Mapping the Multi-Vulnerabilities of Outdoor Places to Enhance the Resilience of Historic Urban Districts: The Case of the Apulian Region Exposed to Slow and Rapid-Onset Disasters

Abstract

Recent critical events brought attention to the increasing exposure of urban environments to both slow and rapid onset disasters, which arise from both anthropogenic and natural causes. These events have particularly severe effects on historic centres, which are characterized by high levels of vulnerability and valuable assets exposed to risk. To minimize the impact on tangible and intangible cultural heritage values, especially in outdoor public areas such as squares and streets, it is crucial to establish coherent mitigative and adaptive solutions for different types of hazards. This research presents a methodology aimed at defining levels of multi-vulnerabilities in historic districts in the Apulia Region (Italy), considering the recurrent hazards to which the latter is prone. It uses a multi-step process based on structured and non-structured methodologies and tools for single risks, examined in combination, to determine the main properties characterizing the vulnerability assessment. The dataset was analyzed in a GIS environment to evaluate the selected Apulian case study (Molfetta) in Multi-Asynchronous Hazard scenarios, showing the compounded levels of criticalities for open areas and streets. This information is intended to support authority and emergency managers in identifying priority interventions and increasing the resilience of the outdoor public places.

1. Introduction

The development of the resilience concept applied to the multi-scale territorial management for the built urban environment is currently one of the most important themes debated in this research field [1]. This topic engages the scientific community and involves urban administrations and authorities as well. The original concept of resilience, which originated in ecology [2], was incorporated in disciplines operating in the context of risk prevention and mitigation. In this context, improving urban resilience involves understanding and managing the vulnerabilities, whether intrinsic or generated, in the built environment that affects structures and their users. This is achieved by focusing on the nature of critical/calamitous events, which reveal vulnerabilities. Such events are typically referred to as Slow Onset (SOD) and Rapid Onset Disasters (ROD) depending on their duration [3].

Many experiences related to risk mitigation and emergency planning in urban areas were evaluated and implemented for the management of specific ROD events [4,5,6]. They were also used to assess the built environment exposed to them (e.g., wall resistance, urban resilience, etc.). In literature, the main goals of urban adaptative plans related to ROD are related to the mitigation of the impacts of climate change on users and the built environments [7]. However, investigations are often conducted without considering the interaction among different SODs and/or RODs, also in asynchronous ways. This lack of consideration can lead to challenges in managing the combinations of these phenomena in the built environment, potentially increasing the magnitude of critical events on structures, open places, and local communities. Moreover, this occurrence may lead to unintended consequences by cross impacts between technical and organizational solutions developed to tackle a single type of risk and neglecting overlapping features [8].

Buildings and places with historical significance represent a part of the urban heritage characterized by a lower resilience value due to the combination of the high intrinsic vulnerability levels and cultural/historical value. However, such a built environment constitutes an exception to mainstream emergency management and urban transformation processes [9,10] due to the administrative–conservative laws that define its protection. This restrictiveness is also linked to the slow process of their creation and modification over time, which generated a multiplicity of morpho-organizational features on the district level and morpho-constitutive attributes (and thus performances) characterizing the building scale. Therefore, their transformation, especially those aimed at disaster management, should account for the preservation of all the values related to the traditional experience of places. This is also in line with recent activities for the conservation of the so-called historic urban landscape, as discussed in the recommendation by UNESCO [11] where member states recognize the strong connection between the sustainable growth of cities and the protection of cultural heritage. In a multi-hazard perspective, this becomes a central point for discussing single or multiple technical solutions that are both compatible with places and effective for the emergency phase [12]. Moreover, historic centres constitute “cultural” parts of the built environment in cities and such relevance welcomes the presence and the interaction of inhomogeneous classes of users, both residential and tourists [13]. Simultaneously, the concepts of “disaster management”, “management of transformations”, and “choice of solutions” extend their dimension towards a district level, conceiving the multiple uses and users of external areas as well as information management on a larger scale. In this context, the assessment of vulnerabilities at a district scale for ancient urban cores is part of the historic urban landscape approach as an opportunity to study their inherent positive features to be preserved near to tangible and intangible values [14,15,16].

In this outline, the present work is part of the resilient cultural urban context to disaster exposure (ResCUDE) project aiming at addressing multi-risk analysis of historic centres according to a holistic method based on an interdisciplinary approach [17]. The project methodology is structured in three macro-actions: (i) the study of Multi-Asynchronous Hazard (MAH) scenarios in representative contexts, (ii) the evaluation of areas most susceptible to MAH scenarios starting from the analysis of inherent vulnerabilities of case studies for single risks, (iii) the assessment of technical solutions to mitigate single risks and their interplay in others to select effective ones, and finally, (iv) the evaluation and calibration of such solutions to consider and align with the human behaviours. Within the large Italian cultural heritage, the project finds its field of application in the Apulian territory, a southern Italian region. It is a strategic choice to determine combinations of RODs and SODs at the regional scale, while the assessment of risk, resilience, as well as the required actions may be determined in a wider perspective for such Cultural Built Environments. The methodology was tested on a representative case study in terms of hazard exposure, inherent fabric features, and the social and cultural attractiveness of places. The peculiarity of the Apulia Region originates from the increasing interest in historical and cultural assets of the area related to the wide diffusion of relevant cultural heritage evidence and from the particular attention to the transformation processes and conservation already noticed in the drawing up of regional planning tools (Apulia Regional Landscape Plan—PPTR). Specifically for discussion, this work presents the method for mapping the most vulnerable sub-areas in Cultural Built Environments in cities exposed to specific MAH scenarios. This mapping starts with the recognition of the main parameters of such built environments prone to regional hazardous events, which are then processed to obtain single and multi-vulnerability maps, and then properly reorganized in digital models, following validated or innovative approaches and tools. In detail, the paper is structured into four main sections:

• The parametrization of the built environment, intended as a system of buildings and unbuilt areas, infrastructures, and people, according to three main sub-classes of characters. Here, the process focuses on the framing of tools, strategies and methodologies already tested—even for built areas without historical character—in the technical-scientific field, highlighting the features of the built environment that influence their physical vulnerability to the selected RODs and SODs (Section 2).
• The discussion of methods and tools for the recognition of vulnerable sub-district areas to MAH scenarios, following main sectorial approaches and tools (Section 3).
• Application to the case study to set and test data and methods (Section 4).
• Discussion for implementing future goals and developments (Section 5).

2. Parametrization of the Historic District Characters for the Analysis of Multi-Asynchronous-Hazardous Events in the Apulia Region

In order to understand the risk and conceive urban spaces resilient to disasters, its standardized assessment [18] suggests the analysis of the built environment according to three main determinants:

• The hazard refers to the possible occurrences of events that could affect the exposed elements in the built environment.
• The exposure concerns the set of elements in the built environment directly exposed to dangerous events.
• The vulnerability is usually related to the susceptibility of exposed elements to suffer damages of dangerous effects during and after disasters. Particularly, the “vulnerability” class can involve the “physical” and “social” dimensions, depending on the elements considered in the analysis.

Thus, the risk assessment requires focusing on all the elements that constitute the built environment. As a critical concept of cities, the built environment is the result of the historical process aimed at meeting human (and societal) needs through land transformation or adapting the built environment to the features of the territory. Therefore, currently built spaces can be categorized into three main classes of elements:

• “Natural environment”, including natural boundary conditions (e.g., geological/geotechnical, climatological).
• “Users” that live and interact with the built spaces (inhabitants, tourists).
• “Built environment” as the system of technical choices and their implementation in constructing cities.

