Search

Abstract An earthquake with a moment magnitude of 7.5 (Mw) struck the northern Noto Peninsula, Ishikawa Prefecture, Japan, at 16:10 local time on January 1, 2024. This earthquake triggered a tsunami that propagated along the coastline of Ishikawa, Toyama, and Niigata Prefectures facing the Sea of Japan and significantly damaged coastal communities and infrastructure. Approximately 70 researchers from 23 universities or other institutes throughout Japan formed a joint research group to conduct a post-tsunami survey along a 340 km stretch of the coast. Based on the watermarks and traces of the tsunami, the inundation and run-up heights were surveyed using total stations, automatic optical levels, laser range finders, and a real-time kinematic (RTK) Global Navigation Satellite System (GNSS). The tidal correction was adjusted using astronomical tidal tables. In total, 303 survey records have been compiled, generating the NP2024TS (Noto Peninsula 2024 Tsunami Survey) dataset. This dataset provides comprehensive information on the inundation and run-up heights of the tsunami, which is useful for understanding tsunami characteristics and validating numerical tsunami models.

Synthetic aperture radar (SAR) imagery is indispensable for acquiring a comprehensive, large-scale topographical perspective of the Earth's surface, facilitating the evaluation of diverse scenarios spanning various events. However, integrating SAR imagery with ground truth data poses a formidable challenge, primarily due to the complex interactions of electromagnetic radar waves with the intricate features of the surface, characterized by multiple scatterings. This study seeks to bridge these disparities between authentic SAR and simulated by applying ray tracing SAR simulations within a meticulously controlled environment. Our investigation into SAR simulation focuses on evaluating the accuracy, precision, and reliability of theoretical models in representing surface features. Through ray tracing SAR simulations, we aim to understand better the complex interactions of electromagnetic radar waves with surface features, ultimately advancing our ability to interpret SAR imagery in diverse scenarios. The research findings indicate a robust correlation between the physical parameters of the surface and the backscattering levels observed in SAR simulations, underscoring the critical dependence of simulation accuracy on the granularity of the 3-D models employed. Moreover, this study delves into the consistency of modeling approaches in accurately replicating the assessment of building damage in postdisaster contexts, providing insights into the potential enhancements in disaster management and response strategies. The implications of these findings not only advance the understanding of SAR simulation fidelity but also pave the way for significant improvements in the application of SAR technology for disaster assessment and environmental monitoring.

Integrated with remote sensing technology, deep learning has been increasingly used for rapid damage assessment. Despite reportedly having high accuracy, the approach requires numerous samples to maintain its performance. However, in the emergency response phase, training samples are often unavailable. Since no ground truth data is available, deep learning models cannot be trained for this specific situation and, thus, have to be applied to unseen data. Previous research has implemented transfer learning techniques to solve data unavailability. However, many studies do not accurately reflect the rapid damage mapping in real-world scenarios. This study illustrates the use of Earth observation and deep learning technologies in predicting damage in realistic emergency response settings. To this aim, we conducted extensive experiments using historical data to find the best model by examining multiple neural network models and loss functions. Then, we evaluated the performance of the best model for predicting building damage due to two different disasters, the 2011 Tohoku Tsunami and the 2023 Türkiye–Syria Earthquake, which were independent of the training samples. We found that a transformer-based model with a combined cross-entropy loss (CEL) and focal loss generates the highest scoring values. The testing on both unseen sites illustrates that the model can perform well in no-damage and destroyed classes. However, the scores dropped in the middle class. We also compared our transformer-based approach with other state-of-the-art models, specifically the xView-2 winning solution. The results show that the transformer-based models have stable generalization toward multiclass classification and multiresolution imagery.

Abstract Tsunami fragility functions (TFF) are statistical models that relate a tsunami intensity measure to a given building damage state, expressed as cumulative probability. Advances in computational and data retrieval speeds, coupled with novel deep learning applications to disaster science, have shifted research focus away from statistical estimators. TFFs offer a “disaster signature” with comparative value, though these models are seldom applied to generate damage estimates. With applicability in mind, we challenge this notion and investigate a portion of TFF literature, selecting three TFFs and two application methodologies to generate a building damage estimation baseline. Further, we propose a simple machine learning method, trained on physical parameters inspired by, but expanded beyond, TFF intensity measures. We test these three methods on the 2011 Ishinomaki dataset after the Great East Japan Earthquake and Tsunami in both binary and multi-class cases. We explore: (1) the quality of building damage estimation using TFF application methods; (2) whether TFF can generalize to out-of-domain building damage datasets; (3) a novel machine learning approach to perform the same task. Our findings suggest that: both TFF methods and our model have the potential to achieve good binary results; TFF methods struggle with multiple classes and out-of-domain tasks, while our proposed method appears to generalize better.

