Two consecutive earthquakes with magnitudes exceeding 7 occurred in the Republic of Turkey on February 6, 2023, causing severe damage to cities along the causative fault lines. Many satellite images were provided soon after the earthquakes for supporting the emergency response. In this study, the authors conducted a quick extraction of collapsed individual buildings using two pre-event and one post-event TerraSAR-X intensity images. The differences of backscattering coefficients and the correlation coefficients were calculated for pre-event and co-event SAR data pairs. Collapsed buildings were then identified based on the thresholds defined by the pre-event pair. Compared with visual interpretation results from high-resolution optical images, 64% of the collapsed buildings were detected successfully, demonstrating the effectiveness of our proposed method at an early stage of disaster response.

Wide spread landslides were extracted from airborne LiDAR data acquired before and after the 6 September 2018 Hokkaido-Eastern-Iburi earthquake. By taking the difference of Digital Surface Models (DSMs) from the LiDAR datasets, the failed slopes and accumulated soils were recognized. Since the pre-event dataset was obtained 12 years before the earthquake, the growth of trees was estimated based on forest registry block data. The adjusted DSM difference was compared with the visual inspection result of aerial photos by the Geospatial Information Authority of Japan and the accuracy of the confusion matrix was discussed.

After a major disaster, a rapid assessment of building damage is highly required for emergency response and prompt recovery. Remote sensing technologies have been widely applied for building damage mapping. Combining machine-learning algorithms (e.g., deep learning) and satellite images has recently demonstrated success in boosting damage recognition methods. Although previous techniques have shown great success, they primarily adopt supervised settings, often requiring a minimum number of training samples to achieve acceptable accuracy. Moreover, previous methods also are developed for specific target areas, which makes it challenging to apply them to other regions in case of future disasters. This paper presents a novel unsupervised approach for building damage mapping, focusing on collapsed structures, using modern convolutional neural network (CNN) models and high-resolution remote sensing imagery. We apply our mapping framework to revise the building damage following the 2007 Peru-Pisco Earthquake and the recent 2023 Turkey and Syria Earthquakes.

A series of earthquakes hit Kumamoto Prefecture, Japan, in April 2016. Numerous landslides occurred in the surroundings of the Aso volcano and in the areas within 10-km from the Futagawa fault. Two months after the earthquake, the stagnation of the rainy season front brought heavy rainfall to the Kyushu Island. The rainfall caused additional landslides and the expansion of the earthquake-induced landslides. In this study, three temporal Lidar data taken before the 2016 Kumamoto earthquake, after the earthquake and after the heavy rainfall, were used to estimate the earthquake-induced and rainfall-induced landslides in the outer ring of the Mt. Aso. The height difference between two temporal digital surface models (DSMs) and that of digital terrain models (DTMs) were calculated. They were combined to extract the collapse zones and sediments of landslides. The results were verified by comparing with the reports of visual interpretation of aerial optical images.

In the rainy season of July 2021, debris flow suddenly occurred in Atami City, Japan, after a heavy rainfall and it caused significant impacts to human lives and houses along the Aizome river. In this study, the pre- and post-event SAR images acquired from the ALOS-2 satellite were used to extract the affected area and the results were compared with optical images and field survey data. The change detection techniques could identify the hard-hit zone to some extent. But the mountainous topography and thick vegetation cover hindered the accurate extraction of the affected area only from the SAR imagery. To interpret the SAR intensity images, a simple SAR simulation was further carried out using DSMs obtained from pre- and post-event LiDAR data.

One span of an aqueduct bridge suddenly collapsed in Wakayama City in western Japan on October 3, 2021. This study investigates the use of remote sensing data for the assessment of bridge situations in the normal time and in accidents/disasters. A field survey was conducted by the authors with the aid of a small UA V. Google Street View photos taken before the accident were also used. Based on these data, it is estimated that more than 5 hangers out of 18 might have failed for the collapsed span when the bridge collapse occurred. The failure of 4 hangers was also confirmed in the adjacent surviving span from the UAV images. The pre- and post-event high-resolution TerraSAR- X intensity images were also introduced to extract the collapsed span from the SAR data.