When cultural and/or landscape relevance are features of the system, the physical sphere should include all the elements that are usually associated with the “traditional” built environment. Here, the original choices of the place, their uses, and technical solutions reflect local human experiences and environmental adaptation, giving rise to a distinctive urban habitat. This is the case of historic buildings organized in a specific part of urban land with a unique significance for people, usually identified as “historic district” or “urban ancient core”. Such Cultural Built Environments (C-BEs) are analyzed to highlight which tangible and intangible elements are involved in the risk assessment. Focusing on the three determinants of risk, all the buildings in C-BE, as single fabrics and their system, represent the physical exposed elements; on the other hand, the social level of exposure requires a critical analysis that takes into account also the touristic relevance of such places. For each of these elements, including “built,” “socio-spatial,” and “environmental” components and features, the assessment is conducted.

In this framework, this section details the recurrent and relevant parameters to be considered for the physical vulnerability assessment of C-BE starting from the discussion of physical and social declinations. As discussed in the previous section, MAH scenarios are selected with a focus on the possible hazards in the Apulia Region. Natural and human events are selected as recorded in the national and regional database, and namely:

• ROD: among the hydrogeological occurrences classified in [19], floods, landslides, and earthquakes are recognized as possible natural RODs, as resulting from national and regional classification [20,21]. Terrorist actions as humans-induced ROD is selected, due to the increasing cultural attraction of the Apulian territory [22], above all in spring and summer.
• SOD: Heatwave is chosen as a SOD caused by meteorological events [19]. The current attention to the Mediterranean area in the climate change process [23], as well as for the increasing likelihood of extreme events in the south of Italy in future climatological scenarios [24], includes the Apulia Region in such exposed territories. However, the main attention is on the amplifying impact resulting from the combination of extreme meteorological events (e.g., heatwaves) and morpho-constructive features of the built environment and uses, which increase locally the temperatures, recognized as the urban heat island effect.

Following this introduction, a brief overview of features and components involved in individual risks is discussed, considering their relevance to the risk determinant.

Table 1. Summary of parameters involved in the qualification of Cultural Built Environments in Apulian MAH scenarios, classified according to “Natural Environment” (orange), “Built Environment” (blue)and “Socio-Spatial” (green) parameters and the determinant of risk (H hazard, V vulnerability, and E exposure) and to the dimension of vulnerability (p physical, s social).

SOD	H	V	E	ROD	H	V	E
Natural Environment Parameters				Natural Environment Parameters
[S.E.1] Climate	x			[R.E.1] Geo-morphological features (intrinsic or amplifying) of the terrain	x
[S.E.2] Environmental boundary conditions	x
Built Environment Parameters				Built environment Parameters
[S.B.1] Morpho-typological features of district or blocks		p		[R.B.1] Morphological asset of the district		p
[S.B.2] Material, constructive and technological features of buildings		p		[R.B.2] Material, constructive and technological features of buildings		p
[S.B.3] Geometric and constructive features of outdoor areas		p		[R.B.3] Geometric and constructive features of outdoor areas (boundaries)		p
[S.B.4] Geometric data of buildings		p		[R.B.4] Geometric data of buildings and components		p
				[R.B.5] Type of accesses for geometric and protective features	x
				[R.B.6] State of conservation of buildings		p
				[R.B.7] Presence of special buildings		p
				[R.B.8] Presence of obstacles and urban furniture			x
[S.B.5] n. of buildings and places exposed to overheating			x	[R.B.9] n. of unusable buildings after critical event			x
Socio-spatial Parameters				Socio-spatial Parameters
[S.H.1] Timely availability and accessibility of risk-related and lifesaving information		s		[R.H.1] Timely availability and accessibility of risk-related and lifesaving information		s
[S.H.2] Socio-demographic and socio-economic profile of the affected population		s		[R.H.2] Socio-demographic and socio-economic profile of the affected population		s
[S.H.3] Purpose and actual use of land, buildings, and open spaces		s		[R.H.3] Purpose and actual use of land, buildings, and open spaces		s
		s		[R.H.4] Social relationships between agents and attachment to places		p/s
		s		[R.H.5] Dynamic spatial interactions between agents and the built environment		p/s
		s		[R.H.6] Roles and responsibilities for each agent		s
[S.H.4] Agents’ direct or indirect knowledge of specific risk dynamics		s		[R.H.7] Agents’ direct or indirect knowledge of specific risk dynamics		s
[S.H.5] Permanent or temporary impairments, abilities/disabilities, and health conditions		s		[R.H.8] Permanent or temporary impairments, abilities/disabilities, and health conditions		s
[S.H.6] Risk perception and reaction to danger and disruption		s		[R.H.9] Risk perception and reaction to danger and disruption		p/s
[S.H.7] Number of people exposed to SOD		s	x	[R.H.10] Number of people exposed to ROD		s	x

summarizes these elements in a matrix of reduced parameters.

2.1. Environmental Parameters for RODs and SODs

Regarding the features belonging to the “natural environment” class, geomorphological characteristics and climate are the main elements to consider in the frontier conditions. Natural events referred to as ROD are thoroughly discussed in national regulations for the hazard, linking features of the ground with effects of previous hazardous events.

The Italian seismic classification maps categorize the entire territory into four classes, based on measured ground acceleration, as well as geological and geomorphological ground features [25]. A lower hazard is associated with the higher class and vice versa.

The national hydrogeological characterization operates at a lower scale and involves regional bodies in classifying regional territories using geolithological, physiographic, and hydrological (both surface and underground) characteristics. The main objective is the assessment of hydrological phenomena and possible consequent landslides, flooding, or subsidence effects. Unlike seismic classes that encompass wider parts of the urban extension, hydro-geological hazards refer to specific sub-parts of the city according to the described features. Thus, the study of historic centres exposed to such phenomena requires a detailed description of its position and the exposed area. However, for such hazards, the characterization focuses on the geo-morphological features of the terrain [R.E.1].

As human-induced occurrences, terrorism events cannot be related to specific “environmental” characteristics. The “boundary” features of the phenomenon are associated with political or religious discussions that typically involve broader territories (e.g., differences among countries or political involvement in international wars) [26]. Thus, any environmental feature can be introduced in the discussion of the “natural environment” class.

As previously introduced, the combination of regional/sub-regional climate and urban district features may locally exacerbate the overheating effects of the urban heat island. While heatwaves originate from macro-climatological processes transcending regional or urban dimensions, geographic location becomes a discriminating factor when comparing cities situated differently in national or international lands due to predominant climatic features. This relates to the climatic classification [S.E.1] of the global territories where the main climate conditions, such as summery temperatures and mean seasonal relative humidity, can intensify rising temperatures during such phenomena. The Apulia Region, owing to its diverse orography, exhibits various climates according to the Koppen–Geiger classification [27]. For instance, the promontories of Gargano in the northeast and the sub-Apennine areas in the northwest feature an oceanic climate. Along the coastline and hills, humid subtropical (Cfa) and hot summer Mediterranean (Csa) climates prevail. These climates, such as Cfa and Csa, are examples of critical starting conditions for summer where mean maximum temperatures are higher than 30 °C. Moreover, on the micro-climatic scale, the presence of green or blue areas near the district can locally affect the boundary conditions [S.E.2]. Thus, the adjacency of the water basin or green areas to the urban districts can affect the local micro-climate (e.g., for decentralized cores).

All the described data are featured by the different scales of focus but are available for a first qualification of the Apulian Region.

