Distribution Of Building Damage Research Articles

A Disparate Disaster: Spatial Patterns of Building Damage Caused by Hurricane Ian and Associated Socio-Economic Factors

The literature shows that communities under different socio-economic conditions suffer different levels of damage in disasters. In addition to the physical intensity of hazards, such differences are also related to the varying abilities of communities to prepare for and respond to disasters. This study analyzes the spatial patterns of building damage in Hurricane Ian in 2022 and investigates the socio-economic disparities related to the damage. Specifically, this study employs NASA’s Damage Proxy Map (DPM2) to analyze spatial patterns of building damage caused by the hurricane. Then, it uses statistical analysis to assess the relationships between building damage and hurricane intensity, building conditions, and socio-economic variables at the building and census tract levels. Furthermore, the study applies geographically weighted regression (GWR) to examine the spatial variation of the damage factors. The results provide valuable insights into the potential factors related to building damage and the spatial variation in the factors. The results also reveal the uneven distribution of building damage among different population groups, implying socio-economic inequalities in disaster adaptation and resilience. Moreover, the study provides actionable information for policymakers, emergency responders, and community leaders in formulating strategies to mitigate the impact of future hurricanes by identifying vulnerable communities and population groups.

Building typology classification using convolutional neural networks utilizing multiple ground-level image process for city-scale rapid seismic vulnerability assessment

Several studies have focused on generating seismic vulnerability maps for earthquake-prone areas, particularly in Indonesia. Building typologies are a key factor in determining vulnerability to earthquakes. However, conducting large-scale field surveys to determine the spatial distribution of building typologies in a city is uneconomical. This paper explores the use of a convolutional neural network (CNN) to automatically detect building typologies from diverse regions in Indonesia, utilizing both conventional and automated building image acquisition processes. In this study, datasets from three distinct image acquisition methods are trained with four unique CNN architectures to identify the best-performing model to classify building typologies. The sample size effect on CNN performance is also investigated. The results showed that randomly sampled Google Street View (GSV) images are the most effective dataset for the CNN model, achieving an f1-score of 84.33%. Among the network architectures tested, MobileNet demonstrated superior performance on the majority of evaluated datasets. As the sample size increases by about 350% in the dataset, there is a positive correlation with up to 2.3% f1-score improvement. Using the best-performing CNN model, two building vulnerability models were employed to assess the spatial distribution of building damage in the urban area of Bandung, considering a hypothetical scenario of an M7 earthquake. Incorporating local construction data, one of the generated maps estimated that approximately 55% of buildings in Bandung would experience moderate to severe structural damage. This study showcases the potential of CNN models in automating regional seismic assessments and providing valuable insights for comprehensive seismic mitigation strategies.

A Deep Learning Application for Building Damage Assessment Using Ultra-High-Resolution Remote Sensing Imagery in Turkey Earthquake

Rapid building damage assessment following an earthquake is important for humanitarian relief and disaster emergency responses. In February 2023, two magnitude-7.8 earthquakes struck Turkey in quick succession, impacting over 30 major cities across nearly 300 km. A quick and comprehensive understanding of the distribution of building damage is essential for efficiently deploying rescue forces during critical rescue periods. This article presents the training of a two-stage convolutional neural network called BDANet that integrated image features captured before and after the disaster to evaluate the extent of building damage in Islahiye. Based on high-resolution remote sensing data from WorldView2, BDANet used pre-disaster imagery to extract building outlines; the image features before and after the disaster were then combined to conduct building damage assessment. We optimized these results to improve the accuracy of building edges and analyzed the damage to each building, and used population distribution information to estimate the population count and urgency of rescue at different disaster levels. The results indicate that the building area in the Islahiye region was 156.92 ha, with an affected area of 26.60 ha. Severely damaged buildings accounted for 15.67% of the total building area in the affected areas. WorldPop population distribution data indicated approximately 253, 297, and 1,246 people in the collapsed, severely damaged, and lightly damaged areas, respectively. Accuracy verification showed that the BDANet model exhibited good performance in handling high-resolution images and can be used to directly assess building damage and provide rapid information for rescue operations in future disasters using model weights.

