Geo-foundation models and UAV data for post flooding damage assessment in Mozambique

Pakistan has faced numerous natural disasters like floods, earthquakes, landslides and environmental degradation which severely affects the Pakistan’s economy and results in various problems like causalities, diseases, water stress and severe damages (e.g., houses, public infrastructure and agricultural land erosion). There is a lack of systematic approaches to analyze pre and post damage assessment for estimation of exact loses and the total cost for rehabilitation of damaged infrastructure in an efficient way. There exist a variety of mechanisms but GIS based flood mapping is considered the most efficient to manage the flood situation. This study is focused on evaluation of flood affected areas especially in Punjab using WorldView-2, 8-band multi-spectral imagery by applying Remote Sensing (RS) and GIS techniques. The research area is comprised of Kot Addu and Muzaffargarh Districts in Punjab province of Pakistan that faced a catastrophic super flood of 2010. The WorldView-2, Quick Bird and multispectral satellite imagery are capable of making better decisions and assessment of flood effected area accurately. RS and GIS techniques can achieve the objectives and significant analyses through visual interpretations. These techniques are also used to identify the flood affected regions. The study site was examined by applying supervised classification on the basis of the training areas which were obtained during the field surveys in the study site. Supervised classification determines that 16900.96 Hectors of agriculture land was damaged while Sparse Riverine Forest had the area 44.52 hectors. The damaged built-up area was 1805.78 Hectors. RS and GIS techniques are efficient for flood mapping.

Pakistan has faced numerous natural disasters like floods, earthquakes, landslides and environmental degradation which severely affects the Pakistan’s economy and results in various problems like causalities, diseases, water stress and severe damages (e.g., houses, public infrastructure and agricultural land erosion). There is a lack of systematic approaches to analyze pre and post damage assessment for estimation of exact loses and the total cost for rehabilitation of damaged infrastructure in an efficient way. There exist a variety of mechanisms but GIS based flood mapping is considered the most efficient to manage the flood situation. This study is focused on evaluation of flood affected areas especially in Punjab using WorldView-2, 8-band multi-spectral imagery by applying Remote Sensing (RS) and GIS techniques. The research area is comprised of Kot Addu and Muzaffargarh Districts in Punjab province of Pakistan that faced a catastrophic super flood of 2010. The WorldView-2, Quick Bird and multispectral satellite imagery are capable of making better decisions and assessment of flood effected area accurately. RS and GIS techniques can achieve the objectives and significant analyses through visual interpretations. These techniques are also used to identify the flood affected regions. The study site was examined by applying supervised classification on the basis of the training areas which were obtained during the field surveys in the study site. Supervised classification determines that 16900.96 Hectors of agriculture land was damaged while Sparse Riverine Forest had the area 44.52 hectors. The damaged built-up area was 1805.78 Hectors. RS and GIS techniques are efficient for flood mapping.

Flooding and submersion-induced morphological and physiological adaptive strategies in Lonchocarpus cultratus

This study focused on flood damage assessment for future floods under the impact of climate change. Four river basins of Southeast Asia were selected for the study. They included the Pampanga River Basin (PRB) in the Philippines, the Solo River Basin (SRB) in Indonesia, the Lower Mekong River Basin (LMRB) in Cambodia and Vietnam, and the Chao Phraya River Basin (CPRB) in Thailand. Flood damage to rice crops was assessed by flood damage functions considering flood depth and duration and the growth stage of rice plants. Flood characteristics such as flood depth, duration, and distribution were computed using the rainfall–runoff–inundation model to assess flood hazards under the present and future climatic conditions produced by MRI-AGCM3.2S. The damage assessment methodology for rice crops employed in this study was verified using data on past flood events. Then, flood damage assessment was conducted for both the present climate (1979–2003) and future climate (2075–2099) conditions, using MRI-AGCM3.2S precipitation datasets. Flood damage was assessed for worst cases chosen from each climate period and for floods of 50- and 100-year return periods with different rainfall patterns chosen from each climate scenario. The results of flood hazard and damage assessment show that the flood inundation area for a 100-year flood may increase in the future by 20% in PRB; by 66% in SRB; by 27% in LMRB; and by 27% in CPRB. The flood damage area of paddy fields for a 100-year flood may also increase in the future by 16% in PRB; by 55% in SRB; by 23% in LMRB; and by 13% in CPRB.

