An official website of the United States government
Building an annotated damage image database is the first step to support AI-assisted hurricane impact analysis. Up to now, annotated datasets for model training are insufficient at a local level despite abundant raw data that have been collected for decades. This paper provides a systematic approach for establishing an annotated hurricane-damaged building image database to support AI-assisted damage assessment and analysis. Optimal rectilinear images were generated from panoramic images collected from Hurricane Harvey, Texas 2017. Then, deep learning models, including Amazon Web Service (AWS) Rekognition and Mask R-CNN (Region Based Convolutional Neural Networks), were retrained on the data to develop a pipeline for building detection and structural component extraction. A web-based dashboard was developed for building data management and processed image visualization along with detected structural components and their damage ratings. The proposed AI-assisted labeling tool and trained models can intelligently and rapidly assist potential users such as hazard researchers, practitioners, and government agencies on natural disaster damage management.
more »
« less
Scalable approach to create annotated disaster image database supporting AI-driven damage assessment
Abstract As coastal populations surge, the devastation caused by hurricanes becomes more catastrophic. Understanding the extent of the damage is essential as this knowledge helps shape our plans and decisions to reduce the effects of hurricanes. While community and property-level damage post-hurricane damage assessments are common, evaluations at the building component level, such as roofs, windows, and walls, are rarely conducted. This scarcity is attributed to the challenges inherent in automating precise object detections. Moreover, a significant disconnection exists between manual damage assessments, typically logged-in spreadsheets, and images of the damaged buildings. Extracting historical damage insights from these datasets becomes arduous without a digital linkage. This study introduces an innovative workflow anchored in state-of-the-art deep learning models to address these gaps. The methodology offers enhanced image annotation capabilities by leveraging large-scale pre-trained instance segmentation models and accurate damaged building component segmentation from transformer-based fine-tuning detection models. Coupled with a novel data repository structure, this study merges the segmentation mask of hurricane-affected components with manual damage assessment data, heralding a transformative approach to hurricane-induced building damage assessments and visualization.
more »
« less
- Award ID(s):
- 2103754
- PAR ID:
- 10508032
- Publisher / Repository:
- Springer Science + Business Media
- Date Published:
- Journal Name:
- Natural Hazards
- Volume:
- 120
- Issue:
- 13
- ISSN:
- 0921-030X
- Format(s):
- Medium: X Size: p. 11693-11712
- Size(s):
- p. 11693-11712
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Ambinakudige_Shrinidhi; Dash_Padmanava (Ed.)This research explores the utilization of the Black Marble nighttime light (NTL) product to detect and assess damage caused by hurricanes, tornadoes, and earthquakes. The study first examines average regional NTL trends before and after each disaster, demonstrating that NTL patterns for hurricanes closely align with the features of a resilience curve, unlike those for earthquakes and tornadoes. The relative NTL change ratio is computed using monthly and daily NTL data, effectively reducing variance due to daily fluctuations. Results indicate the robustness of the NTL change ratio in detecting hurricane damage, whereas its performance in earthquake and tornado assessment was inconsistent and inadequate. Furthermore, NTL demonstrates a high performance in identifying hurricane damage in well-lit areas and the potential to detect damage along tornado paths. However, a low correlation between the NTL change ratio and the degree of damage highlights the method’s limitation in quantifying damage. Overall, the study offers a promising, prompt approach for detecting damaged/undamaged areas, with specific relevance to hurricane reconnaissance, and points to avenues for further refinement and investigation.more » « less
-
Abstract In the aftermath of Hurricane Ike in 2008 in the United States, the “Ike Dike” was proposed as a coastal barrier system, featuring floodgates, to protect the Houston‐Galveston area (HGA) from future storm surges. Given its substantial costs, the feasibility and effectiveness of the Ike Dike have been subjects of investigation. In this study, we evaluated these aspects under both present and future climate conditions by simulating storm surges using a set of models. Delft3D Flexible Mesh Suite was utilized to simulate hydrodynamic and wave motions driven by hurricanes, with wind and pressure fields spatialized by the Holland model. The models were validated against data from Hurricane Ike and were used to simulate synthetic hurricane tracks downscaled from several general circulation models and based on different sea level rise projections, both with and without the Ike Dike. Flood maps for each simulation were generated, and probabilistic flood depths for specific annual exceedance probabilities were predicted using annual maxima flood maps. Building damage curves were applied to residential properties in the HGA to calculate flood damage for each exceedance probability, resulting in estimates of expected annual damage as a measure of quantified flood risk. Our findings indicate that the Ike Dike significantly mitigates storm surge risk in the HGA, demonstrating its feasibility and effectiveness. We also found that the flood risk estimates are sensitive to hurricane intensity, the choice of damage curve, and the properties included in the analysis, suggesting that careful consideration is needed in future studies.more » « less
-
Abstract Nature‐based climate solutions (NCS) are championed as a primary tool to mitigate climate change, especially in forested regions capable of storing and sequestering vast amounts of carbon. New England is one of the most heavily forested regions in the United States (>75% forested by land area), and forest carbon is a significant component of climate mitigation policies. Large infrequent disturbances, such as hurricanes, are a major source of uncertainty and risk for policies relying on forest carbon for climate mitigation, especially as climate change is projected to alter the intensity and extent of hurricanes. To date, most research into disturbance impacts on forest carbon stocks has focused on fire. Here, we show that a single hurricane in the region can down between 121 and 250 MMTCO2e or 4.6%–9.4% of the total aboveground forest carbon, much greater than the carbon sequestered annually by New England's forests (16 MMTCO2e year−1). However, emissions from hurricanes are not instantaneous; it takes approximately 19 years for downed carbon to become a net emission and 100 years for 90% of the downed carbon to be emitted. Reconstructing hurricanes with the HURRECON and EXPOS models across a range of historical and projected wind speeds, we find that an 8% and 16% increase in hurricane wind speeds leads to a 10.7‐ and 24.8‐fold increase in the extent of high‐severity damaged areas (widespread tree mortality). Increased wind speed also leads to unprecedented geographical shifts in damage, both inland and northward, into heavily forested regions traditionally less affected by hurricanes. Given that a single hurricane can emit the equivalent of 10+ years of carbon sequestered by forests in New England, the status of these forests as a durable carbon sink is uncertain. Understanding the risks to forest carbon stocks from disturbances is necessary for decision‐makers relying on forests as a NCS.more » « less
-
null (Ed.)Recent hurricane events have caused unprecedented amounts of damage on critical infrastructure systems and have severely threatened our public safety and economic health. The most observable (and severe) impact of these hurricanes is the loss of electric power in many regions, which causes breakdowns in essential public services. Understanding power outages and how they evolve during a hurricane provides insights on how to reduce outages in the future, and how to improve the robustness of the underlying critical infrastructure systems. In this article, we propose a novel scalable segmentation with explanations framework to help experts understand such datasets. Our method, CnR (Cut-n-Reveal), first finds a segmentation of the outage sequences based on the temporal variations of the power outage failure process so as to capture major pattern changes. This temporal segmentation procedure is capable of accounting for both the spatial and temporal correlations of the underlying power outage process. We then propose a novel explanation optimization formulation to find an intuitive explanation of the segmentation such that the explanation highlights the culprit time series of the change in each segment. Through extensive experiments, we show that our method consistently outperforms competitors in multiple real datasets with ground truth. We further study real county-level power outage data from several recent hurricanes (Matthew, Harvey, Irma) and show that CnR recovers important, non-trivial, and actionable patterns for domain experts, whereas baselines typically do not give meaningful results.more » « less