Dynamic post-earthquake updating of regional damage estimates using Gaussian Processes

The widespread earthquake damage to the built environment induces severe short- and long-term societal consequences. Better community resilience may be achieved through well-organized recovery. Decisions to organize the recovery process are taken under intense time pressure using limited, and potentially inaccurate, data on the severity and the spatial distribution of building damage. We propose to use Gaussian Process inference models to fuse the available inspection data with a pre-existing earthquake risk model to dynamically update regional post-earthquake damage estimates and thereby support a well-organized recovery. The proposed method consistently aggregates the gradually incoming building damage inspection data to reduce the uncertainty in ground shaking intensity geographic distribution and to update regional building damage estimates. The performance of the proposed Gaussian Process methodology is demonstrated on one fictitious earthquake scenario and two real earthquake damage datasets. A comparison with purely data-driven methods shows that the proposed method reduces the number of building inspections required to provide reliable and precise damage predictions.

Influence of the constitutive model in the damage distribution of buildings designed with an energy-based method

It is widely accepted in the seismic design of buildings a certain level of damage under moderate or severe seismic actions but preventing the damage concentration in them. On the other hand, the energy-based design methodology proposes an optimum strength distribution for designing the structure of the building aimed at achieving an approximated even distribution of the damage—energy dissipated by plastic deformations—under seismic actions. Different approaches for the optimum strength distribution have been proposed in both existing literature and standards. Most of them were formulated from the results obtained in non-linear numeric evaluations of elastic-perfectly plastic (EPP) structures, such as the findings proposed recently by the authors of this study. However, studies on the optimum strength distributions of reinforced concrete (RC) structures are scarce. The present study sheds light on this issue. Accordingly, the structures of four prototype buildings with 3, 6, 9, and 12 stories were designed through an energy-based method by using five approaches for the optimum strength distribution: those proposed by the authors and two others from the literature and standards. Then, different prototypes of the structures arose considering the different approaches for the optimum strength distribution, two soil classes (dense and medium dense), and two ductility levels (low and high). Such prototype structures were subjected to two sets of far-field ground motion records by using three different constitutive models for the shear force-interstory drift relationship: EPP, Clough model, and Modified Clough model. The first characterizes the steel structures and the rest are typical for RC structures. A complete analysis was carried out to obtain the distribution of damage for EPP and RC structures, their deviations with respect to the “ideal” even distribution of damage, and the possible damage concentration on specific stories. RC structures showed a higher dispersion for the distribution of damage than EPP structures although those designed with the optimum strength distributions proposed by the authors showed the lowest values in the order of those obtained with EPP structures designed with optimum strength distributions proposed in the literature.

Simulation of soil liquefaction distribution in downtown Mashiki during 2016 Kumamoto earthquake using nonlinear site response

Several sites located between Road No.28 and Akitsu River in downtown Mashiki were liquefied during the mainshock of the 2016 Kumamoto earthquake. According to the building damage survey results, only a few buildings were damaged in areas proximate to the Akitsu River, where liquefaction occurred, however, serious building damage occurred in neighboring regions. Therefore, the effect of soil liquefaction on strong ground motions in Mashiki should be ascertained. Moreover, the distribution of visible and invisible liquefaction is required to be estimated as well. In this study, the distribution of depth of groundwater level in Mashiki was studied, which decreased from 14 to 0 m from northeast to southwest. Thereafter, the nonlinearities of the shallow layers at four borehole drilling sites were identified from the experimental data using the Ramberg–Osgood relationship. Subsequently, the dynamic nonlinear effective stress analysis of the one-dimensional soil column was performed to 592 sites in Mashiki between the seismological bedrock and ground surface to estimate the distribution of strong ground motions during the mainshock. First, the ground motions estimated by the nonlinear analysis corresponded to the ground motions observed at the Kik-net KMMH16. Second, the soil nonlinearity of shallow layers was considerably strong in the entire target area especially in the southern Mashiki, and the PGV distribution was similar to the building damage distribution after the mainshock. Furthermore, the estimated distribution of the soil liquefaction site was similar to the observed results, whereas certain invisible-liquefaction sites were estimated in the north and middle of the target area.