A simple quantitative approach for post earthquake damage assessment of flexure dominant reinforced concrete bridges

Rapid damage assessment of regional beam bridges after earthquake based on simplified models and different seismic inputs

Visual scene understanding is the core task in making any crucial decision in any computer vision system. Although popular computer vision datasets like Cityscapes, MS-COCO, PASCAL provide good benchmarks for several tasks (e.g. image classification, segmentation, object detection), these datasets are hardly suitable for post disaster damage assessments. On the other hand, existing natural disaster datasets include mainly satellite imagery which has low spatial resolution and a high revisit period. Therefore, they do not have a scope to provide quick and efficient damage assessment tasks. Unmanned Aerial Vehicle (UAV) can effortlessly access difficult places during any disaster and collect high resolution imagery that is required for aforementioned tasks of computer vision. To address these issues we present a high resolution UAV imagery, FloodNet, captured after the hurricane Harvey. This dataset demonstrates the post flooded damages of the affected areas. The images are labeled pixel-wise for semantic segmentation task and questions are produced for the task of visual question answering. FloodNet poses several challenges including detection of flooded roads and buildings and distinguishing between natural water and flooded water. With the advancement of deep learning algorithms, we can analyze the impact of any disaster which can make a precise understanding of the affected areas. In this paper, we compare and contrast the performances of baseline methods for image classification, semantic segmentation, and visual question answering on our dataset. FloodNet dataset can be downloaded from here: https://github.com/BinaLab/FloodNet-Supervised_v1.0.

Visual scene understanding is the core task in making any crucial decision in any computer vision system. Although popular computer vision datasets like Cityscapes, MS-COCO, PASCAL provide good benchmarks for several tasks (e.g. image classification, segmentation, object detection), these datasets are hardly suitable for post disaster damage assessments. On the other hand, existing natural disaster datasets include mainly satellite imagery which have low spatial resolution and a high revisit period. Therefore, they do not have a scope to provide quick and efficient damage assessment tasks. Unmanned Aerial Vehicle(UAV) can effortlessly access difficult places during any disaster and collect high resolution imagery that is required for aforementioned tasks of computer vision. To address these issues we present a high resolution UAV imagery, FloodNet, captured after the hurricane Harvey. This dataset demonstrates the post flooded damages of the affected areas. The images are labeled pixel-wise for semantic segmentation task and questions are produced for the task of visual question answering. FloodNet poses several challenges including detection of flooded roads and buildings and distinguishing between natural water and flooded water. With the advancement of deep learning algorithms, we can analyze the impact of any disaster which can make a precise understanding of the affected areas. In this paper, we compare and contrast the performances of baseline methods for image classification, semantic segmentation, and visual question answering on our dataset.

<p>Post-earthquake reconnaissance missions are critical to understand the event characteristics, identify building and infrastructure vulnerabilities, and improve future construction practice. However, in-field missions can present logistic and safety challenges that do not make them viable in every post-disaster scenario. Remote sensing technique can be used to rapidly collect a large amount information that can be used to enrich the post-event learning process. While the possibility to deploy teams in the field remain a valuable asset for an integrated understanding of technical and socio-economic factors, a mix of remote and in-field reconnaissance activities can be a way forward in post-disaster management.</p><p>This work presents the results of a hybrid mission mobilised by the Earthquake Engineering Field Investigation Team (EEFIT) after the 2021 Haiti earthquake. On 14 August 2021, a 7.2 magnitude earthquake struck the Tiburon Peninsula in the Caribbean nation of Haiti, approximately 150km east of the capital Port au Prince. The event was followed by numerous aftershocks up to magnitude 5.7, and tiggered over 1000 landslides. Over 2000 people lost their lives, with over 15,000 injured and over 137,000 houses damaged or destroyed. The estimated economic impact is of the order of US$1.6 billion. Due the complex political and security situation in Haiti, coupled with the global pandemic, a full in field mission was not considered feasible, so a hybrid mission was designed instead.</p><p>First, open-source information was collected and used to characterise the seismic event, analyse the strong ground motion and compare to established national and international earthquake codes and standard. Second, remote sensing techniques including Interferometric Synthetic Aperture Radar (InSAR) and Optical/Multispectral imagery were used to understand the earthquake mechanism, the ground displacement distribution and the possibility to detect landslide on a regional scale. The general applicability of remote sensing technique in the context of post disaster assessment was also evaluated. Finally, the earthquake impact on different building typologies in Haiti was investigated through the damage assessment of over 2000 buildings comprising schools, hospitals, churches and housing. This was done in collaboration with the Structural Extreme Events Reconnaissance (StEER) team, who mobilised a team of local non-experts to rapidly record building damage.</p><p>This talk summarises the mission setup and findings, and discusses the benefits of and difficulties encountered during this hybrid reconnaissance.</p>

