Periodic Reporting for period 1 - LANDSCAPEforCHANGE (A LANDSCAPE approach to cultural heritage management in the context of climate CHANGE)
Reporting period: 2023-07-03 to 2025-07-02
It is acknowledged by International declarations and policy guidance documents that cultural heritage (CH) can contribute directly to many of the Sustainable Development Goals (SDGs), including resilience and adaptation to climate change (SDG 13). Yet, the integration of CH into the wider framework of sustainable urban development and resilience in the context of climate change is faced with different challenges. There is a challenge around assessing the vulnerability of CH to climate change and integrating its vulnerability status into the broader context of sustainable urban development. This challenge is imposed by the lack of climate vulnerability assessment for diverse CH types (cultural, natural, exceptional, and everyday/ordinary), a lack of a framework that addresses landscapes rather than heritage sites in isolation, and a lack of climate change adaptation policy for CH. To overcome these challenges, this project transcends traditional boundaries between tangible and intangible heritage and goes beyond the state of the art by addressing the city as a living heritage. The adopted landscape approach is interdisciplinary in nature. Cultural heritage mapping and resilience assessment require the integration of different disciplines and knowledge from conservation of cultural and natural landscape, climate change impacts and social sciences and humanities. The project leverages social media as a data source and as a transformative tool that drives the co-construction of local heritage values and enhances community resilience. By applying a landscape people-centered approach, this project (i) responds to local cultural contexts and value systems, (ii) integrates distinct theoretical perspectives to address the complex layering of the various spatial and functional process-related dimensions of the landscape, and (iii) addresses policies and governance concerns at international and local levels. The methodological framework incorporates social media metadata with survey data and integrates different qualitative, quantitative, geographical, and visual analysis along with machine learning techniques. The results of the project inform evidence-based decision-making for heritage conservation, enhancing community resilience and recovery, and promoting the integration of cultural heritage considerations into climate adaptation planning and disaster risk management.
The work carried out combined conceptual development, methodological innovation, and empirical analysis, to explore the intersection of cultural heritage, disaster management, and digital technologies.
Development of a people-centered conceptual framework articulating the role of social media in heritage conservation and community resilience in the context of climate change. The framework advocates for more inclusive and locally contextualized practices by engaging communities in the co-construction of heritage values and enhancing multivocality through digital platforms. It highlights the transformative role of digitally mediated heritage practices, from digitization and crowdsourcing to active community participation in crisis response.
Development of a multidimensional framework for the assessment of cultural heritage vulnerability to flood hazards. In this framework, cultural heritage is framed within a landscape approach drawing on the concept of the historic urban landscape as a physical entity, a lived space and a layered socioeconomic environment. The framework integrates built and natural heritage with area-based vulnerability and socioeconomic deprivation, considering households' exposure to flood risk. This involved a systematic review of existing methods, indicator development, data collection, and spatial analysis using GIS-based tools.
Social media data collection and analysis, including Facebook, Instagram, and X (formerly Twitter). The analysis aimed to investigate the role of grassroots mobilizations on social media in building community resilience and addressing specific challenges during disaster events. Several methods were applied for the analysis of textual and visual data. For textual data, Latent Dirichlet Allocation (LDA) from the topicmodels package was used for topic modelling. In addition, the CardiffNLP RoBERTa-based model was used in Python to conduct sentiment analysis. For visual data, analysis was conducted using ChatGPT-4, a Generative Artificial Intelligence (GenAI) model, which was used to generate descriptions, categorise images, and extract emotional tone from photos shared on Instagram.
Interviews with Facebook group administrators whose groups were initiated in response to the crisis and who were actively engaged in mobilizing community members. The questionnaire consisted of five sections addressing: (1) the purpose of social media use during the crisis; (2) the role of social media in building community and coordinating efforts to address specific challenges; (3) perceived community support and interaction on social media; (4) the impact of online interactions on offline support; and (5) the role of social media in showcasing community resilience and unity during the crisis.