Method for rapidly generating urban damage scenarios under non-uniform ground motion input based on matching algorithms and time history analyses

Due to the rapid urbanization, the built environment becomes more complex and denser, making cities especially vulnerable to strong earthquakes. It is important to simulate seismic disaster scenarios of buildings and critical infrastructures accurately in a reasonable time for the development of urban disaster prevention strategies and decision-making frameworks for post-earthquake emergence rescue. In this study, a novel methodology is established to rapidly generate detailed seismic disaster scenarios that virtually illustrate the damage distribution of buildings in city level without executing time history analysis. The non-uniform distribution of ground motions is taken into account for a large area. The large city is divided into small areas, and a scenario database is generated by performing massive nonlinear time history analyses through the city-level simulation platform recently developed by the authors. The proposed method selects and composes the best matching disaster scenarios by assessing the scenario database within a very short time using matching algorithms. Three matching algorithms, including the single wave matching algorithm, multiple waves splicing matching algorithm and multiple waves combination matching algorithm, are applied to form disaster scenario for any given input ground motion. The performance and effectiveness of the three matching algorithms are examined and a case study on simulation of a large city is presented. The results show that the matching algorithms are all feasible, and the multiple waves splicing matching algorithm is the most efficient. The proposed methodology not only produces reasonably similar results and findings to those obtained by the nonlinear time history analysis directly, but also can reduce the computing time by over 98%.

Investigation and modeling of the LPG tank truck accident in Wenling, China

In this study, the liquefied petroleum gas (LPG) accident in Wenling, Zhejiang Province, China, on June 13, 2020, was analyzed and simulated. An LPG tank truck overturned and collided with the concrete guardrail; the subsequent explosion of the tank released 25.36 t LPG. Shortly afterward, the LPG tank was shot into the air, and the gas cloud ignited, thereby triggering a fire and vapor cloud explosion (VCE). This accident killed 20 people, injured 175 people, and caused significant property loss. The accident timeline was established based on multiple accident images, and the accident process was discussed in detail. The distribution of tank debris, evaporation and spreading of the LPG pool, dispersion of the LPG gas cloud, and VCE were simulated with the EFFECTS and ALOHA software. The images of the accident scene helped to determine and analyze the accident process. In the particular case in which a tank is shot into the air with a continuous two-phase jet, the actual flight distance of the tank is much greater than the prediction of the debris distribution model. The gas cloud distribution simulated with the SLAB model approximately corresponds to a major part of the severely damaged zone. In this accident, the TNO multi-energy method and ALOHA provide relatively consistent predictions with the actual building damage distribution when models use a specific confined explosive mass.

Factors affecting earthquake responders’ building damage information needs and use

After an earthquake, many responding organizations need to understand the scale and distribution of building damage to react effectively. However, their building damage information needs and information use remain poorly understood, limiting the efficacy of information production, sharing, and research. To clarify those needs, we conducted a two-part survey, comprising semi-structured interviews and an online questionnaire, of building damage information users and providers. Based on the interview data and questionnaire responses, we characterize six post-disaster tasks that rely on building damage information by their timing and by the necessary qualities of the information they require. Through inductive analysis of the interview data, we show that responders’ use of building damage information also depends on factors beyond the building damage information itself—namely, trust, impediments to information sharing, their varying understandings of disaster, and their attitudes toward emerging technologies. These factors must be considered in the design of any effort to create and/or disseminate post-disaster building damage information.