Abstract In this study, we present an optimization method for determining a cost‐effective sparse configuration for tsunami gauges to realize the reconstruction of high‐resolution wave height distribution throughout the target region based on the concept of super‐resolution. This optimization method consists of three procedures. First, we generate time series data of tsunami wave heights at synthetic gauges by conducting numerical simulations of various earthquake and tsunami scenarios at the target site. Next, we apply proper orthogonal decomposition to the synthetic tsunami data to extract the spatial features of the wave height distribution. Finally, according to these spatial features, an optimization process is performed to determine a sparse configuration of synthetic gauges. In the optimization, the optimal gauges are sequentially selected from the set of synthetic gauges to reconstruct the wave height distribution with the highest accuracy. Targeting hypothetical Nankai Trough earthquakes and tsunamis, we determine the optimal configuration near Shikoku and demonstrate the wave height reconstruction capability of the approach by comparing the performance of networks with optimally and randomly placed gauges. The results indicate that coastal gauges contribute more to improving the reconstruction accuracy and that a configuration with 21 optimal gauges has satisfactory performance. In addition, we assess the performance of the existing NOWPHAS network installed in the Shikoku region and find that the reconstruction performance of the existing network is equivalent to that of the optimal gauge network.

Abstract This study proposes a new method for the uncertainty estimation of coseismic slip distribution on the plate interface deduced from real-time global navigation satellite system (GNSS) data and explores its application for tsunami inundation prediction. Jointly developed by the Geospatial Information Authority of Japan and Tohoku University, REGARD (REal-time GEONET Analysis system for Rapid Deformation monitoring) estimates coseismic fault models (a single rectangular fault model and slip distribution model) in real time to support tsunami prediction. The estimated results are adopted as part of the Disaster Information System, which is used by the Cabinet Office of the Government of Japan to assess tsunami inundation and damage. However, the REGARD system currently struggles to estimate the quantitative uncertainty of the estimated result, although the obtained result should contain both observation and modeling errors caused by the model settings. Understanding such quantitative uncertainties based on the input data is essential for utilizing this resource for disaster response. We developed an algorithm that estimates the coseismic slip distribution and its uncertainties using Markov chain Monte Carlo methods. We focused on the Nankai Trough of southwest Japan, where megathrust earthquakes have repeatedly occurred, and used simulation data to assume a Hoei-type earthquake. We divided the 2951 rectangular subfaults on the plate interface and designed a multistage sampling flow with stepwise perturbation groups. As a result, we successfully estimated the slip distribution and its uncertainty at the 95% confidence interval of the posterior probability density function. Furthermore, we developed a new visualization procedure that shows the risk of tsunami inundation and the probability on a map. Under the algorithm, we regarded the Markov chain Monte Carlo samples as individual fault models and clustered them using the k-means approach to obtain different tsunami source scenarios. We then calculated the parallel tsunami inundations and integrated the results on the map. This map, which expresses the uncertainties of tsunami inundation caused by uncertainties in the coseismic fault estimation, offers quantitative and real time insights into possible worst-case scenarios. Graphical Abstract

Floods increase every year worldwide, and prompt information about the affected areas is essential for early disaster response. There has been extensive development in applying remote sensing data to identify floods. In fact, remote sensing data are the only tool to identify the extent of large-scale floods within hours after their occurrence. However, few studies have addressed methods to estimate inundation depth. Inundation depth can be advantageous for identifying areas where people may need assistance during evacuation and estimating damage loss. We present a practical application of sparse representation that integrates a synthetic aperture radar-based flood binary map with a digital elevation model to estimate inundation depths. We assume that the floodwaters can be modeled as a combination of water bodies at a state of rest. A dictionary of water bodies computed under potential inundation levels is constructed from the digital elevation model. Then, the actual flood extent is represented as a sparse linear combination of the water body dictionary. The inundation depth can be estimated because each water body from the linear combination is associated with an inundation level. To assess our proposed procedure, we computed the inundation depth of the flood in the town of Mabi, Okayama Prefecture, produced during the 2018 heavy rainfall. An average absolute value difference of about 60 cm between our results and a field survey performed by a third party was observed. Two other floods produced by the 2019 Hagibis typhoon were analyzed to illustrate the relevant information that can provide inundation depths.