2.2. Built Parameters for RODs and SODs

Focusing on seismic phenomena, the assessment of risk for the built environment is strictly related to the analysis of its vulnerability, relating the seismic effects on buildings and the human uses of external places. Its evaluation may change in relation to the level of detail. According to the goals of the project, the assessment of the C-BE prone to seismic events must examine both the district and the building scales; this includes all the features related to the construction techniques and prevalent materials that usually depend on the construction period and the prevalent differences derived along their historical period of function that influence single buildings. To support these goals, C-BE parameters involved in the seismic risk are founded on smart approaches instead of detailed mechanical models. The Italian CARTIS sheets [28], as well as simplified approaches to assess the macro-seismic vulnerability of historic buildings in wide districts [29,30,31,32], reflect some recurrent elements to consider for the goals:

• The relation among buildings within blocks, thus as aggregate or isolated ones, aiming at studying their interaction during events (as a single building’s movements or effects on adjacent buildings) [R.B.1].
• Geometric data of buildings, as well as their morpho-typological classes, which encompass the number of floors, height, and surface footprints. Additionally, geometric regularities in plan and vertical development are included [R.B.4].
• Technological information related to construction techniques and materials. In this regard, masonries and floors (or vaults) are prevalent sub-systems to focus on due to their mechanical qualities and interactions [R.B.2].
• State of maintenance of buildings and their components (floors, walls) is evaluated to assess the final mechanical quality of structures. This assessment also considers the state of use/disuse of buildings. Furthermore, the presence of anti-seismic devices can be included in the assessment of parameters [R.B.6].

When systems of buildings are considered, geometric characters related to buildings and/or their aggregates with external areas should be discussed. In detail, the relation between street width and building height can impact evacuation processes and human safety during a seismic event in the external areas due to debris development [32]. In this case, these geometric data link the inherent vulnerability of building walls (construction typology, connections between walls and walls and ceilings) to the morphological features of districts [R.B.3], which are prevalent characteristics of the boundaries of outdoor places (streets, squares).

Moving to the hydrogeological risks, the main building characteristics pertain to interactions between building and soil. Significant damage to buildings results from the collapse or the vertical/horizontal (or combined) movement of buildings caused by flood pressure or soil movements. Recent studies discussing the vulnerability of buildings during floods and landslides [33,34,35,36] highlighted specific recurrent points related to building parameters involved in hydrogeological events, namely:

• Geometric data, including the height and number of floors of buildings [R.B.4]. For floods, the height of the basement floor decreases the likelihood of flooding during the event, while a higher number of storeys increases the vertical resistance to horizontal actions of floods. Instead, geometric factors have less relevance in landslide vulnerability.
• Construction data and material information [R.B.2]. Concerning floods, the continuity of wall increases the flow pressure over larger surfaces, thus, the type of construction techniques and materials affect the overall resistance to external pressure. Similar considerations should be applied to foundations, where the relation between the basement and terrain, as well as their construction technique, can influence the stability of buildings or their parts. These parameters also impact the vulnerability of buildings to landslides, while structural types influence the stability of buildings and some of their parts during the terrain instability.
• Finally, the conservation state of building (foundation and masonries) influences the vulnerability due to a loss of the inherent resistance of the structure [R.B.6].

As the last ROD event to discuss, the terrorist threat and its overall mechanism in risk assessment were not extensively discussed in scientific fields, particularly in Europe. The last studies that referred to such events are discussed [37,38]. Here, the parameters of the built environment are examined to describe and classify the most used mitigation strategies. Moreover, a first recognition of elements or uses in the built environment that can interfere with the terrorist risk assessment in terms of vulnerability, hazard, and exposure is examined. As discussed in the previous section, this type of disaster is related to the intentions of perpetrators who seek to maximize damage or mass media impact through their attacks [26]. Consequently, the parametrization process necessitates a multi-scale approach to account for the presence of such uses for places or buildings, as well as for their cultural or social significance, representing the “attractors”. Specifically:

• If attractors hold significant value for terrorist events due to their inherent significance (special, strategic, or public uses), the external area represents a vulnerable part of the district [R.B.7]. The relation between external areas and exposed buildings is introduced in Li Piani [39], identified as the space of influence (SoI). However, any reference to the extension of such external areas is still discussed in the literature.
• The presence of control systems along the perimeter of the attractor element can influence the intentions of perpetrators in attacking the place. In this case, control systems, considered as “additional” elements in the built environment (e.g., closed circuit television, remote control), constitute a parameter that affects the hazard potential of the place [R.B.5] [37].
• Specific attention is required for obstacles in the BE, including all built components of places that are not directly related to buildings but are part of the real places. Urban furniture, monuments, dehors, and fences are some examples of the inherent capacity to be an obstacle. In line with the other risks, these elements affect the evacuation processes (exposure) of users involved in the attacks [R.B.8] [38]. On the other hand, if obstacles serve as meeting points for inhabitants and tourists, they can be classified as “attractors” increasing the vulnerability of areas where they are located.
• Finally, the morphological configuration of places and areas can be considered as the inherent parameters influencing the vulnerability. The openness of squares or open spaces to external forces, such as vehicles or car bombings, due to their geometric characteristics, constitutes the inherent potentialities for these places to be targeted by perpetrators [R.B.3] [37].

For all the RODs, the morphology of the district as well as its part constitutes an inherent parameter of vulnerability for the place [40,41]. Focusing on the emergency phase, the geometric features of streets and accesses affect the speed of the evacuation process. Here, the morphological characteristic of the district impacts the significance of exposure by increasing the number of people involved in disasters; on the other hand, it represents the inherent vulnerable quality of the district or its part.

Moving to the SOD events related to climate change, specifically amplification parameters that enhance the local urban heat island effect, the parametrization process follows previous literature activities. However, for C-BE located in the Mediterranean area, previous work by Cantatore and Fatiguso [42] already identified the characteristics of the built environment that affect the urban heat island intensity as follows:

• Morphological features of the district, in terms of sky view factor and shape ratio values of open spaces [43]; both the parameters are related to the openness to the direct solar radiation or the closeness to the solar reflective process among the district surfaces [S.B.1], which also depend on building geometries [S.B.4].
• Thermal and optical properties of district surfaces exposed to environmental stress; in this case, external walls and roofs of buildings [S.B.2] and pavements in open areas (square, streets) [S.B.3] influence the capacity to reflect longwave radiation or to absorb and re-emit heat to/from the massive elements. This effect participates in the overall increase in surface temperature, impacting both the outdoor and indoor comfort of pedestrians and inhabitants, respectively. Thus, these characters are useful in determining the inherent vulnerability of external areas in the districts and inner spaces of buildings, affecting energy needs and health of occupants.

Regarding data availability, all these characteristics are closely related to the physical elements of C-BE, at all scales, and their categorization requires a complex phase of collection of data. This passes towards in situ or documental analysis of places and buildings for the establishment of detailed archives.

Finally, the physical dimension of exposure to ROD and SOD events can be summarized in terms of the number of buildings, infrastructures, or places that are overloaded or critically damaged after the events. Specifically, the exposure can be explained as the number of buildings and places exposed to critical overheating after the combined assessment of heatwaves and the urban heat island effect [S.B.5], as well as the number of unusable buildings after the critical event [R.B.9].

2.3. Socio-Spatial Parameters for RODs and SODs

In assessing vulnerable areas within Cultural Built Environments (C-BEs), previous risk assessment studies predominantly focused on analyzing the physical attributes of the BE. This approach, at times, resulted in inaccurate or misleading mitigation decisions [44]. Instead, to make informed choices for risk management and heritage preservation, it is imperative to investigate the socio-spatial aspects of vulnerability. The objective is to enhance the analysis of historical C-BEs when confronted with risks stemming from both natural and man-made RODs and SODs. This need persists despite the availability of disaster plans and preparedness, as a deficient understanding of actual human behaviour in space can adversely affect the emergency management process, rendering interventions at heritage sites increasingly complex. An illustrative case is when the population does not adhere to protective action instructions, contrary to the intentions of decision-makers. On a strategic level, there is a continuous need to reconcile the preservation of valuable sites without altering them while improving their safe utilization, especially during critical situations resulting from SODs and RODs.