The main achievements of the project include:
A conceptual and methodological foundation for integrating digital platforms into heritage conservation and resilience planning, promoting inclusive engagement, co-construction of values, and grassroots mobilization.
A validated multidimensional vulnerability assessment framework, offering a replicable methodology for understanding the association between cultural heritage vulnerability, socioeconomic deprivation and area-based vulnerability, supporting the identification of historic urban landscape qualities and priority areas for reconstruction and social support.
A methodological workflow for textual and visual analysis of social media data shared during disaster events.
Empirical evidence demonstrating the potential of social media as a data source and transformative tool supporting heritage conservation, community resilience and disaster recovery.
Three open-access datasets supporting replication and further research.
Development of a people-centered conceptual framework articulating the role of social media in heritage conservation and community resilience in the context of climate change. The framework advocates for more inclusive and locally contextualized practices by engaging communities in the co-construction of heritage values and enhancing multivocality through digital platforms. It highlights the transformative role of digitally mediated heritage practices, from digitization and crowdsourcing to active community participation in crisis response.
Development of a multidimensional framework for the assessment of cultural heritage vulnerability to flood hazards. In this framework, cultural heritage is framed within a landscape approach drawing on the concept of the historic urban landscape as a physical entity, a lived space and a layered socioeconomic environment. The framework integrates built and natural heritage with area-based vulnerability and socioeconomic deprivation, considering households' exposure to flood risk. This involved a systematic review of existing methods, indicator development, data collection, and spatial analysis using GIS-based tools.
Social media data collection and analysis, including Facebook, Instagram, and X (formerly Twitter). The analysis aimed to investigate the role of grassroots mobilizations on social media in building community resilience and addressing specific challenges during disaster events. Several methods were applied for the analysis of textual and visual data. For textual data, Latent Dirichlet Allocation (LDA) from the topicmodels package was used for topic modelling. In addition, the CardiffNLP RoBERTa-based model was used in Python to conduct sentiment analysis. For visual data, analysis was conducted using ChatGPT-4, a Generative Artificial Intelligence (GenAI) model, which was used to generate descriptions, categorise images, and extract emotional tone from photos shared on Instagram.
Interviews with Facebook group administrators whose groups were initiated in response to the crisis and who were actively engaged in mobilizing community members. The questionnaire consisted of five sections addressing: (1) the purpose of social media use during the crisis; (2) the role of social media in building community and coordinating efforts to address specific challenges; (3) perceived community support and interaction on social media; (4) the impact of online interactions on offline support; and (5) the role of social media in showcasing community resilience and unity during the crisis.
The main achievements of the project include:
A conceptual and methodological foundation for integrating digital platforms into heritage conservation and resilience planning, promoting inclusive engagement, co-construction of values, and grassroots mobilization.
A validated multidimensional vulnerability assessment framework, offering a replicable methodology for understanding the association between cultural heritage vulnerability, socioeconomic deprivation and area-based vulnerability, supporting the identification of historic urban landscape qualities and priority areas for reconstruction and social support.
A methodological workflow for textual and visual analysis of social media data shared during disaster events.
Empirical evidence demonstrating the potential of social media as a data source and transformative tool supporting heritage conservation, community resilience and disaster recovery.
Three open-access datasets supporting replication and further research.
The results obtained from the application of the multidimensional framework for assessing the vulnerability of cultural heritage (CH) to flood hazards go beyond the state of the art by moving beyond conventional approaches that typically focus solely on physical attributes or exposure. Instead, the analysis adopts a holistic methodology that integrates multiple dimensions—physical, social, and economic—offering contextualization of heritage within the urban environment. The research findings highlight a strong correlation between cultural heritage vulnerability, socio-economic deprivation and area-based vulnerability.
