concentrate on five different continental water storage loading models, six different non-tidal atmospheric loading models, and five different non-tidal oceanic loading models. We also investigate all the different combinations of loading products. The results show that the difference in RMS reduction can reach 20% in the vertical component depending on the loading correction applied. We then discuss the performance of different loading combinations and their effects on the noise characteristics of GNSS height time series and horizontal velocities. The results show that the loading products from NASA may be the best choice for corrections in mainland China. This conclusion could serve as an important reference for loading products users in this region. [ABSTRACT FROM AUTHOR]

This study presents a new approach for Urban Functional Zone (UFZ) mapping by integrating two-dimensional (2D) Urban Structure Parameters (USPs), three-dimensional (3D) USPs, and the spatial patterns of land covers, which can be divided into two steps. Firstly, we extracted various features, i.e., spectral, textural, geometrical features, and 3D USPs from very-high-resolution (VHR) images and light detection and ranging (LiDAR) point clouds. In addition, the multi-classifiers (MLCs), i.e., Random Forest, K-Nearest Neighbor, and Linear Discriminant Analysis classifiers were used to perform the land cover mapping by using the optimized features. Secondly, based on the land cover classification results, we extracted 2D and 3D USPs for different land covers and used MLCs to classify UFZs. Results for the northern part of Brooklyn, New York, USA, show that the approach yielded an excellent accuracy of UFZ mapping with an overall accuracy of 91.9%. Moreover, we have demonstrated that 3D USPs could considerably improve the classification accuracies of UFZs and land covers by 6.4% and 3.0%, respectively. [ABSTRACT FROM AUTHOR]

Mastering the spatial and temporal differences of ENSO (EI Niño-Southern Oscillation) and MJO (Madden–Julian Oscillation) and their influence on drought is very important for accurately monitoring and forecasting drought. In this study, spatiotemporal characteristics and variability of the impact of ENSO and MJO on drought were analyzed from the perspectives of meteorological drought and agricultural drought through temporal and spatial correlation analyses of China's 48 eco-geographical regions. The results show a strong correlation between drought and ENSO and MJO in general. The spatial correlation coefficients are different, and the response of extreme events varies in different regions. The influence of ENSO and MJO on agricultural drought is higher than that on meteorological drought. ENSO and MJO have a considerable influence on agricultural drought in regions such as the Qinghai-Tibet Plateau and Xinjiang, with the highest correlation coefficient of 0.72. A significant influence of ENSO and MJO on meteorological drought was found in the Jiangnan region with the highest correlation coefficient of 0.40. In addition, agricultural drought shows a significant time lag in response to ENSO events. When the lag time is six months, the time series presents the highest correlation coefficient with the mean value of the correlation coefficient reaching 0.38 and the maximum value reaching 0.75. This research is of great significance for understanding the spatiotemporal correlation between climate patterns and drought on a large regional scale and it provides further insights into the teleconnection mechanisms of drought. [ABSTRACT FROM AUTHOR]

Machine learning has been successfully used for object recognition within images. Due to the complexity of the spectrum and texture of construction and demolition waste (C&DW), it is difficult to construct an automatic identification method for C&DW based on machine learning and remote sensing data sources. Machine learning includes many types of algorithms; however, different algorithms and parameters have different identification effects on C&DW. Exploring the optimal method for automatic remote sensing identification of C&DW is an important approach for the intelligent supervision of C&DW. This study investigates the megacity of Beijing, which is facing high risk of C&DW pollution. To improve the classification accuracy of C&DW, buildings, vegetation, water, and crops were selected as comparative training samples based on the Google Earth Engine (GEE), and Sentinel-2 was used as the data source. Three classification methods of typical machine learning algorithms (classification and regression trees (CART), random forest (RF), and support vector machine (SVM)) were selected to classify the C&DW from remote sensing images. Using empirical methods, the experimental trial method, and the grid search method, the optimal parameterization scheme of the three classification methods was studied to determine the optimal method of remote sensing identification of C&DW based on machine learning. Through accuracy evaluation and ground verification, the overall recognition accuracies of CART, RF, and SVM for C&DW were 73.12%, 98.05%, and 85.62%, respectively, under the optimal parameterization scheme determined in this study. Among these algorithms, RF was a better C&DW identification method than were CART and SVM when the number of decision trees was 50. This study explores the robust machine learning method for automatic remote sensing identification of C&DW and provides a scientific basis for intelligent supervision and resource utilization of C&DW. [ABSTRACT FROM AUTHOR]