Therefore, adopting a human-cantered perspective is crucial to enhance the planning of strategies, procedures, and operations aimed at preventing, controlling, and managing crises in complex built environments such as historical districts. This perspective applies to events with advanced notice, such as heatwaves and meteorological hazards (e.g., floods and storms), as well as unforeseeable events such as terrorist acts or accidents (e.g., explosions, chemical hazards, etc.). By concentrating on humans affected by disasters, we consider the environment not just as an inanimate backdrop, but as it is perceived, understood, and experienced by users. This approach allows us to analyze situations stemming from behavioural interactions between agents and spaces. Consequently, we can identify individual, social, and environmental factors influencing people’s decision-making processes and motivations for action. Moreover, comprehending how these factors impact subsequent behaviours in space, whether strategic, tactical, or operational, aids in evaluating how these interactions can either reduce or increase risks for individuals. Thus, in accordance with the existing literature [45,46,47,48,49,50,51,52], we propose a classification and explanation of fundamental parameters for consideration:

• Access to risk information and availability of timely lifesaving information [S.H.1–R.H.1]: This pertains to the availability of reliable, useful information sources to inform users about risk behaviours. During events, this can range from a dedicated early warning system for the public to a broadcasting system for recommendations or issuing instructions, ensuring prompt access to informal advice via media and social media through personal communication channels. All of these are intended to provide hazard-related cues to assess the severity of the situation and make appropriate or lifesaving decisions (e.g., whether and how to evacuate).
• Socio-demographic and socio-economic profile of the potentially exposed population [S.H.2–R.H.2]: this encompasses considerations such as gender, race, ethnicity, age, marital status, household size, and the presence of vulnerable individuals (e.g., children, the elderly, and pets), socioeconomic status, educational level, employment status, household income (e.g., income’s relation to the presence of air conditioning), housing type and size, vehicle ownership, and more.
• Knowledge of land use, building purposes, and open space utilization [S.H.3–R.H.3]: Together with information about the location of medical resources, facilities for the vulnerable (e.g., nurseries, kindergartens, and care homes for the elderly), this knowledge helps us understand how activity patterns specific to each place type affect across population groups. This understanding allows us to derive the spatial distribution of people and their accessibility to health services in case of emergencies.
• Social relationships between agents and the extent of attachment to places [R.H.4]: These factors encompass social ties, relationships, and group affiliations (e.g., families, neighbourhood and community groups) through which milling and collaboration or disagreement, negotiation, and time wasting may occur. Conversely, these factors may result in social isolation and marginalization, determining a lack of access to lifesaving resources. “Places” signifies attachment to property, but also familiarity with surroundings (e.g., building layout, routes, etc.), or it may account for the presence of users with limited familiarity with the local area.
• Dynamic spatial interactions between agents and the built environment [R.H.5]: This accounts for both social and spatial interactions among people and their interactions with places based on human spatial behaviour. It requires analysing interactions and interferences between evacuees and rescuers and their interactions and interferences with the environment (e.g., proxemics and spatial surroundings). Evaluating the level of crowding (e.g., changes increasing the number of involved people from dyads to groups and crowds) and the spatial characteristics of the affected area that affect human spatial perceptions (e.g., configurational analysis of the area to assess wayfinding task complexity for individuals and their tendency to follow others).
• Different roles and responsibilities for each involved agent [R.H.6]: This includes emergency managers, coordinators, rescuers, evacuees, etc. For instance, safety delegates trained in safety practices or group leaders responsible for relatives and friends.
• Either direct or indirect knowledge of specific risk dynamics [R.H.7]: This encompasses awareness of long-term risk probability and knowledge of short-term effects of risks. These can be direct (e.g., memories of previous emergency experiences) or indirect (e.g., through training in hazards and emergency management procedures). The ability to process information correctly from credible sources for inhabitants and visitors should also be assessed.
• Presence of people affected by permanent impairments, disabilities, or temporary health conditions [R.H.8]: this includes health status, pre-existing chronic or acute conditions, linguistic barriers, impairments (e.g., vision, hearing, strength, and mobility), and special needs individuals.
• Risk perception and reaction to danger and disruption [R.H.9–S.H.6]: This considers emotional states, risk aversion, subconscious reactions (e.g., shock and panic caused by unexpected events or impediments), judgment errors (e.g., overestimation of capabilities or underestimation of hazards), confidence, and self-efficacy to cope with danger and altered perceptions (e.g., distorted sensation of temperature or thirst).

Access to risk information and timely lifesaving information for heritage conservation and risk management in historic built environments can be achieved through a comprehensive approach. This strategy integrates technology, communication methods, and community engagement. Key components involve implementing early warning systems for the public, establishing broadcasting systems for recommendations and instructions, promoting community engagement through educational programs, leveraging social media for information dissemination, encouraging personal communication channels, and ensuring information comes from trustworthy sources. By adopting these measures, the operationalization of vital information becomes possible, enhancing the resilience of historic assets against natural and man-made risks while fostering a sense of responsibility and preparedness among the community [53,54,55].

Similarly, operationalizing the socio-demographic and socio-economic profile for risk management in historic built environments involves a systematic process. It begins with comprehensive data collection, incorporating aspects such as gender, age, socioeconomic status, and more. Geographic information systems (GIS) are then employed to map and visualize this data, aiding in understanding the distribution of vulnerable populations. Vulnerability assessments follow, using the data to identify at-risk groups and inform tailored interventions and preparedness strategies. Community engagement and education programs help raise awareness, and continuous monitoring and community feedback mechanisms ensure that interventions evolve to meet changing needs. This approach fosters more equitable and targeted risk management strategies, considering the diverse vulnerabilities and characteristics of the community [56,57,58].

Overall, this classification underscores that socio-spatial aspects extend beyond physical characteristics such as dimensional space requirements. Estimating the human factor should also encompass cognitive, motivational, and social variables, which contribute to complex social and spatial behavioural patterns and choices for each agent involved in emergencies. While assessing their combined and simultaneous effects on decision-making processes in dangerous situations may be extremely complex and yield various outcomes for different individuals, recent empirically based studies moved away from oversimplifying human spatial behaviour (e.g., likening it to that of a herd) and instead, aimed to develop more sophisticated and realistic behaviour models [59].

However, there remains a challenge in striking a balance between a realistic representation that may involve some risk and a limited but practical one. This is particularly crucial when the innovative objective is to comprehensively characterize the resilience of Cultural Built Environments (C-BE) by integrating multifaceted information from multidisciplinary analyses. The excessive computational burden of simulation models, their high demand for extensive datasets, and the difficulty of selecting only necessary and sufficient variables to avoid unrealistic outcomes can often impede their applicability or use. Therefore, finding a trade-off becomes essential. Overall, the pursuit of a comprehensive understanding of resilience in C-BE requires considering various aspects while acknowledging the complexities of human behaviour and the challenges in model representation.

3. Method and Tools to Assess the Physical and Spatial Vulnerability of Historic Cities Due to Multi-Hazardous Events

As discussed, the method aims to create multi-vulnerability maps of historic districts, focusing on the peculiar MAH scenarios in the Apulia Region. This is a fundamental first step to promote the assessment of C-BE prone to selected ROD and SOD for the identification of priority areas and associated mitigation strategies and solutions featured by a low level of impacts to the cultural and historical significance of these places.