The current study investigated the probable impact from a tsunami to a populated area located along the northwest ecuadorian coast, specifically in the key oil-industrial city of esmeraldas. a numerical tsunami simulation was performed considering the seismological and tectonic aspects of the area. The damage probability was calculated using fragility functions (ffs). Briefly, 16 cases of source models with slightly different fault parameters were tested, where one was selected as the worst scenario of tsunami inundation. This scenario was a hypothetic earthquake case (mw 8.7) located in front of esmeraldas city, approximately 100 km offshore along the ecuador—colombia trench, with three shallow fault segments (top depth of 10 km), a strike aligned with the trench axis, a middle dip angle of 28°, assuming large slips of 5 to 15 m, and a rake angle of 90°. The results from the numerical simulation were comparable to a similar study previously conducted and with those of historically documented data. The tsunami damage estimation using FFs resulted in estimated damages of 50% and 44% in exposed buildings and population, respectively. Results also showed that the most impacted areas were located next to the coastal shoreline and river. tourism, oil exports, and port activities, in general, would be affected in this scenario; thus, compromising important industries that support the national budget. Results from this study would assist in designing or improving tsunami risk reduction strategies, disaster management, use of coastal zones, and planning better policies.

Abstract Emergency responders require accurate and comprehensive data to make informed decisions. Moreover, the data should be acquired and analyzed swiftly to ensure an efficient response. One of the tasks at hand post-disaster is damage assessment within the impacted areas. In particular, building damage should be assessed to account for possible casualties, and displaced populations, to estimate long-term shelter capacities, and to assess the damage to services that depend on essential infrastructure (e.g. hospitals, schools, etc.). Remote sensing techniques, including satellite imagery, can be used to gathering such information so that the overall damage can be assessed. However, specific points of interest among the damaged buildings need higher resolution images and detailed information to assess the damage situation. These areas can be further assessed through unmanned aerial vehicles and 3D model reconstruction. This paper presents a multi-UAV coverage path planning method for the 3D reconstruction of postdisaster damaged buildings. The methodology has been implemented in NetLogo3D, a multi-agent model environment, and tested in a virtual built environment in Unity3D. The proposed method generates camera location points surrounding targeted damaged buildings. These camera location points are filtered to avoid collision and then sorted using the K-means or the Fuzzy C-means methods. After clustering camera location points and allocating these to each UAV unit, a route optimization process is conducted as a multiple traveling salesman problem. Final corrections are made to paths to avoid obstacles and give a resulting path for each UAV that balances the flight distance and time. The paper presents the details of the model and methodologies, and an examination of the texture resolution obtained from the proposed method and the conventional overhead flight with the nadir-looking method used in 3D mappings. The algorithm outperforms the conventional method in terms of the quality of the generated 3D model.

Abstract Rapid estimation of the coseismic fault model for medium-to-large-sized earthquakes is key for disaster response. To estimate the coseismic fault model for large earthquakes, the Geospatial Information Authority of Japan and Tohoku University have jointly developed a real-time GEONET analysis system for rapid deformation monitoring (REGARD). REGARD can estimate the single rectangular fault model and slip distribution along the assumed plate interface. The single rectangular fault model is useful as a first-order approximation of a medium-to-large earthquake. However, in its estimation, it is difficult to obtain accurate results for model parameters due to the strong effect of initial values. To solve this problem, this study proposes a new method to estimate the coseismic fault model and model uncertainties in real time based on the Bayesian inversion approach using the Markov Chain Monte Carlo (MCMC) method. The MCMC approach is computationally expensive and hyperparameters should be defined in advance via trial and error. The sampling efficiency was improved using a parallel tempering method, and an automatic definition method for hyperparameters was developed for real-time use. The calculation time was within 30 s for 1 × 106 samples using a typical single LINUX server, which can implement real-time analysis, similar to REGARD. The reliability of the developed method was evaluated using data from recent earthquakes (2016 Kumamoto and 2019 Yamagata-Oki earthquakes). Simulations of the earthquakes in the Sea of Japan were also conducted exhaustively. The results showed an advantage over the maximum likelihood approach with a priori information, which has initial value dependence in nonlinear problems. In terms of application to data with a small signal-to-noise ratio, the results suggest the possibility of using several conjugate fault models. There is a tradeoff between the fault area and slip amount, especially for offshore earthquakes, which means that quantification of the uncertainty enables us to evaluate the reliability of the fault model estimation results in real time.