Record-breaking heavy rainfall hit Japan from July 3 to 31, 2020. The water level of rivers rose rapidly, and many bridges were washed away. In this study, two temporal ALOS-2 PALSAR-2 intensity images were introduced to detect damaged bridges over the Kuma River in Yatsushiro City and Ashikita Town, Kumamoto Prefecture, Japan. First, the backscattering models of bridges were investigated. Then the damage conditions of twenty-five target bridges were examined by the difference and correlation coefficient in the backscatter intensity. In addition, the increases of water levels was estimated by the movements of bridges' SAR model. A field survey report and a post-event SPOT-7 satellite optical image were used to verify our results.

An Mw 6.6 earthquake struck the eastern Iburi area of Hokkaido, Japan, on September 6, 2018. Due to the strong shaking, more than 3000 landslides occurred around Atsuma Town and killed 36 people. In this study, two pre-event and one post-event ALOS-2 PALSAR-2 images were used to extract landslides in Abira, Atsuma and Mukawa Towns. The characteristics of landslides were investigated using the backscattering coefficient, the coherence and the slope of elevation. Then the landslides in the target area were extracted by the optimal threshold values. A procedure using both the difference of backscattering coefficient and the difference of coherence was proposed to extract landslides. Finally, the result was verified by comparing with a post-event optical satellite image and a landslide distribution map by the GSI.

Due to the huge tsunamis occurred in the 2011 Tohoku-Oki, Japan, earthquake, more than 100 bridges located in the Pacific coast of the Tohoku region were severely damaged. In this study, the extraction of the damaged bridges in Miyagi Prefecture, Japan, was conducted by two methods using two post-event TerraSAR-X (TSX) intensity images, respectively. First, the statistical features within the outlines of the target bridges were calculated. The thresholding method of the backscatter intensity was applied to extract the damaged bridges. Then the TSX image was transformed into a binary image including water and non-water regions. The percentages of no-water regions within the bridge outlines were used to classify the washed-away and survived bridges. By comparing with the optical images and the report of field surveys, the accuracies of the proposed two methods and the influence of the shooting date were investigated.

In the March 11, 2011 Tohoku-Oki, Japan, earthquake, many bridges in Iwate and Miyagi Prefectures were washed away by associated tsunamis. In this study, nine bridges in the inundated areas of Iwate Prefecture, Japan were selected as targets to estimate their damage status from post-event TerraSAR-X (TSX) satellite images and Pi-SAR-X2 airborne SAR images. By visual interpretation, washed-away bridges could be identified from both the satellite and airborne very high-resolution SAR images. The statistical analysis was carried out to classify non-damaged bridges, debris blocked bridges and washed-away bridges. Pre- and post-event optical images and filed survey reports were introduced as the truth data of bridges' situation.

Unmanned Aerial Vehicles (UAVs) are becoming an efficient tool of image collection for affected areas due to natural disasters. In this study, UAV flights were carried out over a landslide affected site due to the July 2017 Northern Kyushu heavy rainfall in Japan. The UAV flights captured high-resolution still photos, and using them, three-dimensional (3D) models were developed based on a SfM (Structure-from-Motion) technique. The developed models could depict the damage situations vividly, and the location accuracy was evaluated through the comparison with the result of GPS measurements on the site.

Unmanned Aerial Vehicles (UAVs) are becoming an efficient tool of high-resolution image collection for the places that are difficult to access or observe from the ground. In this study, UAV flights were carried out by the authors over various damage sites due to the 2016 Kumamoto, Japan earthquake, such as surface faulting, overturned tombstones, landslides, collapsed buildings and a bridge. The UAV flights captured high-resolution video footages and photos, and using them, three-dimensional (3D) models were developed based on a SfM (Structure-from-Motion) technique. The developed models could depict the damage situations vividly, and the accuracy was evaluated through comparison with aerial photos and field measurement results.

In this study, a practical methodology was presented to map damaged buildings using high resolution synthetic aperture radar (SAR) images and post-event building damage data from the 2013 Super Typhoon Haiyan, in Tacloban city, the Philippines. To detect destroyed structures, we focused on the changes in the radar signal within footprints of buildings between pre- and post-event SAR images. The method was tested using a 1.0 m resolution COSMO-SkyMed SAR images taken over Tacloban city, the Philippines. The method proves, with 73% accuracy in this case, to be suitable for estimating destroyed buildings.