The applied method considers the opportunity to establish a multi-layered system of information in a geographic information system (GIS) environment. This system allows for the study of places exposed to single-hazard and multi-hazard—but asynchronous—scenarios (here called MAH). The GIS model is designed to collect, reorganize, and correlate all the data involved in the vulnerability assessment for both single-hazard cases and in MAH scenarios, according to the proper scale. In fact, the approach involves the entire historic district extension (D), from the municipal-territorial to the detailed scale of the district and its essential units as Outdoor Areas (OAs—streets, squares) or buildings (B) and their components (SB—walls, roofs, etc.). All the required data are gathered starting from the parametrization process outlined in Section 1, and combining detailed information (wall materials, roof technologies, etc.) with complex data obtained from external analyses. Finally, the model became functional in translating vulnerability and critical levels of fabrics into properties of Outdoor Areas, thereby preparing for the assessment of public open spaces in MAH scenarios. In the Apulian cases, the process combines the analysis of the microclimate impact of climate change events and the inherent morpho-constitutive characters of the C-BE, the identification of areas of the C-BE that are vulnerable to natural and anthropic RODs and the assessment of heritage assets within the hazard zones potentially affected [60]. These zones are then related to the potential capacity to solve the emergency process, assessing the spatial vulnerability of the places where this spatial vulnerability results from the analysis of relations between their geometric and morphological features and the perception of asset by users, which can affect the emergency phase of sudden onset events.

Therefore, the method follows three interrelated steps, as shown in Figure 1 and described in depth as follows:

• Creation of the three-dimensional model of the historic centre and data collection. In this phase, the model is created as the basis of subsequent analyses; moreover, its scale refers to the fundamental units, as a digital surface model (DSM), starting from the shared data in the territorial information system (SIT) of the region [61]. All the required data useful for the analysis of the physical vulnerability of places, both in single and multi-hazardous scenarios, are catalogued in the model or homogenized if derived from external analysis. This phase connects all the following ones, constituting the main common thread in homogenizing data and results towards the creation of vulnerability maps, both for single hazardous conditions and MAH scenarios.
• Elaboration of single-vulnerability maps on different scales. This step involves the analysis of single hazardous events based on the characteristics of RODs and SODs and includes the following sub-steps:

  (a)

  For SOD, the analysis of interactions with a climate-built environment (heatwave, HW) is conducted using fluid dynamics models. This step aims to qualify micro-climatic variations in a district scale by combining climatic conditions, as well as geometric and material (affecting global performance) properties of the C-BE. However, it uses fluid–dynamic simulations, obtained in Envi-met^® software (version 5.0.2), for the identification of critical areas in C-BE within the GIS model as a specific layer for the HW vulnerability maps referring to OAs (HW_OA). Moreover, due to the relevance of heatwaves as climatological events, the study should test the case during previous extreme events.

  (b)

  For RODs, the assessment of vulnerable areas of C-BE in the emergency phase follows two main steps: the first analyses the C-BE combining the buildings and OA parameters involved in the vulnerability, as direct assessment of single risks referred to the proper scale [EQ_B—earthquake; HG_B—hydrogeologic; and T_OA—Terrorism]; the second one accounts for the district morphology in terms of emergency impact, moving the assessment of vulnerability from the building to the related OA in evacuation process for each single risk [T_OA,E; EQ_B,E; and HG_B,E]. Specifically, this phase determines the vulnerability levels of OAs considering:

  • Building vulnerability to seismic and/or hydrogeological [EQ and HG] events according to [29,33] methods, respectively. In these cases, the relation between building performances and OA is reached reflecting the proper level of building vulnerability outside them. Specifically, the related OA vulnerability maps result from the projection, along streets or places, of facades or buildings, respectively, for seismic and hydrogeological risk [vulnerability maps of EQ_B; HG_B].
  • The presence of representative or symbolic buildings or district areas exposed to the threat of terrorist events [T]. Here, the process is based on the identification of representative buildings or rendezvous, which constitute main touristic/cultural, political, or social “attractors” for perpetrators. Then, as discussed in Cantatore et al. [38], vulnerable OA areas are quantified and identified through two options: (i) the real perimetration of places when OAs are identified as open spaces (i.e., squares) or featured by special attractive uses (i.e., dehors); and (ii) the identification of the associated space of influence in the case of attractive buildings with special uses (e.g., museum, churches, etc.). As discussed in Section 2.2, there are no references about the extension of such spaces. However, for the study, its quantification considers the product between the commercial extension of the building (A_CommBuild) and the ratio between the maximum density capacity of people in the buildings (C_b) [pp/m²], as defined by the Italian decree D.M. 03/08/2015, and the maximum density of people when public activities are conducted outside (C_out), as considered for public buildings in the National Ministerial Decree 19/8/1996) (Equation (1)).

    A_SoI [m²] = A_CommBuild [m²] × C_b [pp/m²]/C_out [pp/m²]

    (1)

    The equation defines the relation between the maximum extension of the external area that can host people when special activities are supposed to be outside the building. Thus, in this second case, the space of influence is geometrically developed along the building facades with access/es and it completes the map of terroristic vulnerability in OAs [vulnerability map of T_OA].

  As well as the spatial vulnerability of the districts during single ROD concerns, open spaces and streets are examined through the space syntax (SS) analysis to define probability distributions of moving pedestrians inside the historic centre during emergencies [62]. This approach is proposed as a proxy to represent the pattern of human behaviours in space considering [R.H.4], [R.H.5], and [R.H.9] factors though in a derived way. This offers the opportunity to consider physical properties of C-BE in a socio-spatial aspect, combining geometric dimensions with the perception of places by users. Indeed, the space syntax process is based on a topological analysis of the graph representation of the planimetry of the study area to reveal how people tend to use it [49]. This analysis allowed us to correlate the arrangement and the reciprocal relationship of the segments of the road axes in their function of passage paths during the movement inside and out of the historic centre, which does not require a consideration of their staging/safety area capability. Such analysis was carried out with depthmapX software (version 0.5b, https://varoudis.github.io/depthmapX/, accessed on 3 August 2023).

  The assessment of the spatial vulnerability of single OAs (streets or places) in a linear way using SS maps determines, for each observation point, the possible way of evacuation during the emergency. Thus, the SS analysis becomes the key map to read single-risk maps [EQ_B; HG_B; T_OA] as critical ones in emergency scenarios [T_OA,E; EQ_B,E; HG_B,E]. In an operative point of view, this is the result of the map-overlapping process, considering the worst conditions in SS analysis, resulting from the statistical assessment of the entire built environment.

• Finally, the identification of areas with high levels of physical multi-vulnerability in MAH scenarios. Here, external analysis (heatwave and SS) and single-risk maps (hydrological, seismic, and terrorism) are combined in the GIS model and properly detailed in the OA scale, in order to identify areas with high levels of physical multi-vulnerability MAH scenarios. The resulting maps help determine spatial priority classes for vulnerable sub-district areas during potential ROD or SOD events, facilitating the assessment of mitigation solutions that are required to be checked in terms of compatibility with the places. Specifically, a first level of MAH scenarios is discussed in terms of two timely different hazardous events—(HW+T,E)_OA; (HW+EQ,E)_OA; HW+HG,E)_OA—into the emergency phase; the second one overlaps two different sudden events, generated by the different nature of external strengths (natural and man-made), (HW+T,E+EQ,E)_OA and (HW+T,E+HG,E)_OA.

