Abstract

     Remote sensing provides crucial support for building damage assessment in the wake of hurricanes. This article proposes a coupled deep learning-based model for damage assessment that leverages a large very high-resolution satellite images dataset and a flexibility of building footprint source. Convolutional Neural Networks were used to generate building footprints from pre-hurricane satellite imagery and conduct a classification of incurred damage. We emphasize the advantages of multiclass classification in comparison with traditional binary classification of damage and propose resolving dataset imbalances due to unequal damage impact distribution with a focal loss function. We also investigate differences between relying on learned features using a deep learning approach for damage classification versus a commonly used shallow machine learning classifier, Support Vector Machines, that requires manual feature engineering. The proposed model leads to an 86.3% overall accuracy of damage classification for a case event of Hurricane Michael and an 11% overall accuracy improvement from the Support Vector Machines classifier, suggesting better applicability of such an open-source deep learning-based workflow in disaster management and recovery. Furthermore, the findings can be integrated into emergency response frameworks for automated damage assessment and prioritization of relief efforts.

Introduction

     Hurricanes have been the most expensive type of disaster in the USA (CRED 2019) and historically deathliest, continuing to drive the importance of improving techniques to assess post-hurricane urban damage. The process of such damage assessment involves detection, classification, and evaluation of disaster damage to an economy and society on local, county, state, or tribal levels (FEMA 2016). Remote sensing analysis has proved to be indispensable in aiding these assessment efforts (Adams et al. 2009; Waharte and Trigoni 2010; Stow et al. 2015). In particular, the Federal Emergency Management Agency (FEMA) relies on remote sensing analysis to rapidly assess a large-scale impact and monitor areas that cannot be effectively accessed on the ground. The timeliness, accuracy, and semantic information of such remote sensing analysis results are of key interest (Stow et al. 2015).

Conclusions

     In this study, we proposed a damage assessment workflow from VHR big data imagery, xBD, consisting of two CNNs that delineate building footprints and classify hurricane-incurred damage into four categories: un-damaged, lightly damaged, severely damaged, and destroyed buildings. The coupled model allows users to utilize an existing building footprint, unlike unified models. Another key contribution of this study was addressing a class imbalance problem in the xBD dataset with a focal loss function. We examined a case study of Hurricane Michael in 2018 around the Panama City metropolitan area, where our proposed models achieved an overall accuracy of 84.6% for building footprint segmentation and 86.3% for damage classification tasks. The model successfully identified undamaged buildings with an F1-score of 95.4% and predicts three damage classes (minor, major damage, and destroyed) with 59.3%, 68.7%, and 74.2% F1-score, respectively. An output of this model presents a probability vector of each building belonging to damage classes, thus, creating an opportunity for emergency management and first responders to set a higher or lower threshold for alerting about a hurricane damage presence.

Climate change directly affects the future security of cultural resources. Cultural heritage and in particular, archives, are increasingly at risk of degradation due to climate change threats and triggers. This study evaluated present and future consequences of water-related climate change impacts using a mapping methodology to assess exposure of American archives to incompatible weather extremes. Susceptibility to climate change threats like sea level rise, storm surge, surface water flooding, and humidity, all influenced by a combination of temperature rise and increased precipitation, at a worst-case scenario were assessed for 1232 archival repositories. Results indicate that approximately 98.8% of archives are likely to be affected by at least one climate risk factor, though on average, most archives are at low risk of exposure (90%) when risk factors are combined. Future storm surge plus sea level rise was likely to impact 17.7% of archival repositories with 22.1% affected by only storm surge and 4.3% affected by only sea level rise (1.8- m scenario). Fewer archives were likely to be susceptible to surface water flooding (2.4%). More than 90% of archives were estimated to have a temperature change greater than ±1 °C, with 7.5% of sites likely to change by ±10 °C, and 69.5% of archives were likely to receive at least 152 mm more rainfall by 2100 over current annual averages. In terms of sustainability, developing appropriate socio-economic planning schemes that integrate cumulative exposure of archives to future climate patterns is critically important for safeguarding society and its heritage. The outcomes from the risk assessment in this study aid in the decision-making process by promoting strategic adaptation protocols and providing administrators a way to prioritize archival management goals based on the expected severity of future climate change impacts.

Introduction

Climate change is increasingly recognized as a threat to cultural heritage (Rockman et al., 2016; Fatorić and Seekamp, 2017a). Cultural heritage includes both tangible forms such as paintings, buildings, monuments, and other material objects, and intangible forms, such as folklore, customs, and traditional knowledge. The well-accepted water-related climate change threats and climatetriggered phenomena that can impact cultural heritage resources include: sea level rise (Taboroff, 2000; Adger et al., 2013; Marzeion and Levermann, 2014; Anderson et al., 2017); storm surge (Gontz et al., 2011; Balica et al., 2012; Daire et al., 2012; Lickley et al., 2014); surface flooding (Dupont and Van Eetvelde, 2013; Wang, 2015; Vojinovic et al., 2016); precipitation (Haugen and Mattsson, 2011; Wang, 2015); temperature change (Hong et al., 2012; Huijbregts et al., 2012; Leissner et al., 2015); and humidity (Bernikola, et al., 2008; Bratasz et al., 2012; Lankester and Brimblecombe, 2012; Morawitz et al., 2013; Tornari et al., 2013; Camuffo et al., 2014; Bertolin et al., 2015; Tornari et al., 2015). Much of the recent work on assessing climate risks to cultural heritage has largely focused on immovable heritage such as archaeological sites (Anderson et al. 2017; Constantinidis, 2009; Reeder-Myers, 2015), cultural landscapes (Dupont and Van Eetvelde, 2013), buildings (Grossi et al., 2007; Bonazza et al., 2009a; Bonazza et al., 2009b; Haugen and Mattsson, 2011; Wu, et al., 2014; Wang, 2015), and UNESCO World Heritage sites (Smith et al., 2011; Viles and Cutler, 2012; Marzeion and Levermann, 2014; Margottini, 2015; Howard et al., 2016; Vojinovic et al., 2016). This body of literature has often focused on risks around coastal or riverine communities given the long history of human settlement in these environments (Reeder-Myers, 2015). A recent systematic literature review examined the integration of climate change and cultural heritage scholarship, which began about 14 years ago (Fatorić and Seekamp, 2017a). According to Fatorić and Seekamp (2017a), research to date focused primarily on European areas, was highly interdisciplinary, and reflected a large range of quantitative and qualitative research methods. Little of the existing work “explore[s] or acknowledge[s] the barriers, limits, and constraints to adaptation or preservation,” and significant gaps remain on the topic of adaptation planning across cultural heritage professions (Fatorić and Seekamp, 2017b). Perhaps because they largely preserve movable cultural heritage (e.g., paintings, museum objects, furniture, documents), archives1 have received less attention in the climate risk literature. Archives preserve historical records in multiple formats that are critical for legal matters, administrative accountability, and documentary cultural heritage (SAA, 2017). When these records and documents are lost following extreme weather events, for example, their absence severely handicaps socio-cultural and economic reconstruction efforts (Gordon-Clark, 2012).