When a large-scale flood disaster occurs, it is important to identify the flood areas in a short time in order to effectively support the affected areas afterwards. Synthetic Aperture Radar (SAR) is promising for flood detection. A number of change detection methods have been proposed to detect flooded areas with pre- and post-event SAR data. However, it remains difficult to detect flooded areas in built-up areas due to the complicated scattering of microwaves. To solve this issue, in this paper we propose the idea of analyzing the local changes in pre- and post-event SAR data as well as the larger-scale changes, which may improve accuracy for detecting floods in built-up areas. Therefore, we aimed at evaluating the effectiveness of multi-scale SAR analysis for flood detection in built-up areas using ALOS-2/PALSAR-2 data. First, several features were determined by calculating standard deviation images, difference images, and correlation coefficient images with several sizes of kernels. Then, segmentation on both small and large scales was applied to the correlation coefficient image and calculated explanatory variables with the features at each segment. Finally, machine learning models were tested for their flood detection performance in built-up areas by comparing a small-scale approach and multi-scale approach. Ten-fold cross-validation was used to validate the model, showing that highest accuracy was offered by the AdaBoost model, which improved the F1 Score from 0.89 in the small-scale analysis to 0.98 in the multi-scale analysis. The main contribution of this manuscript is that, from our results, it can be inferred that multi-scale analysis shows better performance in the quantitative detection of floods in built-up areas.

In large-scale disasters, such as earthquakes and tsunamis, quick and sufficient transportation of emergency relief supplies is required. Logistics activities conducted to quickly provide appropriate aid supplies (relief goods) to people affected by disasters are known as humanitarian logistics (HL), and play an important role in terms of saving the lives of those affected. In the previous last-mile distribution of HL, supplies are transported by trucks and helicopters, but these transport methods are sometimes not feasible. Therefore, the use of unmanned aerial vehicles (UAVs) to transport supplies is attracting attention due to their convenience regardless of the disaster conditions. However, existing transportation planning that utilizes UAVs may not meet some of the requirements for post-disaster transport of supplies. Equitable distribution of supplies among affected shelters is particularly important in a crisis situation, but it has not been a major consideration in the logistics of UAVs in the existing study. Therefore, this study proposes transportation planning by introducing three crucial performance metrics: (1) the rapidity of supplies, (2) the urgency of supplies, and (3) the equity of supply amounts. We formulated the routing problem of UAVs as the multi-objective, multi-trip, multi-item, and multi-UAV problem, and optimize the problem with Q-learning (QL), one of the reinforcement learning methods. We performed reinforcement learning for multiple cases with different rewards and quantitatively evaluated the transportation of each countermeasure by comparing them. The results suggest that the model improved the stability of the supply of emergency relief supplies to all evacuation centers when compared to other models.