Although hydrodynamic models are used extensively to quantify the physical hazard of hurricane storm surge, the connection between the physical hazard and its effects on the built environment has not been well addressed. The focus of this dissertation research is the improvement of our understanding of the interaction of hurricane storm surge with the built environment. This is accomplished through proposed methodologies to describe, assess and model residential building damage from hurricane storm surge. Current methods to describe damage from hurricane events rely on the initiating mechanism. To describe hurricane damage to residential buildings, a combined wind and flood damage scale is developed that categorizes hurricane damage on a loss-consistent basis, regardless of the primary damage mechanism. The proposed Wind and Flood (WF) Damage Scale incorporates existing damage and loss assessment methodologies for wind and flood events and describes damage using a seven-category discrete scale. Assessment of hurricane damage has traditionally been conducted through field reconnaissance deployments where damage information is captured and cataloged. The increasing availability of high resolution satellite and aerial imagery in the last few years has led to damage assessments that rely on remotely sensed information. Existing remote sensing damage assessment methodologies are reviewed for high velocity flood events at the regional, neighborhood and per-building levels. The suitability of using remote sensing in assessing residential building damage from hurricane storm surge at the neighborhood and per-building levels is investigated using visual analysis of damage indicators. Existing models for flood damage in the United States generally quantify the economic loss that results from flooding as a function of depth, rather than assessing a level of physical damage. To serve as a first work in this area, a framework for the development of an analytical damage model for residential structures is presented. Input conditions are provided by existing hydrodynamic storm surge models and building performance is determined through a comparison of physical hazard and building resistance parameters in a geospatial computational environment. The proposed damage model consists of a two-tier framework, where overall structural response and the performance of specific components are evaluated.

The Indian Ocean Tsunami on 26th December 2004, had caused devastation on land and coastal ecosystems. One of the worst affected ecosystems is the mangrove forest. Many areas of mangrove forest have been destroyed, and it may pose a long-term impact for the region, both in terms of forest and biodiversity conservation and in terms of the ability of the ecosystem to support the livelihoods of the coastal communities. The tragedy of tsunami has requested the rationale for conserving and sustainably managing natural ecosystems among the researchers and scientific community. The research was carried out to make a post-tsunami damage assessment and recovery of mangrove forest in Banda Aceh city area. Several high-resolution satellite imageries data have been effectively used to detect, assess and monitor the changes of mangrove forests in the pre and post-tsunami period using multi-temporal optical satellite data. An on-screen visual interpretation was made to identify the heterogeneous patches for the preliminary classification of fieldwork. A supervised classification was performed to obtain the pre and post Tsunami assessment maps. Accuracy assessment of the classified maps was performed on pixel-level using ground truth. The result highlighted the changes in the spatial extent of the mangrove forests in the study area as a result of Tsunami disaster. The latest condition of the mangrove forest recovery is about 90.3% of the total damage area.

Due to the high population density and economic property in urban environments, urban flooding can result in substantial economic losses. However, accurate urban flood modeling and assessing the associated damages are challenging due to complex urban layouts and rugged topography. Furthermore, the rapid development of urban underground spaces has introduced additional complexities. To address the demanding need for accurate flood modeling and damage assessment in cities, a fine‐scale flood damage assessment method was proposed in this study, based on the 1D/2D coupled hydrodynamic model, which can not only assess the flood damage on the ground surface but also evaluate the flood loss in underground spaces. Taking the Gangxi drainage area in Wuhan City, China as a case, this study extensively discussed the flood inundation processes on the ground surface as well as in the buildings and underground spaces, under different rainfall scenarios with the return period ranging from 2 to 200 years. Based on the high spatial‐temporal hydrodynamic simulations, the flood damage degrees were evaluated for the buildings and underground spaces with different industry types. The results indicated that the inundation of buildings and underground spaces in densely built cities cannot be neglected in urban flood modeling. Buildings with public service attributes can have a higher flood damage degree, as compared with other types of buildings and underground spaces.