Simulation of building damage distribution in downtown Mashiki, Kumamoto, Japan caused by the 2016 Kumamoto earthquake based on site-specific ground motions and nonlinear structural analyses

Most of the buildings damaged by the mainshock of the 2016 Kumamoto earthquake were concentrated in downtown Mashiki in Kumamoto Prefecture, Japan. We obtained 1D subsurface velocity structures at 535 grid points covering this area based on 57 identified velocity models, used the linear and equivalent linear analyses to obtain site-specific ground motions, and generated detailed distribution maps of the peak ground acceleration and velocity in Mashiki. We determined the construction period of every individual building in the target area corresponding to updates to the Japanese building codes. Finally, we estimated the damage probability by the nonlinear response model of wooden structures with different ages. The distribution map of the estimated damage probabilities was similar to the map of the damage ratios from a field survey, and moderate damage was estimated in the northwest where no damage survey was conducted. We found that both the detailed site amplification and the construction period of wooden houses are important factors for evaluating the seismic risk of wooden structures.

Interpreting the socio-technical interactions within a wind damage-artificial neural network model for community resilience.

The use of machine learning has grown in popularity in various disciplines. Despite the popularity, the apparent ‘black box’ nature of such tools continues to be an area of concern. In this article, we attempt to unravel the complexity of this black box by exploring the use of artificial neural networks (ANNs), coupled with graph theory, to model and interpret the spatial distribution of building damage from extreme wind events at a community level. Structural wind damage is a topic that is mostly well understood for how wind pressure translates to extreme loading on a structure, how debris can affect that loading and how specific social characteristics contribute to the overall population vulnerability. While these themes are widely accepted, they have proven difficult to model in a cohesive manner, which has led primarily to physical damage models considering wind loading only as it relates to structural capacity. We take advantage of this modelling difficulty to reflect on two different ANN models for predicting the spatial distribution of structural damage due to wind loading. Through graph theory analysis, we study the internal patterns of the apparent black box of artificial intelligence of the models and show that social parameters are key to predict structural damage.

Seismic Vulnerability Assessment and Mapping of Gyeongju, South Korea Using Frequency Ratio, Decision Tree, and Random Forest

The main purpose of this study was to compare the prediction accuracies of various seismic vulnerability assessment and mapping methods. We applied the frequency ratio (FR), decision tree (DT), and random forest (RF) methods to seismic data for Gyeongju, South Korea. A magnitude 5.8 earthquake occurred in Gyeongju on 12 September 2016. Buildings damaged during the earthquake were used as dependent variables, and 18 sub-indicators related to seismic vulnerability were used as independent variables. Seismic data were used to construct a model for each method, and the models’ results and prediction accuracies were validated using receiver operating characteristic (ROC) curves. The success rates of the FR, DT, and RF models were 0.661, 0.899, and 1.000, and their prediction rates were 0.655, 0.851, and 0.949, respectively. The importance of each indicator was determined, and the peak ground acceleration (PGA) and distance to epicenter were found to have the greatest impact on seismic vulnerability in the DT and RF models. The constructed models were applied to all buildings in Gyeongju to derive prediction values, which were then normalized to between 0 and 1, and then divided into five classes at equal intervals to create seismic vulnerability maps. An analysis of the class distribution of building damage in each of the 23 administrative districts showed that district 15 (Wolseong) was the most vulnerable area and districts 2 (Gangdong), 18 (Yangbuk), and 23 (Yangnam) were the safest areas.