Abstract The tsunami source of the 2016 Fukushima Earthquake, which was generated by a normal faulting earthquake mechanism, is estimated by inverting the tsunami waveforms that were recorded by seven tide gauge stations and two wave gauge stations along the north Pacific coast of Japan. Two fault models based on different available moment tensor solutions were employed, and their locations were constrained by using the reverse tsunami travel time from the stations to the epicenter. The comparison of the two fault slip models showed that the fault model with a strike = 49°, dip = 35°, and rake = −89° more accurately simulated the observed tsunami data. This fault model estimated a fault area of 40 km $$\times$$ × 32 km. The largest slip was estimated as 4.66 m at a 6.09 km depth, larger slips also concentrated between depths of 6.06 and 10.68 km, and located southwest of the epicenter. Assuming a rigidity of $$2.7\times 10^{10}$$ 2.7 × 10 10 N/m $$^2$$ 2 , the estimated moment magnitude was $$3.35\times 10^{19}$$ 3.35 × 10 19 Nm (equivalent to Mw = 6.95). In addition, a comparison of nonlinear tsunami simulations using finer bathymetry around Sendai Coast verified that the above fault slip model could better reproduce the tsunami features observed at Sendai Port and its surroundings. Finally, we analyzed the nonlinear tsunami computed from our best fault slip model. Our simulations also corroborated the height of the secondary wave amplitude observed at Sendai Port, which was caused by the reflected tsunami waves from the Fukushima coast, as described in previous studies. Furthermore, we found that the initial positive wave recorded inside Sendai Bay resulted from the addition of the initial incoming wave and the tsunami wave reflected off Sendai Coast, between Natori River and Sendai Port.

Identifying permanent water and temporary water in flood disasters efficiently has mainly relied on change detection method from multi-temporal remote sensing imageries, but estimating the water type in flood disaster events from only post-flood remote sensing imageries still remains challenging. Research progress in recent years has demonstrated the excellent potential of multi-source data fusion and deep learning algorithms in improving flood detection, while this field has only been studied initially due to the lack of large-scale labelled remote sensing images of flood events. Here, we present new deep learning algorithms and a multi-source data fusion driven flood inundation mapping approach by leveraging a large-scale publicly available Sen1Flood11 dataset consisting of roughly 4831 labelled Sentinel-1 SAR and Sentinel-2 optical imagery gathered from flood events worldwide in recent years. Specifically, we proposed an automatic segmentation method for surface water, permanent water, and temporary water identification, and all tasks share the same convolutional neural network architecture. We utilize focal loss to deal with the class (water/non-water) imbalance problem. Thorough ablation experiments and analysis confirmed the effectiveness of various proposed designs. In comparison experiments, the method proposed in this paper is superior to other classical models. Our model achieves a mean Intersection over Union (mIoU) of 52.99%, Intersection over Union (IoU) of 52.30%, and Overall Accuracy (OA) of 92.81% on the Sen1Flood11 test set. On the Sen1Flood11 Bolivia test set, our model also achieves very high mIoU (47.88%), IoU (76.74%), and OA (95.59%) and shows good generalization ability.

When flooding occurs, Synthetic Aperture Radar (SAR) imagery is often used to identify flood extent and the affected buildings for two reasons: (i) for early disaster response, such as rescue operations, and (ii) for flood risk analysis. Furthermore, the application of machine learning has been valuable for the identification of damaged buildings. However, the performance of machine learning depends on the number and quality of training data, which is scarce in the aftermath of a large scale disaster. To address this issue, we propose the use of fragmentary but reliable news media photographs at the time of a disaster and use them to detect the whole extent of the flooded buildings. As an experimental test, the flood occurred in the town of Mabi, Japan, in 2018 is used. Five hand-engineered features were extracted from SAR images acquired before and after the disaster. The training data were collected based on news photos. The date release of the photographs were considered to assess the potential role of news information as a source of training data. Then, a discriminant function was calibrated using the training data and the support vector machine method. We found that news information taken within 24 h of a disaster can classify flooded and nonflooded buildings with about 80% accuracy. The results were also compared with a standard unsupervised learning method and confirmed that training data generated from news media photographs improves the accuracy obtained from unsupervised classification methods. We also provide a discussion on the potential role of news media as a source of reliable information to be used as training data and other activities associated to early disaster response.

Most mainstream research on assessing building damage using satellite imagery is based on scattered datasets and lacks unified standards and methods to quantify and compare the performance of different models. To mitigate these problems, the present study develops a novel end-to-end benchmark model, termed the pyramid pooling module semi-Siamese network (PPM-SSNet), based on a large-scale xBD satellite imagery dataset. The high precision of the proposed model is achieved by adding residual blocks with dilated convolution and squeeze-and-excitation blocks into the network. Simultaneously, the highly automated process of satellite imagery input and damage classification result output is reached by employing concurrent learned attention mechanisms through a semi-Siamese network for end-to-end input and output purposes. Our proposed method achieves F1 scores of 0.90, 0.41, 0.65, and 0.70 for the undamaged, minor-damaged, major-damaged, and destroyed building classes, respectively. From the perspective of end-to-end methods, the ablation experiments and comparative analysis confirm the effectiveness and originality of the PPM-SSNet method. Finally, the consistent prediction results of our model for data from the 2011 Tohoku Earthquake verify the high performance of our model in terms of the domain shift problem, which implies that it is effective for evaluating future disasters.