River water discharge (WD) is an essential component when monitoring a regional hydrological cycle. It is expressed in terms of surface runoff (R) when a unit of river basin surface area is considered. To compensate for the decreasing number of hydrological stations, remotely-sensed WD estimation has been widely promoted over the past two decades, due to its global coverage. Previously, remotely-sensed WD was reconstructed either by correlating nearby remotely-sensed surface responses (e.g., indices and hydraulic variables) with ground-based WD observations or by applying water balance formulations, in terms of R, over an entire river basin, assisted by hydrological modeling data. In contrast, the feasibility of using remotely-sensed hydrological variables (RSHVs) and their standardized forms together with water balance representations (WBR) obtained from the river upstream to reconstruct estuarine R for an entire basin, has been rarely investigated. Therefore, our study aimed to construct a correlative relationship between the estuarine observed R and the upstream, spatially averaged RSHVs, together with their standardized forms and WBR, for the Mekong River basin, using estuarine R reconstructions, at a monthly temporal scale. We found that the reconstructed R derived from the upstream, spatially averaged RSHVs agreed well with the observed R, which was also comparable to that calculated using traditional remote sensing data (RSD). Better performance was achieved using spatially averaged, standardized RSHVs, which should be potentially attributable to spatially integrated information and the ability to partly bypass systematic biases by both human (e.g., dam operation) and environmental effects in a standardized form. Comparison of the R reconstructed using the upstream, spatially averaged, standardized RSHVs with that reconstructed from the traditional RSD, against the observed R, revealed a Pearson correlation coefficient (PCC) above 0.91 and below 0.81, a root-mean-squares error (RMSE) below 6.1 mm and above 8.5 mm, and a Nash–Sutcliffe model efficiency coefficient (NSE) above 0.823 and below 0.657, respectively. In terms of the standardized water balance representation (SWBR), the reconstructed R yielded the best performance, with a PCC above 0.92, an RMSE below 5.9 mm, and an NSE above 0.838. External assessment demonstrated similar results. This finding indicated that the standardized RSHVs, in particular its water balance representations, could lead to further improvement in estuarine R reconstructions for river basins affected by various systematic influences. Comparison between hydrological stations at the Mekong River Delta entrance and near the estuary mouth revealed tidally-induced backwater effects on the estimated R, with an RMSE difference of 4–5 mm (equivalent to 9–11% relative error). [ABSTRACT FROM AUTHOR]

The ascending and descending InSAR deformations derived from ALOS-2 and Sentinel-1 satellite SAR images and GPS displacements are used to estimate the fault model of the 2018 Mw 6.4 Hualien earthquake. The sinistral strike-slip fault dipping to the west with a high dip angle of 89.4° and a rake angle of 201.7° is considered as the seismogenic fault of this event. This seismogenic fault also triggered the ruptures of the Milun fault, which dips to the east with a dip angle of ~72°, and an unknown west-dipping fault with a dip angle of 85.2°. Two predicted faulting models indicate that the InSAR deformation fields include more postseismic slip than those of the GPS data. The north segment of the Milun fault and west-dipping fault have been triggered by the rupture of the seismogenic fault, but the postseismic slip occurred only in the south segment of the Milun fault. The InSAR-derived co-seismic and postseismic faulting model suggests that the significant slip concentrates at depths of 2.4–15.0 km of the main fault, 0.0–14.0 km of the Milun fault. Only minor slip is detected on the west-dipping fault. The maximum fault slip of ca. 2.1 m is located at the depth of ca. 2.4 km under the Meilun Tableland. The Coulomb failure stress (CFS) change calculated by the co-seismic and postseismic faulting model shows that there is a significant CFS increase in the east of the south segment of the Milun fault, but few of the aftershocks occur in this area, which indicates a high risk of future seismic hazard. [ABSTRACT FROM AUTHOR]