Search

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]

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]

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]

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]

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]