Earth Observation satellite imaging helps building diagnosis during a disaster. Several models are put forward on the xBD dataset, which can be divided into two levels: the building level and the pixel level. Models from two levels evolve into several versions that will be reviewed in this paper. There are four key challenges hindering researchers from moving forward on this task, and this paper tries to give technical solutions. First, metrics on different levels could not be compared directly. We put forward a fairer metric and give a method to convert between metrics of two levels. Secondly, drone images may be another important source, but drone data may have only a post-disaster image. This paper shows and compares methods of directly detecting and generating. Thirdly, the class imbalance is a typical feature of the xBD dataset and leads to a bad F1 score for minor damage and major damage. This paper provides four specific data resampling strategies, which are Main-Label Over-Sampling (MLOS), Discrimination After Cropping (DAC), Dilation of Area with Minority (DAM) and Synthetic Minority Over-Sampling Technique (SMOTE), as well as cost-sensitive re-weighting schemes. Fourthly, faster prediction meets the need for a real-time situation. This paper recommends three specific methods, feature-map subtraction, parameter sharing, and knowledge distillation. Finally, we developed our AI-driven Damage Diagnose Platform (ADDP). This paper introduces the structure of ADDP and technical details. Customized settings, interface preview, and upload and download satellite images are major services our platform provides.

Applications of machine learning on remote sensing data appear to be endless. Its use in damage identification for early response in the aftermath of a large-scale disaster has a specific issue. The collection of training data right after a disaster is costly, time-consuming, and many times impossible. This study analyzes a possible solution to the referred issue: the collection of training data from past disaster events to calibrate a discriminant function. Then the identification of affected areas in a current disaster can be performed in near real-time. The performance of a supervised machine learning classifier to learn from training data collected from the 2018 heavy rainfall at Okayama Prefecture, Japan, and to identify floods due to the typhoon Hagibis on 12 October 2019 at eastern Japan is reported in this paper. The results show a moderate agreement with flood maps provided by local governments and public institutions, and support the assumption that previous disaster information can be used to identify a current disaster in near-real time.

Wetlands provide society with a myriad of ecosystem services, such as water storage, food sources, and flood control. The ecosystem services provided by a wetland are largely dependent on its hydrological dynamics. Constant monitoring of the spatial extent of water surfaces and the duration of flooding of a wetland is necessary to understand the impact of drought on the ecosystem services a wetland provides. Synthetic aperture radar (SAR) has the potential to reveal wetland dynamics. Multitemporal SAR image analysis for wetland monitoring has been extensively studied based on the advances of modern SAR missions. Unfortunately, most previous studies utilized monopath SAR images, which result in limited success. Tracking changes in individual wetlands remains a challenging task because several environmental factors, such as wind-roughened water, degrade image quality. In general, the data acquisition frequency is an important factor in time series analysis. We propose a Gaussian process-based temporal interpolation (GPTI) method that enables the synergistic use of SAR images taken from multiple paths. The proposed model is applied to a series of Sentinel-1 images capturing wetlands in Okanogan County, Washington State. Our experimental analysis demonstrates that the multiple path analysis based on the proposed method can extract seasonal changes more accurately than a single path analysis.

Tsunamis are rare events compared with the other natural disasters, but once it happens, it can be extremely devastating to the coastal communities. Extensive inland penetration of tsunamis may cause the difficulties of understanding its impact in the aftermath of its generation. Therefore the social needs to technologies of detecting the wide impact of great tsunamis have been increased. Recent advances of remote sensing and technologies of image analysis meet the above needs and lead to more rapid and efficient understanding of tsunami affected areas. This paper provides a review of how remote sensing methods have developed to contribute to post-tsunami disaster response. The evaluations in the performances of the remote sensing methods are discussed according to the needs of tsunami disaster response with future perspective.