A strong typhoon hit the Pacific coast of Japan from the night of October 15th to the morning of 16th, 2013, and caused huge damages, especially in Izu-Oshima island, Tokyo. Synthetic aperture radar (SAR), which can observe the earth surface despite of weather conditions, is an effective tool to grasp damage situation caused by typhoons. In this study, pre-event Pi-SAR-L and post-event Pi-SAR-L2 airborne radar images with full polarizations were used to detect landsides and debris flows in Izu-Oshima. First, the extraction of potential landslide areas was carried out using only the post-event image by land-cover classification and from the standardized difference polarization index (NDPI). Then the difference of backscattering intensity between the pre- and post-event SAR images was calculated to extract potential landslides. In addition, a 5-m resolution digital elevation model (DEM) was introduced to remove errors. Finally, the results were verified through the comparison with the result from visual interpretation.

In this study, the extent of the flooded areas by the Super Typhoon Haiyan in the Philippines were extracted using ASTER VNIR images taken over Tacloban city in the Visayas. In order to constraint the affected area, we employed the normalize difference vegetation and water indices (NDVI and NDWI) from the pre- and post-event images. The extension of the flooded area was determined by comparing the index characteristics before and after the event. A phase-based change detection method indices was applied to classify the affected area into three classes according to the changes between the pre- and post-images. Through NDWI the flooded areas were detected despite the moderate resolution of ASTER images. In addition, the phase-based analysis successfully detected level of change within the affected area that may be correlated to the damage observed on field surveys. The results from the phase-based analysis were verified with damage levels obtained through visual damage inspection using high resolution satellite images.

A strong typhoon “Haiyan” affected Southeast Asia on November 8, 2013, caused gigantic destruction in the Philippines. In this study, two pre- and one post-event COSMO-SkyMed SCSB data were used to detect the damaged area around Tacloban City, Leyte Island. First, the severe damaged areas were detected according to the difference between the pre- and post-event speckle divergence values. Then the pre- and co-event coherence (NDCI) and correlation coefficient (NDCOI) were calculated from the three temporal data. The relationships between the four building damage levels and NDCI or NDCOI value were obtained by introducing the visual interoperation result. Using this relationship, the possibility of each damage class was estimated in the whole urban area.

A method for capturing the two-dimensional (2D) surface movements from two temporal TerraSAR-X (TSX) intensity images has been proposed by the authors in previous research. However, it is impossible to detect the three-dimensional (3D) actual displacement from one pair of TSX images. Hence, three pairs of TSX images taken in ascending and descending paths were used to estimate 3D crustal movements in this study. First, the 2D crustal movements due to the 2011 Tohoku earthquake were detected from the three sets respectively. The relationship between the 3D actual displacement and 2D converted movement in SAR images was derived according to the observation model and shooting condition of the SAR sensor. Then the absolute 3D movements were estimated by the combination of the detected 2D movements that occurred within a short time interval. The results were verified by the GEONET observation records.

In this study, the flooded areas following the 2011 central Thailand flood were extracted using VNIR and TIR images of ASTER and ScanSAR-mode images of TerraSAR-X. The existence of water body was easily recognized for open spaces without trees and buildings from the NDVI value. The surface temperature was also found to be effective in detecting floods in a wide open space although it is limited by its coarse spatial resolution. The SAR intensity images were the most effective because water surfaces showed weak backscatter and they can be acquired at nighttime and under cloud-cover conditions. The extracted results were validated by a high-resolution optical satellite image.

Significant crustal movements were caused by the 2011 Tohoku, Japan earthquake. A method for capturing the surface movements from pre- and postevent TerraSAR-X (TSX) intensity images is proposed in this letter. Because the shifts of unchanged buildings were considered as crustal movements in the two synthetic aperture radar images, we first extracted buildings from the pre- and postevent images using a segmentation approach. Then, the unchanged buildings were detected by matching the buildings in the pre- and postevent images at similar locations. Finally, the shifts were calculated by area-based matching. The method was tested on the TSX images covering the Sendai area. Compared with GPS observation records, the proposed method was found to be able to detect crustal movement at a subpixel level.