4. First Results in the Pilot Case of Molfetta

As discussed in Esposito et al. [60] and in line with the need to provide a general methodology applicable to similar C-BEs with similar peculiarities, the historic centre of the municipalities of Molfetta (BA) was identified as a suitable case for the proposed research. In this work, MAH scenarios were tested on the historic centre of Molfetta due to the morpho-typological and morpho-distributive features of its buildings, as well as their arrangement, emblematic of the nearest coastal regional cities. As emerged from the preliminary analysis and according to the main aim of the work, all the parameters involved in the physical vulnerability are collected, following the summarized data previously listed in Table 1 and thus collected and analyzed for the case study (

Table 2. Summary of properties and features of buildings and places in the historic district of Molfetta according to the qualification of parameters involved in the physical vulnerability, “Built Environment” (blue) and “socio-Spatial” (green) classes.

SOD	ROD
Built Environment Parameters	Built Environment Parameters
[S.B.1a_Molf]	[S.B.1a_Molf]
Continuous and homogeneous district asset with long blocks, with narrows, parallel and flat streets	Continuous and homogeneous district asset with long blocks, with narrows, parallel and flat streets
[S.B_Molf]	[S.B.1b_Molf]
Residential buildings featured by vertical development (3–4 floors), organized in double rows (NNW—SSE axis)	Residential buildings featured by vertical development (3–4 floors), organized in double rows (NNW—SSE axis)
[S.B.2a_Molf]	[S.B.2a_Molf]
Massive masonries in squared and calcareous block, flat roofs (wooden mostly) with any thermal layer, one window for each floor	Massive masonries in squared and calcareous block, flat roofs (wooden mostly), constructive technologies homogeneous and continuous in the block, one window for floor
[S.B.2b_Molf]	[S.B.2b_Molf]
Clear calcarenitic for wall surfaces, bituminous/cotto tiles or calcareous paving for roofs, wooden windows	Clear calcarenitic for wall surfaces, bituminous/cotto tiles or calcareous paving for roofs, wooden windows
[S.B.3_Molf]	[S.B.3_Molf]
Clear, smooth and semi-shiny calcarenitic paving (“chiancarelle”), few green areas and two paved squares	Clear, smooth and semi-shiny calcarenitic “chiancarelle” for street, few garden areas and two open paved squares
	[R.B.4_Molf]
	Limited number of accesses to the districts, referred to the ancient City Gates with limitation of vehicular transit
	[R.B.5_Molf]
	Presence of buildings with medium-low State of conservation (static problems related)
	[R.B.6_Molf]
	Presence of special buildings (Cathedral, churches, coffee bar, Master Harbour Office di Porto)
Socio-spatial Parameters	Socio-spatial Parameters
	[R.H.4_Molf] Social relationships between agents and attachment to places
	[R.H.5_Molf] Dynamic spatial interactions between agents and the built environment
	[R.H.9_Molf] Risk perception and reaction to danger and disruption

).

The Municipality of Molfetta is characterized by a Mediterranean climate with hot summers—Csa type (Koppen-Geiger). Moreover, Molfetta has a low seismic hazard, but buildings in the historic district are featured by a medium-low static conservation status, determining interesting discussion about their vulnerability. As far as the hydrological hazard, Molfetta is featured by critical exposure to hydraulic risk in the northern part of the city. However, these areas are located at a considerable distance (>500 mt) from the analyzed historic area, excluding it from the risk analysis. Finally, as concerns the terroristic risk, Molfetta was analyzed for such violent acts due to the increasing trend of touristic fluxes determined for the higher cultural attractiveness of the city. Even though its political or economic significance may be considered lower than administrative centres in Apulia or other regions, the applicability to such a case study is representative of the method’s replicability. Moreover, according to [38], the study of terroristic events for outdoor public places may be independent of the religious, political, or economic relevance of the city/region, which could constitute boosting factors of the whole hazard exposure. In addition to its touristic vocation, enhanced by the presence of the “Duomo” Cathedral as the main cultural attraction, the historic district of Molfetta has several gathering spaces as meeting points of variable classes of users during the day. Additionally, the presence of the harbour master’s office along the western dock enhances the strategic significance of the case study to terroristic events.

It is essential to note that, unlike physical vulnerability, socio-spatial vulnerability parameters suffer from greater variability and lack of data.

Traditional emergency spatial analyses focus mainly on quantifiable environmental and physical elements, which can be approximated. In contrast, detailed information about the population and users, let alone their dynamic behaviour, is considered neither relevant nor measurable. Consequently, for the present study, it was assumed that the vulnerability parameters [R.H. 1, 2, and 3] are partially met and, therefore, do not constitute significant vulnerability aggravation to be considered in the conducted analysis. In fact, there is a municipal emergency plan [53] for disaster management, which includes indications of safe waiting and reception areas. Additionally, some communication campaigns were activated for its dissemination and a real-time alerting service via smartphones are active and available (https://molfetta.infoalert365.it/infoalert365/, accessed on 3 August 2023). Moreover, data on the resident and transient population are collected annually, albeit spatially aggregated and not distributed by type or temporally differentiated within the year. Information on the use of open and enclosed spaces is fully available but only accessible to decision-makers and experts. On the other hand, parameters from [R.H. 6 to 8] are not assessable at all since no open data are available, and this would require ad hoc survey campaigns. Finally, the space syntax analysis developed aims at partially account for [R.H. 4], [R.H. 5], and [R.H. 9] factors, as outlined in Table 2 and described in the next paragraph.

Application to the Case Study

A three-dimensional model of Molfetta was generated starting from the data available from the Geographic Information System of the Apulia Region [61], in order to have a topographical base in the simulations conducted in the following section. The base map used was the 2019 orthophoto provided by the Apulia Region via a web map service. The 1:5000 scale digital cartography (characterized by metric accuracy) was used to create a preliminary digital surface model (DSM) of the study area. The analysis in the GIS environment allowed for implementation of the geometric model of the data collected and parameterized in the preliminary phase. This primarily included information about the district’s assets, geometric characteristics of buildings and blocks, parameters influencing the seismic vulnerability of buildings, technological and material information of the built envelope, and the main uses to recognize users’ attractors. All these data were employed for the single-hazard analysis according to the methodology, creating specific maps of vulnerability in the GIS environment. These single-vulnerability maps were integrated in the same reference frame, namely RDN2008—UTM zone 33N to deepen the relation between the different potential threats.

Subsequently, the fluid dynamics results for the assessment of the effect of the urban heat island in the district were analyzed, starting from data acquired in the local micro-climatic survey in 2017 by the authors [42]. Specifically, the results highlight the resistive–adaptive behaviour of compact C-BE during the summer middays (Figure 2—from 13:00 to 16:00), reflecting significant differences between environmental temperatures (T_env) and simulated ones in all the 13 reference checkpoints (Figure 3a). On the other hand, the urban heat island effect is maximized during the nighttime hours, particularly in public open areas (OAs) (Figure 2—from 20:00 to 04:00, and Figure 3a), due to the combination of the direct solar exposure during the day, the reduced ventilation, and the high thermal capacity of the flooring material (calcareous).

The higher temperatures in OAs during the nighttime highlight different levels of outdoor comfort for pedestrians. On the other hand, during the day, daily temperatures and solar exposure identify different classes of indoor comfort and health criticalities for inhabitants [42]. According to the method, the fluid–dynamic results are implemented in the GIS environment. Thus, critical areas are identified in the model, considering the points affected by higher temperatures than environmental and narrow-street ones (T_env = 28 °C; T_nstr < 28.65 °C) (Figure 3b). The vulnerability of the compact district is assessed by comparing the adaptive behaviour of narrow streets and the physical variation in the district morphology [42]. This process led to the creation of the heatwave (HW) vulnerability map for Molfetta, emphasizing vulnerabilities in open areas (Figure 3b).