The rapid and accurate mapping of large-scale landslides and other mass movement disasters is crucial for prompt disaster response efforts and immediate recovery planning. As such, remote sensing information, especially from synthetic aperture radar (SAR) sensors, has significant advantages over cloud-covered optical imagery and conventional field survey campaigns. In this work, we introduced an integrated pixel-object image analysis framework for landslide recognition using SAR data. The robustness of our proposed methodology was demonstrated by mapping two different source-induced landslide events, namely, the debris flows following the torrential rainfall that fell over Hiroshima, Japan, in early July 2018 and the coseismic landslide that followed the 2018 Mw6.7 Hokkaido earthquake. For both events, only a pair of SAR images acquired before and after each disaster by the Advanced Land Observing Satellite-2 (ALOS-2) was used. Additional information, such as digital elevation model (DEM) and land cover information, was employed only to constrain the damage detected in the affected areas. We verified the accuracy of our method by comparing it with the available reference data. The detection results showed an acceptable correlation with the reference data in terms of the locations of damage. Numerical evaluations indicated that our methodology could detect landslides with an accuracy exceeding 80%. In addition, the kappa coefficients for the Hiroshima and Hokkaido events were 0.30 and 0.47, respectively.

Synthetic aperture radar (SAR) images have been used to map flooded areas with great success. Flooded areas are often identified by detecting changes between a pair of images recorded before and after a certain flood. During the 2018 Western Japan Floods, the change detection method generated significant misclassifications for agricultural targets. To evaluate whether such a situation could be repeated in future events, this paper examines and identifies the causes of the misclassifications. We concluded that the errors occurred because of the following. (i) The use of only a single pair of SAR images from before and after the floods. (ii) The unawareness of the dynamics of the backscattering intensity through time in agricultural areas. (iii) The effect of the wavelength on agricultural targets. Furthermore, it is highly probable that such conditions might occur in future events. Our conclusions are supported by a field survey of 35 paddy fields located within the misclassified area and the analysis of Sentinel-1 time series data. In addition, in this paper, we propose a new parameter, which we named “conditional coherence”, that can be of help to overcome the referred issue. The new parameter is based on the physical mechanism of the backscattering on flooded and non-flooded agricultural targets. The performance of the conditional coherence as an input of discriminant functions to identify flooded and non-flooded agricultural targets is reported as well.

This work presents a detailed analysis of building damage recognition, employing multi-source data fusion and ensemble learning algorithms for rapid damage mapping tasks. A damage classification framework is introduced and tested to categorize the building damage following the recent 2018 Sulawesi earthquake and tsunami. Three robust ensemble learning classifiers were investigated for recognizing building damage from Synthetic Aperture Radar (SAR) and optical remote sensing datasets and their derived features. The contribution of each feature dataset was also explored, considering different combinations of sensors as well as their temporal information. SAR scenes acquired by the ALOS-2 PALSAR-2 and Sentinel-1 sensors were used. The optical Sentinel-2 and PlanetScope sensors were also included in this study. A non-local filter in the preprocessing phase was used to enhance the SAR features. Our results demonstrated that the canonical correlation forests classifier performs better in comparison to the other classifiers. In the data fusion analysis, Digital Elevation Model (DEM)- and SAR-derived features contributed the most in the overall damage classification. Our proposed mapping framework successfully classifies four levels of building damage (with overall accuracy >90%, average accuracy >67%). The proposed framework learned the damage patterns from a limited available human-interpreted building damage annotation and expands this information to map a larger affected area. This process including pre- and post-processing phases were completed in about 3 h after acquiring all raw datasets.

The fine resolution of synthetic aperture radar (SAR) images enables the rapid detection of severely damaged areas in the case of natural disasters. Developing an optimal model for detecting damage in multitemporal SAR intensity images has been a focus of research. Recent studies have shown that computing changes over a moving window that clusters neighboring pixels is effective in identifying damaged buildings. Unfortunately, classifying tsunami-induced building damage into detailed damage classes remains a challenge. The purpose of this paper is to present a novel multiclass classification model that considers a high-dimensional feature space derived from several sizes of pixel windows and to provide guidance on how to define a multiclass classification scheme for detecting tsunami-induced damage. The proposed model uses a support vector machine (SVM) to determine the parameters of the discriminant function. The generalization ability of the model was tested on the field survey of the 2011 Great East Japan Earthquake and Tsunami and on a pair of TerraSAR-X images. The results show that the combination of different sizes of pixel windows has better performance for multiclass classification using SAR images. In addition, we discuss the limitations and potential use of multiclass building damage classification based on performance and various classification schemes. Notably, our findings suggest that the detectable classes for tsunami damage appear to differ from the detectable classes for earthquake damage. For earthquake damage, it is well known that a lower damage grade can rarely be distinguished in SAR images. However, such a damage grade is apparently easy to identify from tsunami-induced damage grades in SAR images. Taking this characteristic into consideration, we have successfully defined a detectable three-class classification scheme.