The 11 March 2011 Tohoku, Japan earthquake caused gigantic tsunamis and widespread devastations. Various satellites quickly captured the details of affected areas, and were used for emergency response. In this study, high-resolution pre- and post-event TerraSAR-X (TSX) intensity images were used to identify damaged buildings. Since the damaged buildings show changes in backscattering intensity, they can be detected by calculating the difference. A GIS map was introduced to identify individual damaged buildings and investigate their characteristics. According to the side-looking nature of SAR sensors, the buildings' shapes obtained from the GIS map were converted to match their locations in the TSX images. Then washed-away and damaged buildings were extracted using the changed area of SAR intensity within a building's wall and outline. The results were compared with visual interpretation results, and the accuracy of the proposed method was confirmed.

Urban areas grow and change rapidly all over the world. Hence, regular and up-to-date information on urban changes is required for urban planning and disaster management. In this study, two temporal TerraSAR-X images are used to monitor urban changes. The study area is focused on a part of central Tokyo, Japan. Firstly, the changes between two images are checked by color composition. Then the difference and the correlation coefficient between the two images are calculated with a sliding window. A new factor that combines the difference and the correlation coefficient is proposed to detect changed areas. Finally, two high resolution optical images are introduced to verify the accuracy of the detection results.

Shadows in remote sensing images often result in problems for many applications such as land-cover classification, change detection, and damage detection in disasters. Due to these reasons, it is very useful if the radiance of shadowed areas is corrected to the same radiance as shadow-free areas. In this study, a shadow detection and correction method is proposed. Shadowed areas are detected by object-based classification, using brightness values and a neighbor relationship. Then the detected shadowed areas are corrected by a liner function to produce a shadow-free image. The shadowed areas with different darkness are corrected with different ratios to improve the accuracy of the result. The spectral characteristics of sunlit and shadowed areas in several QuickBird images were studied and then the shadow-free radiance was obtained.

SAR images obtained before and after a natural disaster are considered to be useful for emergency response due to its all-weather and sunlight-independent characteristics. Recently, the spatial resolutions of SAR systems have been improved significantly. In this paper, SAR intensity images acquired before and after the 2008 Iwate-Miyagi, Japan, earthquake from ALOS/PALSAR (L-band) and TerraSAR-X (X-band) are employed to investigate the radar backscattering characteristics for various acquisition and surface conditions. The spatial resolution, radar frequency, flight path, and incidence angle were shown to affect SAR backscattering echo, depending on surface materials and roughness. It is also observed that the difference of the backscattering coefficients at the pre- and post-event times gets large and their correlation coefficient becomes small at the locations of landslides and slope failures.

The effects of shadow in remote sensing imagery are investigated. The measurement of radiance in sunlit and shadowed areas was carried out to investigate the spectral characteristics of sunlight. Based on this observation, it is found that the radiance ratio (shadow/sunlit) increases as the sunlight gets weaker and the ratio is dependent on the wavelength of sunlight. The darkness of shadow is also found to vary depending on the surrounding condition. Thus the condition to restore a shadow-free image depends on the spectral bands and the location even in one image. A QuickBird image is then introduced and the spectral characteristics of sunlit and shadowed areas are investigated. Based on these observations, a method to detect shadowed areas and restore the shadow-free radiance for the multi-spectral bands is proposed. The effectiveness of the shadow correction method is demonstrated for the QuickBird image.

A new object-based method is developed to extract the moving vehicles and subsequently detect their speeds from two consecutive digital aerial images automatically. Several parameters of gray values and sizes are examined to classify the objects in the image. The vehicles and their associated shadows can be discriminated by removing big objects such as roads. To detect the speed, firstly the vehicles and shadows are extracted from the two images. The corresponding vehicles from these images are linked based on the order, size, and their distance within a threshold. Finally, using the distance between the corresponding vehicles and the time lag between the two images, the moving speed can be detected. Our test shows a promising result of detecting the moving vehicles' speeds. Further development will employ the proposed method for a pair of QuickBird panchromatic and multi-spectral images, which are at a coarser spatial resolution.