In line with the methodology, the space syntax analysis is conducted using depthmapX Version 0.8.0 software to assess how people perceive the entire C-BE. Specifically, this approach is proposed as a proxy to represent patterns of human behaviour in space, considering factors [R.H.4], [R.H.5], and [R.H.9], albeit in a derived manner. The space syntax process relies on a topological analysis of the graph representation of the study area’s planimetry, revealing how people tend to use it [63,64]. The normalized angular choice map shown in Figure 4a represents, on a scale ranging from red to blue, the road segments preferred by pedestrians when moving from any point to another within the area. For the purposes of the present work, this map highlights critical evacuation routes, as well as roads with the highest probability of becoming crowded due to pedestrians fleeing from the historical centre and rescuers coming in the area. As depicted in Figure 5a, these critical routes are located on the spine axis and the perimeter roads along the ancient defensive walls, leading to the outside through the main gateways. From this perspective, the space syntax analysis indicates the user’s condition of exposure during and immediately after a disaster, becoming a relevant element for assessing and representing the potential vulnerability distribution of the historic planimetry’s configurational asset based on the use of urban space by people.

Considering the main scope of the work, which relates rapid events to perceived spatial vulnerability, the space syntax outputs were analyzed to determine their significance and the potentialities in combining them with seismic and terrorism analysis. All data resulting from the space syntax process underwent statistical analysis to determine their socio-spatial relevance for the C-BE of Molfetta. The choice values are divided into twenty homogenous classes, and the percentage frequency of each class is calculated based on the number of segments within each class. This value was then multiplied by the average choice value of each class, and the results are presented in Figure 4. In detail, the graph, plotted on a logarithmic scale, clearly indicates a sign inversion at the choice value of 0.7125, revealing it as a significance threshold. Consequently, it is selected as the threshold value for the second part of the analysis, where continuous results are mapped within the model based on two main scales (<0.7125 and >0.7125) (Figure 5b).

Following the third phase of the process, the mapping of the seismic vulnerability of buildings was carried out. In detail, this focuses on the main parameters involved in the vulnerability assessment identified in Lagomarsino and Giovinazzi and Quagliarini et al. [29,65], along with recurrent elements related to construction and material features, as well as the state of conservation. Several recurrent and discriminating parameters are found to discuss the systemic buildings (e.g., residential) in the district and specifically:

• All the buildings are featured by resistant masonry made of squared blocks and erected according to the compound technique [66] with a few variations of thickness (from 60 to 80 cm). Referring to floors, barrel vaults on the ground floors and wooden planar ceilings represent the common horizontal sub-components of buildings.
• Tower and palace houses are the predominant morpho-typologies of civil construction in the district. These buildings typically have three to four floors, maintaining a relatively constant total height of the block facade. Moreover, the ratios between openings and massive walls along the frontier have similar ranges, varying from 8 to 12%.
• All the buildings are included in systems of long blocks and each buildings contributes to the overall structural resistance of the block; moreover, it is not in contrast with the homogeneity of construction technique [65].
• Referring to the state of conservation, buildings are featured by three main different levels of technological vulnerability strictly related to their use and maintenance conditions. These levels correspond to the previous process of recovery in the district where (i) a good state of conservation is associated with the wider condition of civil structures, (ii) a medium-low one for buildings was already recovered but remained unused, and (iii) a low level for those affected by a critical level of masonry conservation, and lack of roofs and windows, remaining unused. Moreover, for the latter, the compound walls have a serious level of criticality in the inner layer, made of non-coherent materials. On the other hand, special buildings, such as churches, recently underwent interventions for static recovery and are featured by a high level of conservation.

Based on this analysis, the seismic vulnerability is assessed using the methodology described in [29] using the GIS database enriched with the data. Two classes of vulnerability are recognized for the analyzed case study, revealing in the conservation state the discriminant feature (Figure 6a). Specifically, the classes L1_b and L2_b represent ascending levels of seismic vulnerability, considering the building’s capacity to either generate damages (such as facade rotation and debris production) (L1_b) or not (L2_b). This is fundamental for establishing the correlation between the vulnerability of buildings and outer places.

In fact, for each class of Lx, the associated level of vulnerability is rated to the related part of streets, squares, or open areas (L1_OA or L2_OA) and specifically (Figure 6a):

-

The projection of the façade when the geometric ratio h/W < 1 (h is the height of buildings and w width of the street); this is the case of Amente Square, which is spatially affected by the global rotation of building facades in its northern part.

-

When h/W > 1, a specific area is identified, characterized by the same length as the vulnerable building and a width equivalent to the height of the street. This is the case of all the narrow streets within the analyzed C-BE delimited (by one or two sides) by buildings affected by a critical level of conservation.

Results of vulnerable areas to building damages after a seismic event were combined with the SS results, with the aim of highlighting the most vulnerable OAs. Figure 6b shows the overlapping outcomes of this process, emphasizing the higher significance of narrow streets affected by the façade rotation. On the other hand, the lower spatial vulnerability along the square minimizes its inherent vulnerability.

Considering the vulnerability of the district related to terrorist events and the large scale of analysis, the hazard is assessed by focusing on the presence of special buildings, their uses, and their social and cultural significance. Most in detail, all the cultural attractors are identified, thanks to the regional categorization of cultural heritage [67]:

• The “Duomo” of San Corrado was built during the XII and XIII centuries along the ancient city walls, with its current appearance being the result of an extension process in 1400–1460. This construction represents the Apulian Romanesque style and is a major religious attraction for tourists and inhabitants.
• The “Dogana Vecchia” palace was built along the city walls in the western part of the ancient city. Its final morphological extension is the result of major intervention during the XVI century when it became the Seminary of Molfetta and adjacent cities, and later, the Customs (XIX century). The actual relevance of the building is reflected by the creation of a luxury resort that attracts wealthy visitors.
• Palace “Giovene” is located at the eastern square of the ancient district. It is an XV century fabric, made with calcareous blocks for two floors, and it is very representative due to its architectural decorations on the main façade. Its categorization for terroristic events depends on the presence of the Town Hall.
• Several other palaces enrich the cultural relevance of the district, including “Tattoli”, “Giovanni De Agno”, “Riganti”, and “Ribera”. However, they are less relevant architectures for the attacks, also due to their private ownership.
• The “San Pietro” church, located in the eastern part of the district, is a baroque construction inserted in the compact ancient core. Its relevance is relatively low due to its position in the district and its minor attractiveness to tourists.

However, since their main residential destination of use, Pappagallo, Ribera, Riganti-De Agno, and Tattoli Palace are excluded from the assessment.

In addition to architectural cultural heritage, vulnerability to terrorist attractions for Molfetta is also related to the presence of the Harbour Master Office. While not distinguished by peculiar architectural features, it serves as a significant political attractor for potential perpetrators.

Finally, the assessment of vulnerable places in the district ends with the study of social attractors. In this case, locations hosting public services and nightlife are considered. Municipio and Amente Square are characterized by the presence of nocturnal and external services (bar, pubs) as well as benches for social meetings. Similar attention is given to the accessible docks situated in the western and the north-western parts of the districts, along the borders of the C-BE. Finally, four bars with external activities (dehors) are identified as special social attractors: two in the western area, one in the northern dock, and the last in Municipio Square.

Table 3. Summary of cultural buildings and attractive places with details in uses and area extension or space influence SOI.