CHARACTERISTICS OF TEXTURE INDEX OF DAMAGED BUILDINGS USING TIME-SERIES HIGH-RESOLUTION OPTICAL SATELLITE IMAGES

Abstract. A large-scale disaster has occurred due to the earthquake. In particular, 20% of the world's earthquakes with a magnitude of 6 or more occur near Japan. Damage analysis of buildings by image analysis have been effectively carried out using optical high-resolution satellite images and aerial photograph with spatial resolution of about 2 m or less. In this study, the damaged buildings caused by large-scale and continuous earthquakes in Kumamoto, Japan that occurred in April 2016 was selected as a typical example of damaged buildings. For these earthquake event, the applicability of damage distribution of buildings and recovery/restoration status by texture analysis was examined. The applicability of the representative in the dissimilarity texture analysis methods Gray- Level Co-occurrence Matrix (GLCM) method by image interpretation in the case of a large number of collapsed and wrecked buildings in a wide area was assessed. These results suggest that dissimilarity was applicable to the extraction of damaged and removed buildings in the event of such an earthquake. In addition, the analysis results were appropriately evaluated by comparing the field survey results with the image interpretation results of the pan-sharpened image. From these results, we confirmed the effectiveness of texture analysis using time-series high-resolution satellite images in grasping the damaged buildings before and immediately after the disaster and in the restoration situation 1 year after the disaster.

A numerical investigation of building damage during the 6 May 2012 Tsukuba tornado using hybrid meteorological model/engineering LES method

To validate applicability of the novel hybrid meteorological model/engineering LES method to damage estimation of actual tornado events, simulation is performed focusing on the severely damaged area of the strong tornado on May 6, 2012 in the city of Tsukuba with the realistic tornado structure reproduced in the meteorological model, the actual building distribution, and the topography. The flows over the disaster-affected low-rise residential area show an evident converging character at a height close to the rooftops. Intense strong wind bands are formed by the influence of the buildings. The simulated maximum wind gust is compared with the distribution of actual building damage. It reveals that the hybrid method can reasonably estimate distribution of the building damage over the target area in the tornado event. The track nevertheless exhibits building-scale perturbation that may be attributed to chaotic randomness of the tornado behavior on the complex rough surface. Using the simulated aerodynamic force, simplified case studies on the occurrence and process of damage to two individual buildings is also presented.

Evaluation of the damage distribution of buildings considering the characteristics of far-fault long-period ground motions

In this study, the damage distribution of buildings was evaluated considering the characteristics of far-fault long-period ground motions. The results demonstrated that the damage distribution trend of the analytical model concentrated on the lower floors with an increase in the seismic wave predominant period and concentrated on the top floor with an increase in the fundamental natural period of the building. The regression line for the yield shear force coefficient distribution, according to the seismic wave predominant period and the fundamental natural period of the building, was calculated using the multiple regression analysis. The curvature of the yield shear force coefficient distribution and the top floor yield shear force coefficient was observed to increase as the fundamental natural period of the building increased and as the seismic wave predominant period decreased.

Modeling of the Subsurface Structure from the Seismic Bedrock to the Ground Surface for a Broadband Strong Motion Evaluation in Kumamoto Plain

During the 2016 Kumamoto earthquakes, two earthquakes of seismic intensity 7 were observed in Mashiki Town, the foreshock (MJMA6.5) of April 14 and the main shock (MJMA7.3) of April 16, resulting in significant damage to structures near the fault. The distribution of damage of houses and other buildings [1] showed a tendency in which damage was concentrated in areas near the surface earthquake fault where the main shock presumably occurred. However, there were locations with slight damage even though they were immediately above the fault and locations with a relatively significant damage even though they were far from the fault. These phenomena are highly likely to be a result of soil structure. First, we built an initial geologic model by collecting boring data in areas of the Kumamoto plain near the fault where damage was severe. Next, we observed microtremors, collected earthquake observational records, and adjusted the layer thickness and S-wave velocity of the initial geologic model. Finally, we built a shallow and deep integrated ground model, compared it to the building damage distribution, and discussed the implications.

Shear‐wave velocity structure from MASW and SPAC methods: The case of Adra town, SE Spain