Complementing this, the research advances the field by applying a data-driven, cross-platform analysis combined with qualitative inquiry to examine the role of social media in disaster response and community resilience. It moves beyond conventional approaches that tend to analyze social media use through a single platform—most often Twitter—and instead offers a comparative and complementary examination of both Facebook and Twitter during the 2021 flood in Belgium. By integrating topic modeling and sentiment analysis of posts with in-depth interviews of Facebook group administrators, the study demonstrates how digital platforms serve distinct but interconnected roles: Facebook functions as a space for grassroots mobilization and localized mutual aid, while Twitter serves as a tool for global communication and climate advocacy. This multidimensional approach captures not only the thematic content and emotional dynamics of online interactions but also their offline implications, offering a rare longitudinal perspective on digital engagement throughout crisis and short-term recovery phases. The results have significant implications for improving disaster preparedness, supporting volunteer coordination, and enhancing real-time situational awareness. To ensure broader uptake and success, the findings underscore the need for further research on underrepresented platforms and populations, targeted outreach to digitally excluded groups, and integration of social media analytics into formal disaster response systems. In addition, the research delivered results that go beyond the state of the art by applying a multimodal analytical framework that systematically integrates visual, emotional, and textual analyses of Instagram posts to capture the evolving dynamics of disaster response and recovery during the 2021 floods in western Germany. In contrast to most prior research, which focuses predominantly on text-based platforms such as Twitter and relies on unimodal analyses, the analysis demonstrates the unique value of visual narratives in conveying both the material impacts of the disaster and the emotional trajectories of the affected communities. The results reveal six key visual themes—flood damage, rescue operations, solidarity and resilience, media coverage, post-disaster recovery, and government involvement—alongside the evolving emotional landscape. The results provide empirical evidence that visual storytelling on Instagram offers a richer, more nuanced understanding of disaster impacts and recovery dynamics. The potential impacts of this work include informing more effective crisis communication strategies, enhancing the monitoring of public sentiment and community resilience during recovery, and integrating visual social media data into disaster management and policy frameworks.
Complementing this, the research advances the field by applying a data-driven, cross-platform analysis combined with qualitative inquiry to examine the role of social media in disaster response and community resilience. It moves beyond conventional approaches that tend to analyze social media use through a single platform—most often Twitter—and instead offers a comparative and complementary examination of both Facebook and Twitter during the 2021 flood in Belgium. By integrating topic modeling and sentiment analysis of posts with in-depth interviews of Facebook group administrators, the study demonstrates how digital platforms serve distinct but interconnected roles: Facebook functions as a space for grassroots mobilization and localized mutual aid, while Twitter serves as a tool for global communication and climate advocacy. This multidimensional approach captures not only the thematic content and emotional dynamics of online interactions but also their offline implications, offering a rare longitudinal perspective on digital engagement throughout crisis and short-term recovery phases. The results have significant implications for improving disaster preparedness, supporting volunteer coordination, and enhancing real-time situational awareness. To ensure broader uptake and success, the findings underscore the need for further research on underrepresented platforms and populations, targeted outreach to digitally excluded groups, and integration of social media analytics into formal disaster response systems. In addition, the research delivered results that go beyond the state of the art by applying a multimodal analytical framework that systematically integrates visual, emotional, and textual analyses of Instagram posts to capture the evolving dynamics of disaster response and recovery during the 2021 floods in western Germany. In contrast to most prior research, which focuses predominantly on text-based platforms such as Twitter and relies on unimodal analyses, the analysis demonstrates the unique value of visual narratives in conveying both the material impacts of the disaster and the emotional trajectories of the affected communities. The results reveal six key visual themes—flood damage, rescue operations, solidarity and resilience, media coverage, post-disaster recovery, and government involvement—alongside the evolving emotional landscape. The results provide empirical evidence that visual storytelling on Instagram offers a richer, more nuanced understanding of disaster impacts and recovery dynamics. The potential impacts of this work include informing more effective crisis communication strategies, enhancing the monitoring of public sentiment and community resilience during recovery, and integrating visual social media data into disaster management and policy frameworks.