	Use	Internal Area [m2]	n Floors	Crowding Rate (pers/m2)	Max Outer Density [pers/m2]	ASoI [m2]
Cultural buildings
Capitaneria	Public use/military use	1480	1	0.4	1.2	493.35
Dogana	Touristic/residential	751	4	0.4	1.2	1001.35
Giovene	Public offices	452	2	0.4	1.2	301.35
Pappagallo	Private residential	excluded
Ribera	Private residential	excluded
Riganti-De Agno	Private residential	excluded
San Corrado	Church	813.59	1	1.2	1.2	813.59
San Pietro	Church	358	1	1.2	1.2	358.00
Tattoli	Private residential	excluded
Torrione Passari	Exposition, public use	94.67	2	1.2	1.2	189.33
Attractive Use
Bar Municipio Square	Commercial	67
Bar 1 Duomo	Commercial	124
Bar 2 Duomo	Commercial	228
Bar Dock 1	Commercial	25
Open Spaces
Amente Square	Public use	458
Municipio Square	Public use	1135
Dock 1 (west)	Public use	3441
Dock 2 (north)	Public use	4421

provides an overview of the total extent of vulnerable outdoor areas for the considered buildings, while Figure 7a shows the location of sensitive buildings and places and the associated relative space of influences. Consistent with the methodology, areas vulnerable to terroristic attacks were combined with the spatial data in emergency conditions. Figure 7b summarizes the overlapping process, showing the higher relevance of open areas for the hazard. Both the western and northern docks and squares are affected by a higher level of overlapping, making them the most relevant spaces to study and design. However, the vulnerability maps also underscore the greater susceptibility of these locations to terrorist events due to their openness, aligning results to previous general studies [38].

Finally, these single-risk maps were combined in order to identify the most critical OAs in MAH scenarios for the case study: (HW+T,E)OA (Figure 8), (HW+EQ,E)OA (Figure 9), and (HW+T,E+EQ,E)OA (Figure 10). The process uses the information retrieved by each threat combination to evaluate the additional effect of the HW by performing a spatial analysis in the GIS environment. As shown in the figures, the overlapping process highlighted the vulnerable areas in the MAH scenarios identified for the historic district of Molfetta. These results provide several key insights into the matters.

Considering the combination of ROD and a SOD, both the analyzed scenarios evidence specific areas to focus on.

Open spaces such as Amente Square, the northern dock, and Municipio Square (Figure 8) emerge as significant focal points. These spaces are of higher importance for both sudden and slow events due to their relevance in terms of thermal discomfort for users and the potential for emergency situations, particularly in the context of terrorist threats. On the other hand, inner walkable streets, due to their compactness (as an inherent quality of redundancy and adaptability of the Mediterranean traditional asset) offer an important point of discussion as an inherent value in a multi-disciplinary point of view. While they may be excluded in specific considerations about mitigation strategies, they emphasize the need to explore user behaviour and their preference for cooler areas, which could increase potential exposure values in hazardous scenarios.

An opposite resulting scenario can be described when earthquakes and heatwaves are combinedly assessed. The higher distribution of “special buildings” in terms of morphological construction types and their good level of conservation, derived from continuous interventions at the public level, move the attention within the walkable areas (Figure 9). Here, despite similar geometric features along the eastern part of the central axis, what is discriminant in this MAH scenario is related to the lower state of maintenance. This is particularly evident in areas under private management, which lack prioritization for interventions. In this MAH scenario, Amente Square constitutes an exception. This is not a consequence of the geometric features of the place, but it results from the exceptionality of lacking special buildings. In fact, Amente Square is delimited by private buildings characterized by a low state of maintenance, as the prevalent and recurrent features of seismic vulnerability are combined with higher exposure to potential nighttime discomfort.

Such conditions converge towards the final MAH scenario (HW+T,E+EQ,E)_OA (Figure 10), emphasizing the lower level of resilience for Amente Square. Several factors contribute to its higher vulnerability in this MAH scenario, including limited shading during the night, high thermal capacity of pavements, low conservation levels of private buildings, and the strong attraction of the square for users. In that sense, Amente Square results in the representative OA of the historic centre of Molfetta to which mitigation activities should be properly assessed in a MAH point of view coherently with the inherent significance of the work.

5. Conclusions

In the present study, the assessment of Multi-Asynchronous Hazard scenarios in C-BE was based on qualitative analysis. The proposed analysis aims at identifying the spatial priority of interventions by studying inherent qualitative vulnerabilities to single ROD and SOD and in their specific combination, considering natural and anthropic hazards in the Apulia Region. This first level of assessment is functional in a most comprehensive goal aimed at improving emergency plans, designing effective and suitable mitigative solutions, and establishing requirements for long-term urban planning choices from a multi-risk perspective. Moreover, even if the work starts from single hazards, the standard vulnerability assessment goes beyond their formulation at the building scale towards a wider one. Thus, all single events are combined to explore potential scenarios of multi-vulnerability when different events may occur. This approach recognizes the importance of incorporating spatial information at the district scale as a key factor in assessing the consequences of vulnerability in a pre-emergency process. It also involves considering potential choices made by users residing in these places. In detail, the space syntax analysis supports the goal starting from the spatial configuration of places (i.e., the planimetric layout of the area and the spatial distribution of the built elements), which is peculiar in historically unplanned areas, as the ancient cores, and within Mediterranean area.

The complex planimetric distribution and compactness of such C-BEs constitute the most interesting points of discussion for several hazards. For Slow-Onset Disasters such as heatwaves, the local increasing temperatures trend (urban heat island effect) takes advantage of such peculiarities, moving the criticalities towards open public places, i.e., squares and docks. On the other hand, such compact district sub-areas and their open public places must be considered with the buildings’ uses and state of maintenance for sudden onset disasters. As demonstrated in the case study, squares usually host attractive buildings for cultural, religious, or political relevance, making them central in altering the inherent levels of vulnerabilities to terroristic events and in combination with other hazards (MAH scenarios). Oppositely, when special buildings have high cultural and touristic appeal for the city, they often exhibit a high level of conservation, reducing vulnerability at the fabric scale and associated outdoor spaces. This note moves the attention towards narrow streets where the seismic OAs’ vulnerability is combined with the higher level of the spatial one.

In line with the overall project goals, these preliminary investigations represent the first steps in addressing the problem and will be extended to and compared with other case studies. Thus, in next steps, the hydrogeological vulnerabilities could be investigated and the multi-risk compatibility of mitigative solutions will be analyzed and verified, also coherently with the historic urban landscape approach.

The preliminary results of these analyses underscore the potential of using a digital representation of the urban centres (medium resolution digital models in this case study) to drive and constrain different kinds of surveys. Moreover, it allows for verification through the integrated approach, the real gap between traditional planning systems, mitigation solutions and emergency plans, and innovative tools and methodologies based on multi-risk scenarios. Mitigative and resilient solutions, as well as emergency plans, require comparison to real urban areas of historical–cultural relevance, which should have the highest priority in the implementation of proper tools for their preservation, safeguarding the real uses of places and analysing the potential exposure levels of critical public open areas. In that sense, the design of mitigative solutions and emergency plans requires agent-based simulations in trying to solve the real information about users’ behaviours and their familiarity with the places (e.g., inhabitants and tourists) [68]. On the other hand, all technical outcomes resulting from scientific and technical experiences should be shared with local administrations to communicate the potential risk exposure, show the potential results, and start virtuous planning activities to improve local awareness and readiness of public bodies in emergency. With similar weight, the assessment of risks, their impact, and solutions require to be communicated with local users’ enhancing their responsiveness coherently to the wider goals of resilient communities as part of cities [69,70].

These further results are in line with current practices in risk communication through digital models of the cities. The use of the GIS-based model in the present study facilitates the collection, understanding, and assessment of spatial and physical features that influence the vulnerability of places. Moreover, the digital model can be easily implemented with parametric ones aiming at structuring thematic platforms for managing technical data and planning [71], monitoring of effects of natural phenomena [72,73], and setting up a digitalized environment for risk communication and promotion of mitigative solutions [69,74], encouraging collaborative strategies [75].