ABSTRACTThe damage distribution in the town of Adra (south‐eastern Spain) during the 1993 and 1994 Adra earthquakes of magnitude Mw~5.0 and maximum intensity degree of VII (European Macroseismic Scale) was mainly concentrated in the southeast sector, where the low‐diagenetic (soft) sediments outcrop. As new urbanizations are being planned in this sector, a soil classification based on the shallow shear‐wave velocity () structure is needed. For the purpose of earthquake disaster mitigation, the Spatial Autocorrelation (SPAC) and the Multichannel Analysis of Surface Waves (MASW) methods were used to propose integrated 2D models for the seismic response characterization of the shallow geology. Joint inversion of H/V spectral ratios of ambient noise, interpreted under the Diffuse Field Approach and the dispersion curves derived from the SPAC method allowed us to obtain more constrained models. Both SPAC and MASW methods provided similar results for the surveyed geological formations. From these models, a classification of the geological formations was carried out in terms of values and Eurocode 8 (European Committee for Standardization ) classes. Lower values in the 180‐360 m/s range were found in the southeastern sector of the town, where soft sediments outcrop and some building damage was reported during the 1993‐1994 earthquakes. The highest values exceeding 800 m/s appear in the northern sector, where the hardest rocks outcrop and no building damage was reported. The combination of the well‐suited database prepared for different geological formations with the 1:5,000 scale geological mapping was an important step to obtain the detailed soil microzonation map of Adra. This approach offered a new predictive insight into the building damage distribution, which would contribute to the appropriate urban planning for the future growth of the town.

Temporary Seismic Stations Installation and Building Damage Assessment of December 7th, 2016 Earthquake in Pidie Jaya, Aceh, Indonesia

A signicant Mw 6.5 earthquake occurred in Pidie Jaya, Aceh on December 7th, 2016. The event affected104 people death and more than 1000 people suered injuries due to the rubble of the building. Geologically, the region is composed by of Quaternary alluvial deposits. This is one of factor that amplication occurred in some area. On the other hand, an understanding of the source and mechanism of the earthquake needs to be done. A few days after the earthquake, we deployed 9 seismometers that covered the area of Pidie, Pidie Jaya and Bireuen. This experiment aims to record the aftershock and understanding of earthquake source and mechanism. In addition, we conducted building damage survey to know the pattern of distributionof building damage.

Measuring the Impact of Enhanced Building Performance on the Seismic Resilience of a Residential Community

The relationship between the earthquake performance of an inventory of buildings and the seismic resilience of a residential community is examined, by quantifying the immediate post-earthquake reduction and recovery of the shelter-in-place housing capacity. The effect of several mitigation strategies that involve replacing portions of the existing building stock with an enhanced seismic performance system is evaluated. The impact on community resilience is assessed based on the immediate and cumulative loss of permanent housing occupancy as well as the time to recover some fraction of the pre-earthquake housing capacity. The results show how the slope of the recovery curve for occupiable housing during different periods following the event can be linked to the distribution of building damage. In addition to limiting major damage, enhanced building seismic performance is shown to reduce aggregate losses over the recovery period, thereby having a significant effect on both the safety and resilience of residential communities.

Earthquake loss estimation for the Kathmandu Valley

Kathmandu Valley is geologically located on lacustrine sediment basin, characterized by a long history of destructive earthquakes. The past events resulted in large structural damage, loss of human life’s and property, and interrupted the social development. In recent years, the earthquake risk in this area has significantly increased due to uncontrolled development, poor construction practices with no earthquake safety provisions, and lack of awareness amongst the general public and government authorities. In this context, this study explores the realistic situation of earthquake losses due to future earthquakes in Kathmandu Valley. To this end, three municipalities: (a) Kathmandu Metropolitan City, (b) Lalitpur Sub-Metropolitan City and (c) Bhaktapur Municipality are selected for a case study. The earthquake loss estimation in the selected municipalities is performed through the combination of seismic hazard, structural vulnerability, and exposure data. Regarding the seismic input, various earthquakes scenario considering four seismic sources in Nepal are adopted. For what concerns the exposure, existing literature describing the construction typologies and data from the recent national census survey of 2011 are employed to estimate ward level distribution of buildings. The economic losses due to the earthquake scenarios are determined using fragility functions. Finally, the ward level distribution of building damage and the corresponding economic losses for each earthquake scenario is obtained using the OpenQuake-engine. The distribution of building damage within the Kathmandu Valley is currently being employed in the development of a shelter model for the region, involving various local authorities and decision makers.