The satellite remote-sensing-based damage-mapping technique has played an indispensable role in rapid disaster response practice, whereas the current disaster response practice remains subject to the low damage assessment accuracy and lag in timeliness, which dramatically reduces the significance and feasibility of extending the present method to practical operational applications. Therefore, a highly efficient and intelligent remote-sensing image-processing framework is urgently required to mitigate these challenges. In this article, a deep learning algorithm for the semantic segmentation of high-resolution remote-sensing images using the U-net convolutional network was proposed to map the damage rapidly. The algorithm was implemented within a Microsoft Cognitive Toolkit framework in the GeoAI platform provided by Microsoft. The study takes the 2011 Tohoku Earthquake-Tsunami as a case study, for which the pre- and post-disaster high-resolution WorldView-2 image is used. The performance of the proposed U-net model is compared with that of deep residual U-net. The comparison highlights the superiority U-net for tsunami damage mapping in this work. Our proposed method achieves the overall accuracy of 70.9% in classifying the damage into “washed away,” “collapsed,” and “survived” at the pixel level. In future disaster scenarios, our proposed model can generate the damage map in approximately 2–15 min when the preprocessed remote-sensing datasets are available. Our proposed damage-mapping framework has significantly improved the application value in operational disaster response practice by substantially reducing the manual operation steps required in the actual disaster response. Besides, the proposed framework is highly flexible to extend to other scenarios and various disaster types, which can accelerate operational disaster response practice.

Although supervised machine learning classification techniques have been successfully applied to detect collapsed buildings, there is still a major problem that few publications have addressed. The success of supervised machine learning strongly depends on the availability of training samples. Unfortunately, in the aftermath of a large-scale disaster, training samples become available only after several weeks or even months. However, following a disaster, information on the damage situation is one of the most important necessities for rapid search-and-rescue efforts and relief distribution. In this paper, a modification of the supervised machine learning classification technique called logistic regression is presented. Here, the training samples are replaced with probabilistic information, which is calculated from the spatial distribution of the hazard under consideration and one or more fragility functions. Such damage probabilities can be collected almost in real time for specific disasters such as earthquakes and/or tsunamis. We present the application of the proposed method to the 2011 Great East Japan Earthquake and Tsunami for collapsed building detection. The results show good agreement with a field survey performed by the Ministry of Land, Infrastructure, Transport and Tourism, with an overall accuracy of over 80%. Thus, the proposed method can significantly contribute to a rapid estimation of the number and locations of collapsed buildings.

A powerful tsunami triggered by the Mw 9.0 Tohoku earthquake struck the northern Pacific coast of Japan in 2011, destroying several coastal communities in Iwate Prefecture, Miyagi Prefecture, and Fukushima Prefecture. Here, we investigate a new spatio-temporal slip model (model B) developed by the Cabinet Office of the Government of Japan. This slip model was compared against the non-uniform slip model estimated using tsunami waveform data (model A). We focused our analysis on two areas that were destroyed by the tsunami. The town of Onagawa and the coastal Sendai Plain area were selected because they were located in front of the epicentre, where the most significant slips were registered. Our simulation results revealed that the spatio-temporal slip distribution better replicated the observed data. Regarding the tsunami waveforms from the coastal tide gauge station and offshore stations, the Cabinet Office’s slip model showed an approximately 30% better accuracy relative to the non-uniform slip model. Furthermore, by comparing the local inundation features at two locations with unique topographic and coastal morphological characteristics, we also found that model B better replicated the measured inundation depths. Finally, considering that tsunami-induced damage is a direct function of various inundation features such as the flow depth and flow velocity, this new slip model can generate more realistic damage scenarios for future tsunami assessments.