Measuring and monitoring of river channel evolution or changes under laboratory conditions is an important scope in hydraulic assessment. Measurement such as changes in river channel profile for instance, provides an important indicator on erosion and accretion rates in hydraulic modelling. Under controlled conditions, the changes in channel profile are usually measured using a high precision point gauge. However, when large numbers of points of river profile need to be measured, the use of point gauge method becomes laborious and time consuming. This study proposed a digital close range photogrammetry technique to measure natural river channel bed profile changes in laboratory. The objective of this study was to investigate the changes of physical river model profile for pre and post flooding simulations using digital close range photogrammetry technique. Small scale physical model experimental works were conducted in the Hydraulics and Hydrology Laboratory to observe the river profile evolution during the events. The flood flume utilized in this study is 4.95 m long, 1.38 m wide and 1.26 m deep with carved V-shaped natural main channel with a bed slope of 1:500. Data measured from digital close range photogrammetry technique during pre-flooding at t = 0 second and post flooding events at t= 9660 second were compared to determine the channel profile evolution or changes after simulation of flood event. The results show the changes in invert level or bed level between non-flooding and flooding events due to erosion is varies from 1 mm (minimum) to 6 mm (maximum) along the channel between chainage 0mm and chainage 2000 mm with total volume of erosion is 1157 x 103 mm3 respectively. It can be concluded that the digital close range photogrammetry technique can be used as a complimentary method to measure and monitor the changes of river channel profile in the laboratory

Assessing potential damage of flood events is an important component in flood risk management. Determining direct flood damage is commonly done using depth-damage curves, which denote the flood damage that would occur at specific water depths per asset or per land-use class. Many countries have developed flood damage models using depth-damage curves based on analysis of past flood events and on expert judgement. However, the fact that such damage curves are not available for all regions hampers damage assessments in some areas. Moreover, due to different methodologies employed for various damage models in different countries, damage assessments cannot be directly compared with each other, obstructing also supra-national flood damage assessments.\n\nTo address these problems a globally consistent database of depth-damage curves has been developed. This dataset contains damage curves depicting fractional damage function of water depth as well as maximum damage values for a variety of assets and land use classes. Based on an extensive literature survey concave damage curves have been developed for each continent, while differentiation in flood damage between countries is established by determining maximum damage values at the country scale. These maximum damage values are based on construction cost surveys from multinational construction companies, which provide a coherent set of detailed building cost data across dozens of countries. A consistent set of maximum flood damage values for all countries was computed using statistical regressions with socio-economic World Development Indicators. Further, based on insights from the literature survey, guidance is also given on how the damage curves and maximum damage values can be adjusted for specific local circumstances, such as urban vs. rural locations or use of specific building material. This dataset can be used for consistent supra-national scale flood damage assessments, and guide assessment in countries where no damage model is currently available.

Abstract. Reliable information on building stock and its vulnerability is important for understanding societal exposure to floods. Unfortunately, developing countries have less access to and availability of this information. Therefore, calculations for flood damage assessments have to use the scarce information available, often aggregated on a national or district level. This study aims to improve current assessments of flood damage by extracting individual building characteristics and estimate damage based on the buildings' vulnerability. We carry out an object-based image analysis (OBIA) of high-resolution (11 cm ground sample distance) unmanned aerial vehicle (UAV) imagery to outline building footprints. We then use a support vector machine learning algorithm to classify the delineated buildings. We combine this information with local depth–damage curves to estimate the economic damage for three villages affected by the 2019 January river floods in the southern Shire Basin in Malawi and compare this to a conventional, pixel-based approach using aggregated land use to denote exposure. The flood extent is obtained from satellite imagery (Sentinel-1) and corresponding water depths determined by combining this with elevation data. The results show that OBIA results in building footprints much closer to OpenStreetMap data, in which the pixel-based approach tends to overestimate. Correspondingly, the estimated total damage from the OBIA is lower (EUR 10 140) compared to the pixel-based approach (EUR 15 782). A sensitivity analysis illustrates that uncertainty in the derived damage curves is larger than in the hazard or exposure data. This research highlights the potential for detailed and local damage assessments using UAV imagery to determine exposure and vulnerability in flood damage and risk assessments in data-poor regions.